Management Plan and Public Outreach for
WWII Submerged Resources in Saipan
Jennifer F. McKinnon and Toni L. Carrell
2014
Cover Photographs by Jon Carpenter 2012
Management Plan and Public Outreach for
WWII Submerged Resources in Saipan
Grant Agreement No. GA-2255-11-018
Final Report
Jennifer F. McKinnon and Toni L. Carrell
2014
This material is based upon work assisted by a grant from the Department of the Interior, National Park Service. Any opinions,
findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the
views of the Department of the Interior.
Copies can be obtained from:
Kristen L. McMasters
Government Technical Representative
National Park Service
American Battlefield Protection Program
1201 I Street NW (2287), Washington, DC 20005
Title page photograph by Jon Carpenter, 2012
Executive Summary
The focus of this project is the underwater cultural heritage (UCH) remains from the World War II
(WWII) Battle of Saipan that occurred in June and July of 1944 in the Commonwealth of the Northern
Mariana Islands (CNMI). This project builds upon a 2009 ABPP grant (WWII Invasion Beaches Underwater
Heritage Trail GA-2255-09-028) to survey and map the submerged archeological sites from the Battle
(McKinnon and Carrell 2011).
The foundations upon which the 2009 grant was built were a remote sensing survey of key areas by
SEARCH, Inc. (2008a, 2008b) and the Maritime History and Archaeology of the Commonwealth of the
Northern Mariana Islands (Carrell 2009). The implementation of the maritime heritage trail bought into
sharp focus the need to introduce preservation planning for the WWII UCH in Saipan’s waters. The
logical next step in the multi-year effort to preserve the UCH was this project and this report:
Management Plan and Public Outreach for WWII Sites in Saipan funded under a 2011 APBB grant (GA2255-11-018).
Throughout the 2009-2010 archeological survey, it was clear that certain submerged heritage sites were
being negatively impacted by both natural and cultural factors (McKinnon and Carrell 2011). These
impacts were identified as contributing to an overall loss of archeological and historical context and
affecting the structural integrity of the sites and their long-term survival. To address these concerns and
create a framework for long-term preservation, this project focused on two key areas: in situ
conservation survey with additional archaeological investigations to gather baseline data and a farreaching public outreach effort using a 17-minute interpretive film.
The WWII submerged sites face a variety of threats that include natural forces, cultural impacts,
development and visitation. The report includes recommendations for inter-agency and community
partnerships, improving effectiveness of current legislation and enforcement, long term monitoring
programs and strategic plans to mitigate cultural and natural impacts, visitor impacts and development.
It also addresses the limitations inherent in the funding, personnel, and management at the CNMI
Historic Preservation Office (HPO) and CNMI Coastal Resources Management office (CRM).
Additionally public outreach and education is a necessary component for the long-term protection of
cultural heritage. For fragile sites, and those that are underwater where monitoring is difficult,
stakeholders and users must take an active role. Thus a film was produced to raise awareness of both
divers and non-divers alike for WWII UCH in the CNMI.
iii
iv
Table of Contents
Executive Summary...................................................................................................................................... iii
Table of Contents .......................................................................................................................................... v
List of Figures ............................................................................................................................................... ix
List of Tables ................................................................................................................................................ xi
Acknowledgements.....................................................................................................................................xiii
Chapter 1: Introduction ................................................................................................................................ 1
Introduction .............................................................................................................................................. 1
In Situ Conservation Survey .................................................................................................................. 1
Public Outreach and Interpretive Film .................................................................................................. 2
Scope ......................................................................................................................................................... 2
The Sites ................................................................................................................................................ 3
Limitations............................................................................................................................................. 4
Legislation ................................................................................................................................................. 4
Jurisdictional Responsibilities ................................................................................................................... 4
Chapter 2: Historical Background 1911-1944 (after SEARCH, Inc. 2008) ..................................................... 7
Historical Context...................................................................................................................................... 7
The Mariana Islands in WWII, 1941-1945 ............................................................................................. 7
Japanese Defences on Saipan ............................................................................................................... 9
The Aftermath of the Battle of Saipan ................................................................................................ 12
Chapter 3: Cultural Resources .................................................................................................................... 15
Daihatsu Landing Craft ............................................................................................................................ 15
Daihatsu 1 ........................................................................................................................................... 15
Daihatsu 2 ........................................................................................................................................... 16
Daihatsu 3 ........................................................................................................................................... 17
Inter-site Preservation and Deterioration .......................................................................................... 19
Sherman Tanks ........................................................................................................................................ 19
Tank 1 .................................................................................................................................................. 19
Tank 2 .................................................................................................................................................. 20
Tank 3 .................................................................................................................................................. 22
v
Inter-site Preservation and Deterioration .......................................................................................... 23
Landing Vehicle Tracked LVT (A)-4.......................................................................................................... 24
Landing Vehicle 1 (LVT-1) .................................................................................................................... 24
Landing Vehicle 2 (LVT-2) .................................................................................................................... 25
Inter-site Preservation and Deterioration .......................................................................................... 26
Japanese Merchant Ship, Presumably Shoan Maru aka “Chinsen” ........................................................ 26
Preservation and Deterioration .......................................................................................................... 28
Possible Auxiliary Submarine Chaser ...................................................................................................... 28
Preservation and Deterioration .......................................................................................................... 31
Unidentified Steamship .......................................................................................................................... 31
Preservation and Deterioration .......................................................................................................... 32
Aichi E13A “JAKE” ................................................................................................................................... 32
Preservation and Deterioration .......................................................................................................... 34
Kawanishi H8K “EMILY” .......................................................................................................................... 34
Preservation and Deterioration .......................................................................................................... 35
Martin PBM Mariner ............................................................................................................................... 36
Preservation and Deterioration .......................................................................................................... 37
TBM Avenger........................................................................................................................................... 37
Preservation and Deterioration .......................................................................................................... 39
Consolidated PB2Y Coronado ................................................................................................................. 39
Preservation and Deterioration .......................................................................................................... 40
Chapter 4: Threats and Impacts .................................................................................................................. 41
Cultural Threats and Impacts .................................................................................................................. 41
Anchor and Mooring Damage ............................................................................................................. 41
Looting and Moving Artifacts .............................................................................................................. 43
Acts of Vandalism................................................................................................................................ 47
Tourism Services Impacts.................................................................................................................... 48
Memorialization .................................................................................................................................. 49
Environmental Threats and Impacts ....................................................................................................... 49
In Situ and Ex Situ Artifacts ..................................................................................................................... 52
Chapter 5: Public Outreach ......................................................................................................................... 55
Introduction ............................................................................................................................................ 55
vi
Consultation, Public Meetings, Presentations, Press and Digital Media ................................................ 55
Consultation ........................................................................................................................................ 55
Public Meetings and Presentations .................................................................................................... 56
Press and Digital Media....................................................................................................................... 58
Interpretive Materials and Film .............................................................................................................. 59
Poster and Dive Guide ........................................................................................................................ 59
The Film - WWII Maritime Heritage Trail: Battle of Saipan ................................................................ 63
Chapter 6: Recommendations and Relevant Issues.................................................................................... 67
Management of Underwater Cultural Heritage Sites ............................................................................. 68
Consultation and Strategic Planning ....................................................................................................... 68
Legislation and Effective Enforcement ................................................................................................... 69
Complete National Register Nominations .............................................................................................. 70
Site Database and At-Risk Artifacts......................................................................................................... 71
Programmatic Research and Inventory .................................................................................................. 72
Monitor Material Remains to Identify Natural Impacts ......................................................................... 73
Recognize UCH as Sites of Natural Significance ...................................................................................... 74
Protect Material Remains from Cultural Threats.................................................................................... 75
Monitor Impacts of Development .......................................................................................................... 80
Monitor Visitation to Sites ...................................................................................................................... 81
Strengthen Relationship with the Dive Community ............................................................................... 81
Increase Availability of Interpretive Materials........................................................................................ 82
Chapter 7: Conclusion ................................................................................................................................. 87
Further Contact ....................................................................................................................................... 89
References Cited ......................................................................................................................................... 91
Appendix A: Conservation Survey and Management Program - Saipan WWII Underwater Archaeological
Wreck Sites ................................................................................................................................................. 95
vii
viii
List of Figures
Figure 1. Map of sites investigated (McKinnon 2012). ................................................................................. 3
Figure 2. Invasion Beaches of Saipan. Map design James W. Hunter, Ships of Discovery. .......................... 9
Figure 3. Daihatsu Landing Craft 1 - stern to bow view (Carpenter 2012). ................................................ 16
Figure 4. Daihatsu Landing Craft 2 – port side view (Carpenter 2012)....................................................... 17
Figure 5. Daihatsu Landing Craft 3 – port side view (Carpenter 2012)....................................................... 18
Figure 6. M4 Sherman Tank 1 – port side view (Carpenter 2012). ............................................................. 20
Figure 7. The tank shows evidence of rust, surface spalling, and loss of structural integrity (Carpenter
2012). .......................................................................................................................................................... 21
Figure 8. The tank sits on a bed of seagrass that extends in all directions (Carpenter 2012). ................... 21
Figure 9. M4 Sherman Tank 3 – stern view (Carpenter 2012). ................................................................... 22
Figure 10. Landing Vehicle Tracked 1 – port side view (Carpenter 2012). ................................................. 24
Figure 11. Landing Vehicle Tracked 2 – front view (Carpenter 2012)....................................................... 26
Figure 12. Japanese merchant Ship (bow view) (Carpenter 2012). ............................................................ 27
Figure 13.Photomosaic of possible auxiliary submarine chaser site. ......................................................... 29
Figure 14. Unidentified steamship – boilers (Carpenter 2012). ................................................................. 32
Figure 15. Overview photograph of Aichi E13A JAKE (Carpenter 2012). ................................................... 33
Figure 16. Kawanishi H8K (EMILY) (Carpenter 2012). ................................................................................. 35
Figure 17. Overview shot of central portions of Martin PBM Mariner site; note dihedral wing (D.
McHenry, June 2010). ................................................................................................................................. 37
Figure 18. Grumman TBM Avenger (Carpenter 2012). ............................................................................... 38
Figure 19. Consolidated PB2Y Coronado (Carpenter 2012). ....................................................................... 40
Figure 20. Map of Mañagaha Marine Conservation Area (CRM). .............................................................. 42
Figure 21. Avenger landing gear, note shiny metal where concretion has been rubbed away due to
mooring (Carpenter 2012). ......................................................................................................................... 42
Figure 22. Korean monument with .50 caliber rounds (Gauvin 2010). ...................................................... 43
Figure 23. Cockpit configuration changes over time on Kawanishi H8K (Bell 2010). ................................. 44
Figure 24. Japanese monument surrounded by gas cylinders and other moveable objects (Seymour
2012). .......................................................................................................................................................... 44
Figure 25. Wireless radio Identification plate recovered from the Kawanishi H8K sites (masadivesaipan.com, accessed June 2010). .............................................................................................................. 45
Figure 26. A portion of a Kawanisihi H8K reported to the HPO (Rogers 2010). ......................................... 45
Figure 27. Artifact pile created by divers and altered over time (Bell 2010).............................................. 46
Figure 28. Faux airplane wreck created for a submarine tour with artifacts gathered from sites in the
vicinity. Note four-blade propeller (foreground) and gun in background (Carpenter 2012). .................... 47
Figure 29. Graffiti etched into the Kawanishi H8K (Bell 2010) ................................................................... 48
Figure 30. Front of Japanese shipwrecks poster......................................................................................... 61
Figure 31. Back of Japanese shipwrecks poster. ......................................................................................... 62
Figure 32. Aichi 313A dive guide. ................................................................................................................ 63
Figure 33. Diving Shipwreck brochure produced by South Australia Heritage........................................... 77
Figure 34. Anchoring brochure produced by South Australia Heritage Branch. ........................................ 79
ix
x
List of Tables
Table 1. Table of sites investigated. .............................................................................................................. 2
Table 2. Quick reference to recommendations, action items and time frame .......................................... 83
xi
xii
Acknowledgements
This project would not have occurred without the help of a number of individuals and groups. First and
foremost we would like to thank the CNMI Historic Preservation Office for their continuous support
from long before this project started to long after it will have finished. Special thanks are extended to
Ronnie Rogers (no longer at HPO), John Castro (no longer at HPO), Herman Tudela (no longer at HPO),
“JP” John Palacios and Mertie Kani.
The Coastal Resources Management Office provided support for this project. We would especially like to
thank John “John John” D. San Nicolas for your endless support and for being an integral member of the
team.
The agencies and organizations that have supported this project on island include: the National Park
Service American Memorial Park; the NMI Humanities Council; Power 99; Poseidon’s Maidens Dive
Charter and Aqua Connections.
Special thanks are extended to two generous individuals, Scott Eck for the use of his vessel and
Genevieve Cabrera who continues to educate us about the island.
Thank you to Peter Harvey, Jason Raupp, Cos Coroneos, Vicki Richards, Jon Carpenter, Toni Carrell and
Jennifer McKinnon for forming the “A-Team,” who collected an incredible amount of conservation
survey data in a very short period of time.
Without the hard work of Brett Seymour, National Park Service, Maryanne Morin and Louis Lamar,
Woods Hole Institution, Veronica Veerkamp and Richard Coberly, Windward Media, there would be no
interpretive film. The WWII Maritime Heritage Trail: Battle of Saipan is not only visually beautiful but, as
importantly, was scripted and narrated with sensitivity and respect for all who lost their lives in Saipan.
We owe the film team a debt of gratitude for their guidance and hard work.
Special thanks go to Flinders University for providing staff time and equipment and East Carolina
University for providing staff time to this project.
A big thank you should be extended to the American Battlefield Protection Program for providing the
funding for this project. We would particularly like to thank Kristen McMasters for her availability and
willingness to help us through this process.
Jennifer McKinnon, Greenville
Toni Carrell, Santa Fe
xiii
xiv
Chapter 1: Introduction
Introduction
The implementation of the WWII Invasion Beaches Underwater Heritage Trail under a 2009 ABPP grant
(GA-2255-09-028) brought into sharp focus the need to develop a management and preservation plan
for the submerged WWII heritage in Saipan’s waters. The foundations upon which the 2009 grant was
built were a remote sensing survey of key areas by SEARCH, Inc. (2008a, 2008b) and the Maritime
History and Archaeology of the Commonwealth of the Northern Mariana Islands (Carrell 2009). The
logical next step in the multi-year effort to preserve WWII UCH was this project funded under a 2011
APBB grant (GA-2255-11-018) and this report: Management Plan and Public Outreach for WWII Sites in
Saipan.
Throughout the 2009-2010 archeological survey, it was clear that certain UCH sites were being
negatively impacted by both natural and cultural factors (McKinnon and Carrell 2011). These impacts
were identified as contributing to an overall loss of archeological and historical context and affecting the
structural integrity of the sites and their long-term survival. To address these concerns and create a
framework for long-term management and preservation, this project focused on two key areas: in situ
conservation survey with additional archaeological investigations and a far-reaching public outreach
effort. This report details the results of those efforts and makes recommendations for the preservation
of WWII submerged heritage. This is the technical report for the current grant (GA-2255-11-018).
In Situ Conservation Survey
In situ conservation surveys are different from standard archeological surveys because they include the
collection of data related to the natural environment (chemical and physical) that can allow for a better
understanding of the destructive forces affecting sites and artifacts (MacLeod and Richards 2011). They
also record modern cultural impacts that can be used in the management of sites, including allowing or
restricting access to sites and controls for altering behaviors of visitors. Finally, they often collect specific
cultural information including data on material composition of objects. This data is vital to
understanding the construction of these sites from an archeological or cultural perspective, but also
contributes to understanding the longevity of sites with relation to the material composition of metals
and organics, how they react to and survive within the environment, and their overall structural
integrity.
In situ surveys and studies are critical to regions such as the Pacific because there are limited resources
(i.e. funding, staff, and facilities) to conduct recovery and conservation of submerged objects and sites.
While the CNMI does benefit from grant funds distributed by the National Park Service (NPS), this
funding is limited and often only covers a small portion of the compliance needs of the HPO (Ronnie
Rodgers personal communication, 2010). This means that the conservation and management of the
resources must be done in situ. Further, understanding the condition of the resources through in situ
and archeological surveys is an important step in the management process. An agency cannot manage a
site if they have no knowledge of its condition.
1
In addition to the conservation study, additional archeological survey was needed on sites included on
the heritage trail and also on sites that had yet to be identified and recorded. The purpose of this
archeological survey was two-fold: 1) to update information on natural impacts to known sites that
affect site formation processes (i.e. scouring, sediment build-up) and cultural impacts (i.e. looting,
vandalism, etc.); and 2) to locate, record and identify new “control” sites not on the trail that could be
monitored long-term for comparison purposes. Baseline data collected on new sites is useful because
they can help managers understand the differential impacts of site visitation on those included in the
trail.
Public Outreach and Interpretive Film
The interpretive film was conceived during many discussions with the staff of the NPS American
Memorial park, NPS War in the Pacific National Park, the CNMI HPO, local divers, dive tour operators
and visitors. In those discussions it was clear that although the posters and dive guides produced under
the 2009 heritage trail grant were well received, their impact was limited to those individuals who
obtained copies. Once those printed copies were used up the ability of any group or individual to raise
funds to reprint them was going to be difficult. To create a broad base of support, the preservation
message had to reach a much larger audience and be unlimited in its application and use. An
interpretive film about the WWII submerged sites targeting divers, non-divers, visitors and residents
would not only have an inherent appeal but could use the power of film to reach all audiences.
Scope
The scope of this report includes the sites on the WWII Maritime Heritage Trail, an additional four sites
not included on the trail, and more generally all submerged WWII heritage sites in the waters of Saipan,
CNMI (Table 1).
Table 1. Table of sites investigated.
Name of Site
Japanese merchant ship, Presumably Shoan
Maru/Chinsen (Japanese)
Steamship (Unidentified)
Possible Auxiliary Submarine Chaser (Japanese)
Sherman Tank 1 (U.S.)
Sherman Tank 3 (U.S.)
Landing Vehicle Tracked-(A)-4 (U.S.)
LVT 2 (U.S.)
Daihatsu Landing Craft 1 (Japanese)
Daihatsu Landing Craft 2 (Japanese)
Daihatsu Landing Craft 3 (Japanese)
Kawanishi H8K “EMILY” (Japanese)
Aichi E13A “JAKE” (Japanese)
Martin PBM Mariner (U.S.)
TBM Avenger (U.S.)
PB2Y Coronado (U.S.)
Trail (Y/N)
Y
N
Y
Y
Y
Y
N
Y
Y
N
Y
Y
Y
Y
N
2
The Sites
All sites are within Garapan, Tanapag and Chalan Kanoa Lagoons on the western side of the island of
Saipan. The lagoons were created by a shallow barrier reef and range in width from 375 m to 3.5 km and
in depth up to 14 m deep in Tanapag Lagoon with an average of less than 3 m deep in Garapan Lagoon
(Amesbury et al. 1996). Much of Chalan Kanoa lagoon is difficult to navigate because it is shallow and
interspersed with coral heads and reef flats. A small islet called Mañagaha is located within the lagoon
just west of the modern harbor facilities.
The sites are scattered throughout the area with the main concentration in Garapan Lagoon (Figure 1 ).
They include aircraft, shipwrecks and assault vehicles of varying states of articulation. Each site will be
reviewed and described in detail in below chapters.
Figure 1. Map of sites investigated (McKinnon 2012).
3
Limitations
An inherent limitation in this project is that the remains of many sites have not yet been identified or
are buried beneath the sediments. Nonetheless, the in situ conservation data and the recommendations
generated for preservation are broadly applicable to all of the WWII wrecks in Saipan.
Legislation
Historic shipwrecks in the CNMI are protected through laws passed by the Commonwealth and the US
Federal government. The following is a list of legislation relevant to UCH and military heritage:
The Abandoned Shipwreck Act 1987 protects historic shipwrecks “embedded in State’s submerged
lands.”
The Sunken Military Craft Act 2005 confirms right, title and interest of the US to any sunken military
craft anywhere in the world as well as the same rights and protection to non-US military craft sunk in US
controlled bottomland.
The National Historic Preservation Act 1966 under Section 106 and 110 provides protection for
shipwrecks and other submerged sites with regards to the permit and mitigation processes and requires
inventory and assessment of such sites as standard procedure.
The Archaeological Resources Protection Act 1979 prohibits damage to archeological sites that are 100
years or older and provides archeological and permit guidelines.
CNMI Historic Preservation Act of 1982 (Public Law 3-39) protects submerged sites stating, “It shall be
unlawful for any person, partnership, business, corporation or other entity who willfully remove or take
any artifact that is of historic or cultural significance to the people of the Northern Mariana Island, or
knowingly destroy, remove, disturb, displace, or disfigure any cultural or historic property on public or
private land or in the water surrounding the Northern Mariana Islands as designated by or eligible for
designation by the HPO as a cultural or historic property, unless such activity is pursuant to a permit
issued under Section 5 of this Act.”
Jurisdictional Responsibilities
The Historic Preservation Office (HPO) has the overall administrative responsibility for cultural heritage
in the CNMI. This obligation extends off shore to all UCH, whether they have been identified or are as
yet to be investigated.
To accomplish the legislative purpose of the Mañagaha Marine Conservation Act, the CNMI Department
of Lands and Natural Resources (DLNR) was delegated the exclusive authority to manage the Mañagaha
Marine Conservation Area (MMCA), as well as other marine conservation areas in the CNMI (Section 6 of
PL 12-12). The role of the DLNR in regard to the management of UCH is to promote public access to the
sites while protecting the physical remains. They are also responsible for protecting sites.
4
Coastal Resources Management (CRM) was established on 11 February 1983, with the implementation
of Public Law 3-47 within the Office of the Governor. The CRM program was established in order to
promote the conservation and wise development of coastal resources. CRM is responsible for general
permitting activities that impact coastal resources in Saipan and in particular permits for dive boats and
dive tour operations in all Saipan waters. Similarly, the Department of Environmental Quality (DEQ)
ismainly concerned with water quality and pollution and the US Fish and Wildlife Service (FWS) with fish
and marine life. Their responsibilities extend to all Saipan waters.
5
6
Chapter 2: Historical Background 1911-1944 (after SEARCH, Inc. 2008)
Historical Context
As destructive as World War I was in other parts of the world, the Mariana Islands were spared. Still, the
end of the war had important ramifications for the islands. The League of Nations approved Japan’s
occupation of the Mariana (sans Guam), Caroline, Marshall, and Palau Islands in 1921 with the
stipulation that Japan not develop fortifications in the region. For over a decade Japan complied with
the agreement and focused on infrastructural development that strengthened the economy of their
possessions.
Increasingly militaristic and expansionist, Japan sought to strengthen its presence in the Pacific in the
late 1930s. For example on Saipan, Aslito Airfield on the southern end of the island and a seaplane base
at Flores Point (northeast of Garapan) were constructed. Barracks, ammunition storage, air raid shelters,
and other facilities preparatory for an offensive war were installed elsewhere on Saipan in 1941 (Russell
1984). Japan’s moves to fortify the islands abrogated their agreement with the League of Nations, thus
losing their legal jurisdiction.
The Mariana Islands in WWII, 1941-1945
Across the Pacific, word spread that war was coming, however, only Japan knew when and where it
would start. Shortly after the December 7, 1941 attack on Pearl Harbor in Hawaii, Japanese forces
initiated air attacks over Guam, which they had been openly monitoring since November. Commercial
airline buildings, fuel supplies, the US Navy yard, vessels in Apra Harbor, and the capital at Agana were
bombed. Poorly defended by the Americans, Guam fell on December 10th within six hours of the
subsequent Japanese invasion. Guam became the only part of the US to fall under enemy occupation
during WWII. With this prize under its belt, the Japanese Imperial Army and Navy mounted very
successful aggressive operations across the Pacific in the early years of the war (Rogers 1995; Rottman
2004a).
As these events were occurring in the Marianas, Allied powers meeting in Egypt agreed upon a Pacific
strategy consisting of two offensive drives. The first, led by General Douglas MacArthur, was an advance
from New Guinea to the Philippines. The second, led by Admiral Chester Nimitz, was a push through the
Gilbert and Marshall Islands across the Central Pacific to take the Marianas. Once accomplished, a
strategic bombing campaign could be mounted against the Japanese mainland. By the start of 1944, this
broad strategy proved successful for the US, and the Japanese government realized an invasion of the
Marianas was imminent. As Marine historian O.R. Lodge wrote, “A strangling noose was tightening
around the inner perimeter guarding the path to their homeland” (Lodge 1954).
Operation Forager
The US’s plan to take control of the Marianas from Japanese forces was code-named Operation Forager
and included the island of Guam. The plan to invade and take control of Saipan was called Operation
Teattersalls. This operation involved thousands of troops from all branches of the military and a
bewildering number of vessels, vehicles, and weapons. Japan’s defenses focused on five islands in the
7
Marianas: Guam, Rota, Tinian, Saipan, and Pagan. Saipan, home of the administrative center of the
Japanese Marianas, was chosen as the first target with Tinian and Guam as secondary. US war planners
opted to bypass a full assault on Rota and Pagan (Rottman 2004a).
Air attacks were the first phase of Operation Forager. In February 1944, US bombers destroyed the
Orote Peninsula airstrip on Guam. General air raiding also began across the Marianas, resulting in US
dominance of the skies. In response, Japanese forces prepared Operation A-Go, which relied upon the
support of the Imperial Japanese Navy and air forces to support troops on the ground in the Marianas.
Operation Forager was in full swing by June 1944 when the invasion of Saipan was planned. The invasion
was set for June 15, D-Day. Responsible for this task was the Marine V Amphibious Corps (VAC) that
consisted of the 2nd and 4th Marine Divisions, the 27th Infantry Division (Army), and the XXIV Corps
Artillery (Army). US military planners considered landing points on all sides of Saipan. On the eastern
side of the island an area designated Brown Beach on the Kagman Peninsula was considered but
rejected because it was well-defended and would offer a poor exit. Three other beaches, Purple Beach
on Magicienne Bay and White Beach 1 and 2 near Cape Obian on the southern reach of the island were
kept as alternates. At Tanapag Harbor were the well-defended Scarlet Beaches 1 and 2, and to the north
of this area were Black Beaches 1 and 2, which offered insufficient space for the large unit landing that
was planned (Rottman 2004a).
The lower western side of Saipan was chosen as the primary invasion area. Stretching approximately
four miles and divided into several zones, this area was most favorable because the size would allow
two Marine divisions to land simultaneously. A landing here also allowed the immediate capture of the
airstrip at Chalan Kanoa placing direct pressure on nearby Aslito Airfield. Once secure, these airstrips
were used to support the penetration northward across Saipan. Afetna Point divided the landing
beaches. To the north were Red Beaches 1, 2, and 3 and Green Beaches 1 and 2 while to the south were
Green Beach 3, Blue Beaches 1 and 2, and Yellow Beaches 1, 2, and 3 (Rottman 2004a) (Figure 2).
As favorable as the lower western side of Saipan was for the US invasion, certain limitations remained.
Because of its distance from the northern sector of the island, beaches in this area (Scarlet 1 and 2 and
Black 1 and 2) had to be captured in order to facilitate the off-loading of supplies from landing craft.
Another problem at the beaches off the lower western side of Saipan was the extensive coral reef that
abutted the shore. These had to be negotiated with amphibious vehicles including AMTRACS or Landing
Vehicle Tracked (LVTs) and amphibious tanks, also known as Landing Vehicle Tanks (LVT(A)s). Both the
2nd Marine Division and the 4th Marine Division had three AMTRAC battalions each, plus one amphibian
tank battalion, for the initial shore assault. This amounted to approximately 1,400 AMTRACS, the largest
use to date of amphibious vehicles. Once ashore, the AMTRACS were required to push inland from the
beach, a tactic that was equally unprecedented and also potentially dangerous. AMTRACS, with their
thin armor and low ground clearance, were not designed for cross-country movement.
8
Japanese Defences on Saipan
The Japanese defensive strategy used for Saipan and the other Mariana Islands was identical to that of
earlier battles at the Gilbert and Marshall Islands: the enemy was to be met and destroyed at the
beaches and, if allowed inland, they were to be pushed into the sea by way of counterattacks (Denfield
1992). Japanese defenses on Saipan, while substantial, were incomplete by D-Day. Prior to the outbreak
of the war, Japan’s need for troops and material elsewhere in the Pacific hampered the pace of
developments on Saipan and the other islands of the Marianas. The Japanese forces established two
bases on Saipan by early 1944. The airfield at Aslito (built in the 1930s) served as a repair and
maintenance facility for aircraft involved in battles to the east and south. At Tanapag Harbor, a naval
base served as a staging area for troops and ships bound elsewhere (Denfield 1992). Due to needs in
other parts of the Empire, Japanese troop numbers were fairly low; in February 1944, there were only
1,500 troops on Saipan.
Figure 2. Invasion Beaches of Saipan. Map design James W. Hunter, Ships of Discovery.
9
The impending US campaign against the Marianas hurried Japanese defensive measures and by the end
of February thousands of troops were sent to the Marianas to prepare for battle. US submarine attacks
on Japanese transports prevented an estimated 2,000 troops from reaching the islands (Denfield 1992),
but by early June some 30,000 troops were prepared for battle on Saipan (Rottman 2004b).
Once they had sufficient troops and materials available, the Japanese forces commenced defensive
improvements on Guam, Tinian, Rota, Pagan, and Saipan. On Saipan, another airfield was built by Lake
Susupe near the town of Chalan Kanoa, but this was little more than an emergency landing strip. More
airstrips were planned on Saipan and throughout the islands, but most were never completed (Denfield
1992). Forty-six gun installations were established on the beaches and ridges of Saipan, but twelve were
not operational on D-Day. An additional three units lay on railcars awaiting installation at the start of the
battle and another 42 were in storage at the navy base at Tanapag. Construction of three blockhouses
on the beaches of Saipan, each a concrete structure with four ports housing heavy guns, began in early
1944. Between Camp Obiam and Agingan Point (White Beach 1), one of these blockhouses stood
incomplete and unarmed on D-Day. Another was located to the east of Cape Obiam (White Beach 2) and
the third was on the eastern side of the island at Magicienne Bay (Purple Beach) (Denfield 1992).
The Battle of Saipan – Operation Tattersalls 1944
An intensive barrage presaged the beach invasion foreshadowing the massive confrontation that was set
for June 15. On June 12, 200 carrier aircraft bombed Japanese airfields in the southern Marianas,
decimating the enemy’s air force. That same day, American B-24 bombers began around-the-clock raids
on Saipan and Tinian. The following day (June 13), seven fast battleships leveled the towns of Chalan
Kanoa and Garapan and on June 14, more battleships as well as 11 cruisers and 26 destroyers attacked
Japanese coastal defenses.
On June 14, other crucial preliminary actions took place along the coast of Saipan. Early that morning,
Underwater Demolition Teams began their mission to demolish reefs, enemy mines, and mark lanes
along the Red, Green, Blue, and Yellow beaches. This effort was largely successful. Simultaneously, a
mock attack was held by a “Demonstration Group” off the northwestern coast at Scarlet and Black
beaches (refer to Figure 2) in an attempt to trick the Japanese into thinking that the invasion was to
occur there. Another was held in the early morning hours of June 15, D-Day. The Demonstration Group
consisted of the 2nd Marines, the 1st Battalion, 29th Marines, and the 24th Marines. The feint was
supported by naval gunfire as landing craft approached the beach to within 5,000 yards, circled for a few
minutes, wheeled about, and returned to their ships. Troops were not embarked, the landing craft drew
no fire, and no activity was observed on the shore. Although clearly not convinced by the ploy, the
Japanese stopped short of removing troops from here and redeploying them to other beaches.
Interestingly, Japanese commanders did wire Tokyo to report that they had repelled an invasion in this
area (Rottman 2004a).
As the sun rose on June 15, the enormous US amphibian force was assembled for the invasion.
AMTRACS and DUKWs were unloaded and LSDs (Landing Ship, Dock) launched LCMs (Landing Craft,
Mechanized) that held tanks. Battleships, cruisers, and destroyers closed in on the invasion beach with
10
aircraft screaming overhead toward the shore. With nearly 1,500 vessels in all, this great unpacking of
men, weaponry and ammunition was surely an intimidating sight for the Japanese forces and the
islanders. At exactly 0830 hours, the AMTRACS carrying hundreds of marines stormed for the beach as
supporting ships and aircraft opened fire. The Japanese troops generally held their fire until the
AMTRACS reached the lip of the coral reef when they poured artillery, mortar, and machine gun fire on
the Marines. Due to the condition of the seas, the Marines had a great deal of difficulty reaching the
beach. Numerous AMTRACS were rendered inoperable or destroyed as a result. Dozens of Marines lost
their lives when AMTRACS were overturned in the rough surf, but by 0900 hours nearly 8,000 Marines
were ashore (Crowl 1960; Rottman 2004a).
The beaches, especially the northern Red and Green Beaches, were chaotic and crowded. Because of the
heavy swell and long shore current, the US landing in this area was several hundred yards north of its
intended location. In all areas Japanese fire was very heavy and came from the high ground beyond, as
well as trenches and spiderholes along the immediate shoreline. The overland AMTRAC charge, a
calculated risk, proved disastrous as the flimsy AMTRACS became stranded in marshy areas, craters, and
other obstructions on the ground. As casualties for the Marines mounted, many chose to abandon their
machines in favor of walking or crawling. Problems increased as the morning wore on. At the northern
beaches, two 2nd Marine Division command posts were destroyed, killing battalion commanders and key
staff and on the south, the 4th Division indecisively struggled with Japanese tanks for hours. The arrival
of US tank battalions, howitzer batteries, and other support forces improved the overall situation by the
end of the day, although the beachhead remained unconsolidated. On this day, the first of the battle for
Saipan, some 2,000 Marines were killed. Mortar and artillery fire were the principal causes of death.
Offshore hospital vessels were completely overwhelmed and wounded men were transferred to other,
non-hospital ships for treatment (Crowl 1960; Rottman 2004a).
Over the next two days, the Marines secured and expanded the beachhead. The morning of June 16 was
spent closing the gap between the two divisions, which were separated by a strong Japanese force on
Afetna Point. By noon, this goal was accomplished. A huge Japanese counterattack was successfully
repelled that afternoon and the Marines engaged in the largest tank battle of the Pacific War. Aslito
Airfield was under pressure at the close of the day. In contrast, the 27th Infantry Division (Army) came
ashore to support the Marines on the morning of June 17 and bore the brunt of a door-to-door fight
through Garapan village, another innovation in the Pacific War.
After the beaches were secured, support groups landed on the beachheads to facilitate the massive
unloading of supplies. The post-invasion landscape they witnessed was grim. Disabled AMTRACS and
boxes of c-rations were scattered across the beach. Bodies of dead Marines that were not yet recovered
bobbed in the surf. “The leaves on battered trees and underbrush were covered with a fine, gray dust,”
wrote former Naval Construction Battalion (Seabee) commander David Moore. “This whole scene gave
an eerie feeling of war” (Moore 2002).
On the evening of D-Day, Seabees were ordered to launch the floating causeways held in the LSTs
(Landing Ship, Tanks). Throughout the night, the Seabees maneuvered the various pieces through the
11
channel and assembled them at the Chalan Kanoa beachhead. By daybreak supplies were being
unloaded from landing craft. Other craft towed pontoon sections that were made into floating piers to
unload supplies. Pontoon barges were used to haul ammunition. While vital to the continued success of
the American operation on Saipan, unloading supplies and ammunition from barges on the beachhead
became monotonous as combat moved farther inland. This “grueling mission of moving ammunition and
other supplies to the beach…became a routine of eating bland c-rations and sleeping on the barge for a
boring 54 days of blazing sun or miserable rain,” remembered one Seabee who served at Saipan. “The
Coxswain often grumbled: it is noble to suffer. We [the Seabees] were granted the undistinguished title,
‘Bastards of the Beaches’” (Moore 2002).
There were nevertheless many soldiers on the mainland that would have traded places with the
Seabees. Japanese attempts to call in reinforcements from neighboring islands were fruitless, but they
continued to fight with resolve. On June 18, they attempted a daring counter-landing from their tattered
navy base at Tanapag Harbor. Japanese infantrymen were hastily loaded onto 35 barges and sent
toward the US landing beaches. In route, US infantry gunboats and Marine artillery intercepted them,
destroying many and deterring all. A larger-scale encounter played out in the Philippine Sea between
June 19 and 20 when the US Navy defeated the Japanese Imperial Navy in what is known at the
“Marianas Turkey Shoot.” In the meantime, the US secured Aslito Airfield and the southern reaches of
the island and prepared the northward fight (Crowl 1960; Rottman 2004a).
More gruesome scenes followed in this monstrous battle as June became July and US troops pushed
deeper into the island of Saipan. Garapan was secured on July 3, the Tanapag seaplane base on July 4,
and on July 6 the Japanese forces staged their last massive banzai charge. After Saipan, Japanese
commanders deemed such suicidal charges as wasteful. Rarely effective, they were not used in later
battles. Through mountainous terrain, tropical conditions, and intense resistance, US forces reached the
northern end of Saipan at Marpi Point on July 9, 1944 (Crowl 1960; Rottman 2004a).
The Aftermath of the Battle of Saipan
The loss of life during the Battle of Saipan was tremendous. Approximately 3,426 of the 67,451 US
troops who participated in the battle were killed or reported missing in action. Four times this number
were confirmed wounded. Japanese losses were far greater. Of the approximately 31,629 Japanese
troops who participated in the battle, 29,500 were killed or missing (Rottman 2004b). Japanese sources,
however, estimated the total number killed on Saipan to be well over 40,000 (Bulgrin 2005).
The long-fought and costly victory at Saipan, the fiercest of the three major battles in the Marianas, was
politically and militarily decisive. US successes in the opening two weeks of the battle induced Japanese
Emperor Hirohito to attempt a diplomatic end to the war. When news of Saipan’s fall reached Japan,
political pressure forced Prime Minister, War Minister, and Chief of Army General Staff Tojo Hideki and
his cabinet, as well as navy officials, to resign (Rottman 2004a).
Equally as important to the decisiveness of the victory at Saipan was the island’s proximity to Japan. This
was also true of Guam and Tinian. The Mariana Islands were ideal for the development of bases for long
12
range, B-29 bombers that were capable of reaching the Japanese mainland. On Saipan, the US wasted
no time in developing these bases. Aslito Airfield was renamed Conroy Field and then Isley Field. By
December 1944, it was used as the main operating field on the island. Two new airfields, Kobler Field
and another at Kagman Peninsula, were also built at this time. The seaplane base at Flores Point was
rebuilt and improvements were made to the Marpi Point airfield.
Saipan and the Mariana Islands paved the road for more decisive battles at Okinawa and Iwo Jima.
Together, the great sacrifices of these and other battles of the Pacific War gave US military planners an
impression of what could be expected from an invasion of the Japanese mainland. Plans for such an
invasion were in the making when the decision was made to drop the atomic bombs on Hiroshima and
Nagasaki, resulting in the surrender of the Japanese Empire and the conclusion of WWII.
13
14
Chapter 3: Cultural Resources
This chapter provides basic identification, condition, and environmental dynamics on the archeological
sites that were the subject of research on differential preservation and deterioration for this project. A
full report of findings resulting from the in situ conservation survey is included as Appendix A (hereafter
referred to as Richards and Carpenter 2012). These data provide the basis for selected
recommendations in Chapter 6.
With the exception of the Daihatsu Landing Craft 3, Landing Vehicle (LVT) 2, an Unidentified Steamship
in Tanapag Harbor and the Consolidated PB2Y Coronado aircraft, all sites are included on the WWII
Maritime Heritage Trail – Battle of Saipan. Divers and snorkelers are actively encouraged to visit the
sites provided they follow the visitation guidelines and do not interfere with the site (i.e. disturb or
attempt to remove any components).
Daihatsu Landing Craft
The remains of two Japanese Daihatsu landing crafts lying in the vicinity of each other were originally
located by Pacific Basin Environmental Consultants (1985) and NPS (Miculka et al. 1984), reported by
Carrell (1991:500, 502), relocated by SEARCH, Inc. in 2008 (2008b:54, 59) and archeologically mapped in
2010 (McKinnon and Carrell 2011). A third site was located during the 2012 survey and was not
previously recorded. Because Daihatsu 3 is not included on the trail, it can serve as a control to monitor
visitor impacts.
Daihatsu 1
The Daihatsu Landing Craft 1 (Figure 3) is located on the southwest side of Saipan, inside Tanapag
Harbor at a depth of about 11 m (35 ft) (
). The wreck, positively identified as
a Daihatsu or 14m Japanese Landing Craft (McKinnon and Carrell 2011:93), is constructed primarily of
welded steel and the dimensions are 14.58 m in length, 3.35 m in width with a 0.76 m draught
(http://pwencycl.kgbudge.com/D/a/Daihatsu class.htm).
The vessel is upright and is reasonably intact except for the port side amidships region that has
collapsed. A deck winch with wire remains in position on the upper stern deck level. The steering wheel
has been displaced and is also lying on this upper deck level. Hull plates are holed in many places and
the armor shield lies on the seabed on the port side of the wheelhouse. The vessel is partially buried in
the stern and the upper half of the rudder is visible. Limited burial has occurred with sand accumulation
in the cargo bay. The maximum height of the main structure rises approximately 2-3m above the
seabed. There are some areas of active corrosion evident on the site, indicated by the presence of the
typical red/brown “rust” spots on the surface of the vehicle.
15
The wreck lies on a flat, slightly undulating sandy seabed that is comparatively devoid of marine biota. It
is covered in brown algal forms and some sporadic secondary colonization was evident (e.g. tunicates,
soft and hard corals, seaweed, etc.). Isolated hard coralline growths have formed in places with one
large coral formation on the upper stern deck. It appears that the wreck is not subjected to regular
burial/exposure cycles. Sediment has partly covered the lower profile areas but the establishment of
some hard corals indicates that significant accumulation of sediment does not readily occur. The site lies
in the comparatively protected lagoon area but the site has the potential to be affected by storm
conditions.
Figure 3. Daihatsu Landing Craft 1 - stern to bow view (Carpenter 2012).
Daihatsu 2
The Daihatsu Landing Craft 2 (Figure 4) is located about 45 m (150 ft) southwest of Daihatsu 1, on the
southwest side of Saipan, inside Tanapag Harbor at a depth of about 11 m (UTM
).
The wreck, positively identified as a Daihatsu or 14 m Japanese Landing Craft (McKinnon and Carrell
2011:98), is constructed primarily of welded steel and measures 14.58 m in length, 3.35 m in width with
a 0.76 m draught (http://pwencycl.kgbudge.com/D/a/Daihatsu class.htm).
The vessel is upright on the seabed and is in poor structural condition. Daihatsu 2 is considerably more
disarticulated than Daihatsu 1. Large sections lie separate and astern of the wreck and substantial
remains have collapsed on the port side of the vessel. The engine is missing and presumed salvaged but
the rudder and propeller shaft remain in situ. The maximum height of the main structure rises
approximately 2-3m above the seabed. Limited burial has occurred with sand accumulation in the
16
loading zone and around the lower profile areas. There are some areas of active corrosion evident on
the site, indicated by the presence of the typical red/brown “rust” spots on the surface of the vehicle.
The wreck is in a similar environment as Daihatsu 1, a flat, slightly undulating sandy seabed relatively
devoid of marine biota. The wreck is covered in brown algal forms and some sporadic secondary
colonization was evident (e.g. tunicates, soft and hard corals, seaweed, etc.). Isolated hard coralline
growths have formed in some places. It appears that the wreck is not subjected to regular
burial/exposure cycles. Sediment has partly covered the lower profile areas but the establishment of
some hard corals indicates that significant accumulation of sediment does not readily occur. The site lies
in the comparatively protected lagoon area but the site has the potential to be affected by storm
conditions.
Figure 4. Daihatsu Landing Craft 2 – port side view (Carpenter 2012).
Daihatsu 3
The Daihatsu Landing Craft 3 (Figure 5) is located on the southwest side of Saipan, inside Tanapag
Harbor at a depth of about 7 m (UTM
). This site was found during the 2012
survey and has not been previously archeologically recorded. The wreck appears to be very similar to
Daihatsu 1 and Daihatsu 2, which are constructed primarily of welded steel; the dimensions are 14.58 m
in length, 3.35 m in width with a 0.76 m draught
(http://pwencycl.kgbudge.com/D/a/Daihatsu class.htm).
17
The vessel is upright on the seabed and is reasonably intact. The maximum height of the main structure
rises approximately 2 m above the seabed. Some scouring has occurred around the stern and the rudder
and propeller are exposed. Many hull plates are either missing or considerably corroded with significant
areas of loss. Most of the port and starboard side hull structure in the amidships area has collapsed. The
engine room of this vehicle is visible through perforated hull plates. The engine remains in situ and is
complete with the exhaust system and other associated features and ancillary equipment. For example,
a funnel is located in the starboard forward corner of the engine room.
Figure 5. Daihatsu Landing Craft 3 – port side view (Carpenter 2012).
There are some areas of active corrosion evident, especially in the shallower, more exposed areas (i.e.
upper surfaces of the stern section) indicated by the presence of the typical red/brown “rust” spots on
the surface of the vehicle. The fact that the engine, other associated machinery and some artifacts (e.g.
funnel) are still present in situ supports the fact that this site is not visited frequently and human
interference has been minimal to date.
Similar to the other landing craft, the wreck lies on a flat, slightly undulating sandy seabed relatively
devoid of marine biota with some isolated coral outcrops in close proximity. The wreck was covered in
concretion, some brown algal forms and some secondary colonization was evident (e.g. tunicates, soft
and hard corals, etc.) especially in the more protected areas (i.e. in the engine room, under the stern).
Isolated hard coralline growths have formed in places with one large coral formation on the upper stern
deck adjacent to the windlass. It appears that the wreck is not subjected to regular burial/exposure
cycles. Sediment has partly covered the lower profile areas and the loading area but the establishment
of hard and soft corals indicates that significant accumulation of sediment does not readily occur.
18
Inter-site Preservation and Deterioration
Based upon the 2012 conservation study by Richards and Carpenter, it appears that Daihatsu Landing
Craft 2 is corroding at a slightly faster rate than both Daihatsu Landing Craft 1 and Daihatsu 3. In
addition, it appears that Daihatsu 1 and Daihatsu 3 are corroding at fairly similar rates, despite the fact
that Daihatsu 3 is a much shallower site where it is expected that the corrosion rate would be slightly
higher. This seems to suggest that human interference (i.e. recreational diving activities) is having some
impact on the deterioration rate of the deeper Daihatsu 1 site. Because of its good condition and
photogenic qualities Daihatsu 1 is frequently visited so regular monitoring is recommended.
Sherman Tanks
Two Sherman tanks were documented by NPS in 1984 (Miculka et al. 1984:2) and were identified again
during a remote sensing survey by SEARCH, Inc. in 2008 (2008a). In 2010, those two and a third tank
nearby were surveyed and reported on (McKinnon and Carrell 2011). All three are located near Chalan
Kanoa and Susupe beaches on the southwest side of the island on shallow, flat sandy and seagrass
seabed inside the barrier reef (refer to Figure 1). They are semi-submerged with their turrets and decks
awash during low tides and are the subject of many tourist photographs.
Tank 1
Tank 1 is the northernmost, located approximately 120 meters offshore and sitting in approximately 1.52 m of water (UTM
). The tank is semi-submerged and at low tide all components
above the upper hull including the turret and gun are exposed to the atmosphere (Figure 6). Shiny nickel
welds are evident on the upper hull edges.
The tank was identified as a M4A2 Dry model, constructed principally of rolled and cast homogenous
steel, is 5.84m in length, 2.62m wide and 2.74m in height (Grove 1976:130-131). The hull of Tank 1 is
oriented with its bow toward the shore on a bearing of 133˚. The main 75mm gun is fixed on a bearing
of 197˚. The main body of the vehicle is mostly intact but other smaller components are missing. This
loss may be due to corrosion and/or cultural impacts, such as salvage. There are many areas of active
corrosion evident on the site, indicated by the presence of the typical red/brown “rust” spots on the
surfaces of the tank. There are also signs of accelerated corrosion on the upper sections of the tank
(flaking, surface spalling and cracking of the metal surfaces and gun barrel) that are cyclically exposed to
the atmosphere. Children were observed playing on the tank, walking along the main gun barrel and
jumping or diving into the water. In time, this practice may become a health and safety issue due to the
extensive corrosion exhibited by these exposed sections and the ever increasing probability that these
areas may collapse or fragment.
When combined, Saipan’s environment and the tanks’ bulk react, taking a toll on these tanks in the form
of holes, cracks, and corrosion. Certain components of the tanks have disappeared altogether, while
others are in danger of being lost. Natural processes such as corrosion and cultural impacts like salvage
are the primary explanations for many missing components. This is particularly true of long, thin parts of
the vehicles like the main gun barrel as well as movable parts such as brackets and hatches.
19
Figure 6. M4 Sherman Tank 1 – port side view (Carpenter 2012).
The surrounding seabed is relatively flat interspersed with large patches of seagrass. The tank lies in a
shallow depression with gently sloping edges and the lower section of the track and roller assembly is
mostly buried. A circular area about 12m2 surrounding the tank is free of seagrass but algal forms are
present on the seabed and on the submerged parts of the hull. High nutrient levels in the lagoon may be
contributing to this extensive algal growth (Denton et al. 2001).
Tank 2
Tank 2 is located nearly 300 m south of Tank 1 and about 450 m offshore. This tank is also a M4A2 Dry
tank with a 75mm cannon. Therefore, it was also equipped with twin General Motors 6-71 diesel
engines. Similar to Tank 1, the fixtures are still present, but none of the auxiliary weapons remain. The
welded hull of Tank 2 is oriented with its bow shoreward on a bearing of 145˚. The main gun is fixed on
a bearing of 270˚.
The tracks, roller assembly, and suspension bogies of Tank 2 are exposed. Because of its shallow
location, the tank is semi-submerged and at low tide all components above the upper hull including the
turret and gun are exposed to the atmosphere (Figure 7). Many of the removable components have
been detached and long, thin parts have become brittle and cracked. Components at risk of being lost
include the 75mm main gun barrel, hatches, and tow hooks. The tank exhibits the signs of many years of
corrosion and environmental pressure, but less evidence of human disturbance. This may be due to the
fact that this tank is further offshore making it less accessible to shore-based traffic. Unlike the sandy
environment around Tank 1, seagrass is present right up to the tracks and stretches approximately 20 m
from the center of the turret in all directions (Figure 8).
20
Figure 7. The tank shows evidence of rust, surface spalling, and loss of structural integrity (Carpenter
2012).
Figure 8. The tank sits on a bed of seagrass that extends in all directions (Carpenter 2012).
21
Tank 3
Tank 3 is located approximately 1 km south of Tank 1 at a depth of 2 m about 175 m offshore from
Chalan Kanoa Beach, near Saipan World Resort and Saipan Grant Hotel (UTM
).
The tank is semi-submerged and at low tide all components above the upper hull including the turret
and gun are exposed to the atmosphere (Figure 9). Shiny high nickel welds are evident on the upper hull
edges.
The tank, identified as a M4A3 Wet model, constructed principally of rolled and cast homogenous steel,
is 5.91m in length, 2.62m wide and 2.74m in height (Grove 1976:130-131). Tank 3 is orientated with its
bow pointing seaward on a bearing of 295° and the 75mm gun fixed on a bearing of 60° (McKinnon and
Carrell 2011:109-110). The main body of the vehicle is mostly intact but other smaller components are
missing. Most obvious is the loss of the engine cover and cowling and the gun barrel is broken. These
remains are lying on the seabed in close proximity to the tank. This loss may be due to corrosion and/or
physical damage by natural and/or human impacts. There are many areas of active corrosion evident on
the site, indicated by the presence of the typical red/brown “rust” spots on the surfaces of the tank.
There are also signs of accelerated corrosion on the upper sections of the tank (flaking, spalling and
cracking of the metal surfaces and the broken gun barrel) that are cyclically exposed to the atmosphere.
Figure 9. M4 Sherman Tank 3 – stern view (Carpenter 2012).
No human activity was observed on the site at the time of the survey but on a previous survey in 2011,
tour boats and ‘banana’ boats frequently passed near the site, tourists used jet skis on a race course just
north of the site and there was significantly more rubbish around the tank than observed on the Tank 1
22
site (McKinnon and Carrell 2011:114). The fact that the site is located in close proximity to two large
hotels would account for this increase in human interference. However, divers and snorkelers are still
actively encouraged to visit provided they follow the visitation guidelines and do not interfere with the
site (i.e. disturb or attempt to remove any components).
The surrounding seabed is relatively flat, comprising of calcareous sediment interspersed with large
patches of seagrass. The tank is above of the seabed and the lower track is visible. A circular area about
8m2 surrounding the tank is free of sea grass but dead coral and algal forms are present on the seabed.
Extensive algal mats are present on the submerged parts of the hull. High nutrient levels in the lagoon
may be contributing to this extensive algal growth (Denton et al. 2001).
Inter-site Preservation and Deterioration
Based on the results from Richards and Carpenter (2012) there is a statistically significant increase in the
corrosion rate of the upper sections of Tank 1 (above 1m of water depth) versus lower sections. The
primary site variable that affects corrosion is the amount of water movement, which is correlated to
water depth. The higher the position on the tank, the greater the amount of water movement and
oxygen impact to the concreted iron surface, consequently the corrosion rate will increase. This increase
in corrosion rate is further exacerbated by wetting/drying cycles that are experienced by areas of the
tank in the splash zone.
The corrosion rates noted on Tank 3 were very inconsistent in comparison to those measured on Tank 1.
No statistically valid relationships between depth of immersion and corrosion could be observed for this
tank. The areas that are constantly immersed (water depth greater than 1m) will tend to provide more
consistent corrosion results. However, similar to Tank 1, there were very few secondary colonizing
organisms on the concretion with the exception of algal forms and some seaweed species sporadically
located on the upper surfaces of the tank. This suggests that there is a significant amount of water
movement on this shallow site and possible sediment effect during periods of rough sea conditions,
which would significantly reduce colonization rates and increase corrosion rates.
The natural and cultural impacts of the local environment on Tank 3 are more aggressive than those
experienced by Tank 1. More importantly, because there is more tourist activity associated with Tank 3
it is likely that this increase in human interference is causing the accelerated deterioration. Although
Tank 2 was not tested, because it is farther off shore and visited less by casual swimmers and snorkelers,
the rate of corrosion may be more similar to Tank 1, simply due to less human impact. Testing of Tank 2
is recommended. All three tanks are fragile and some efforts to educate visitors about both the dangers
and the fragile nature of the sites are warranted.
23
Landing Vehicle Tracked LVT (A)-4
The remains of two landing craft were included in the 2012 project. Landing Vehicle 1 was first located
during remote sensing surveys conducted by SEARCH, Inc. in 2008 and in 2010 was the subject of
intensive archeological survey (McKinnon and Carrell 2011). A second landing vehicle, which appears to
be the same type, was located during the 2012 survey and but has not been archaeologically
documented.
Landing Vehicle 1 (LVT-1)
The remains of a landing vehicle (Figure 10 ) are located approximately 1,100 m (3,600 ft.) from the
seaplane base at Tanapag (UTM
) (SEARCH, Inc. 2008a:84). The landing craft is a
LVT (A)-4, constructed principally of rolled homogenous steal. Its dimensions are 7.95m in length, 3.25m
wide and 3.11m in height, and it is resting at a slight angle in 2-10 ft. (.7 - 3 m).
Figure 10. Landing Vehicle Tracked 1 – port side view (Carpenter 2012).
The LVT(A)-4 is mostly intact with major structural features and a number of field expedient
modifications still evident, but many other components are missing, such as armor plating across the
deck, tracks and engine room, the guns and many of the controls. The turret has collapsed into the deck
space of the wreck. These losses may have been caused by corrosion but it is more likely that they were
salvaged, possibly during the disarming and disposal process outlined by the U.S. military (McKinnon and
Carrell 2011:123-124). There are some areas of active corrosion evident on the site, indicated by the
presence of the typical red/brown “rust” spots on the surface of the vehicle but these are minimal when
compared to the tanks.
24
The wreck lies on a flat, slightly undulating sandy seabed, comprising of calcareous sediment. There are
small and larger patches of reef surrounding the wreck but none are in direct contact with the wreck.
The stern of LVT(A)-4, which faces shoreward, is partially buried in the seabed and there appears to have
been significant scouring around the bow (seaward side) of the vessel. The rear and lower track are
entirely buried and the upper surfaces are fully exposed to the marine environment. Stormy sea
conditions could result in sand movement that is likely to affect the extent of burial/exposure, however
it is not anticipated that the entire vehicle would ever become totally buried. The wreck was covered in
brown algal forms and some sporadic secondary colonization was evident (e.g. tunicates, soft and hard
corals, seaweed, etc.).
Landing Vehicle 2 (LVT-2)
LVT-2 is located in close proximity to Tank 1, which lies on the southwest side of Saipan, inside the
barrier reef about 180m off shore from Susupe Beach (UTM
). This site was
previously unknown and has not been archeologically recorded. It lies at a depth of about 1.5 m
dependent on the tide (Figure 11). The model has not been positively identified, but it is similar in design
to LVT-1, which is an LVT(A)-4, constructed primarily of rolled homogenous steel. It is orientated parallel
to the shoreline with its bow pointing NNE. The vessel is fully submerged at all times.
The vehicle has almost totally collapsed and a track sprocket is detached and lies on the port side close
to its previously installed position (Figure 11). The tracks themselves are also detached and lie, exposed,
in close proximity to the major vessel remains. The vehicle is incomplete with the main components,
such as the engine and superstructure missing. The remains are partially buried (lower track wheel
bogies were not visible) with the starboard side of the vehicle possessing more sediment coverage. The
starboard side of the vessel is extremely damaged and the loss of structure is quite extensive with the
bow region almost absent, which facilitates sediment ingress into the interior of the vessel. Stormy sea
conditions could result in sand movement that is likely to affect the extent of burial/exposure, however
it is not anticipated that the entire vehicle would ever become totally buried. There are many areas of
active corrosion evident on the site, indicated by the presence of the typical red/brown “rust” spots on
the metal surfaces and at the sediment interface.
The surrounding seabed consists primarily of calcareous sediment with dead coral interspersed around
the site and is relatively flat with short period undulating sand ripples caused by winnowing. Occasional
living coral can be observed on the surrounding seabed but there is very limited coralline growth on the
vehicle structure itself. The wreck was densely covered in brown algal forms. High nutrient levels in the
lagoon may be contributing to this extensive algal growth (Denton et al. 2001).
The poor condition and collapsed state of the LVT2 could be due to a number of human factors including
WWII but storm damage through increased wave action, in such shallow water, has very likely
contributed to its gradual destruction. Recent impact damage was noted on the higher profile part of
the structure and is probably due to a small boat collision. It is not included on the maritime trail, so
visitation to the site by divers and snorkelers is expected to be less than to those wrecks that are listed.
25
Figure 11. Landing Vehicle Tracked 2 – front view (Carpenter 2012).
Inter-site Preservation and Deterioration
Based on the results of Richards and Carpenter (2012) both LVT sites exhibit less corrosion of hull at the
level of the seabed than the shallower, more exposed positions. These lower sections of the vessel near
the sediment/seawater interface are subjected to periodic burial cycles, which would reduce the total
amount of dissolved oxygen impacting the concretion surface, consequently reducing the overall
corrosion rate. It is difficult to say whether LVT- 2 is corroding at a faster rate than LVT-1. Considering
the extent of deterioration of LVT-2 in comparison to LVT-1, particularly when one considers its
shallower depth, it is not surprising that the natural and cultural impacts on LVT2 are greater than those
experienced by LVT1. Visitation at LVT-2 should be limited because of its greater deterioration.
Japanese Merchant Ship, Presumably Shoan Maru aka “Chinsen”
The remains of a WWII merchant ship sunk in Tanapag Lagoon were first examined by PBEC and NPS
divers in 1984. The site is a popular dive location, locally referred to in Japanese language as Chinsen, or
“the shipwreck” and was tentatively identified by Jim Brandt in 1990 as Shoan Maru (Jim Brandt
personal communication to Carrell, 1990). The ship was subsequently revisited in 1984, 1991, 2003,
2008, and 2010 (PBEC 1985:9-10, plates 3A, B, C; Carrell 1991:331-335; Lord and Plank 2003:B12-14;
SEARCH, Inc. 2008b:69-70; McKinnon and Carrell 2011).
If this is Shoan Maru, it was torpedoed in 1943 but did not sink and was towed to Saipan for repairs or
salvage. In 1944, it was attacked again by aircraft and damaged beyond repair. During the post-war
cleanup of Tanapag Harbor the ship was heavily salvaged and cut-down to the waterline because it was
considered a navigational hazard. There are also reports that it was used for explosives training during
26
this time (McKinnon and Carrell 2011:38-40). The remains lie in 35 ft (10.66 m) of water and the overall
length of the site is approximately 274 m (900 ft) (UTM
) (Figure 12). A complete
site plan for the shipwreck has not been produced due to its sheer size. The vessel lies on its starboard
side and although most of the wreck is disarticulated and has collapsed in many areas, major elements
such as the engines, boilers, steering mechanism and superstructure are generally located in close
proximity to their original position. No apparent cargo was observed.
Figure 12. Japanese merchant Ship (bow view) (Carpenter 2012).
The wreck lies on a flat, slightly undulating sandy seabed interspersed with coral outcrops, especially
around the bow area. The vessel is not heavily concreted (e.g. welded overlapping hull plates are
discernible) and it has a general layer of encrustation, which may be derived from calcareous and other
forms of algae. Patches of low profile hard coral formations and larger hard corals have become
established particularly on the side of the port bow, which is angled towards the sea surface. Isolated
coral growth exists on other parts of the vessel but it is minimal. As with other sites in the lagoon
freshwater run-off and associated pollution may be influencing marine growth.
It appears that the wreck is not subjected to regular burial/exposure cycles. Limited sediment coverage
has occurred on some hull structure lying on the seabed but due to the relatively high profile of the
shipwreck remains, this negates extensive burial. The establishment of hard corals indicates that
significant accumulation of sediment does not readily occur. Maximum exposure height of the main
structure is approximately 8m. Curved hull structure lying close to the seabed is undercut and free of
sand accretion. Relatively strong currents are experienced on this site and scouring under ship structure
27
is a likely consequence. The site lies in the comparatively protected lagoon area but due to the extensive
profile of the vessel above the seabed it is conceivable that storms would have an impact on the ship’s
structure. Some of the higher profile sections of the vessel are exposed to the atmosphere at low tides
but the majority of the wreck is fully submerged at all times.
There are a few areas of active corrosion evident on the site, indicated by the presence of the typical
red/brown “rust” spots on the surface of the vessel but these are sporadic compared to the size of the
wreck remains.
Preservation and Deterioration
Richards and Carpenter (2012) reported that there were more hard corals on this wreck when compared
to the shallower wrecks, such as the LVTs and the tanks, which are subjected to more natural and
human interference. And more corals than the other three Daihatsu landing craft. This is probably a
reflection of the larger surface area available for colonization rather than an effect of changes in the
local environment. Interpretation of the differences in the corrosion rates is problematic because of the
sheer size of the site. It appears that in general those portions that have a very high profile, such as
vertical plates, or are more damaged, disarticulated and have collapsed completely, have slightly higher
corrosion rates than those areas that are lower profile and/or are more protected from wave action and
oxygen by deck supports or other structural hull features. It also appears that the bow section is
corroding at an elevated rate compared to the stern section, which would be a reflection of the bow
section’s higher profile in the water column. Some places on the site the metal is completely gone and
can present a hazard to unaware divers.
Possible Auxiliary Submarine Chaser
Divers first examined the remains of a possible Japanese Auxiliary Submarine Chaser (Figure 13) sunk in
Tanapag Lagoon in 1984 (PBEC 1984:10-11). At that time local tour operators referred to the site as the
“submarine.” Careful examination of the remains of this WWII-era ship revealed that it was not a
submarine, but was characteristic of an auxiliary submarine chaser or patrol boat (PBEC 1984:10-11,
Plates 4A, B, C). The site was re-examined in 1990 by NPS divers (Carrell 1991:335-336), in 2008 by
SEARCH, Inc. (SEARCH, Inc. 2008a:67-68) and in 2010 (McKinnon and Carrell 2011). The wreck has not
been positively identified but it is thought to be an auxiliary submarine chaser, possibly the Kyo Maru 8
or Kyo Maru 10, which were both steamers of 341 tons built primarily of steel in 1938 and later
requisitioned for use during WWII (McKinnon and Carrell 2011:45). Their sleek hull design was suitable
for high-speed chases of submarines and they were to be used for whaling after the war.
The site is located on the southwest side of Saipan, inside Tanapag Harbor (UTM
It lies on its starboard side in approximately 30 ft (9.1 m) of water. It is discrete and comprised of a
tightly-clustered linear deposit of bow, disarticulated hull plates and minor structural elements
approximately 60 m (200 ft) long by 15 m (50 ft) maximum width (Figure 13). There is little scattered
debris away from the main concentration and it is fully submerged at all times.
28
).
Figure 13.Photomosaic of possible auxiliary submarine chaser site.
29
Approximately 12 m (40 ft) of the sharp bow section remains intact and it appears that the upper deck
of the bow has been cut away from the vessel. Small hatch holes are present just aft of the stem leading
into the narrow below deck space. The beam of the aft end of the remaining bow section is 3.9-4.8 m
(13-16 ft) wide. The remainder of the ship is badly disarticulated, although the wreckage generally
follows the original line of the ship. No evidence of the engine or boilers is present but there is what
appears to be a section of a funnel in the amidships section. Just aft of the amidships section is where
the ship is highly fragmented and missing components. There is no obvious evidence of the stern,
steering gear, or propellers. The disarticulated nature and missing features makes it difficult to
understand what has happened to the site; however this could be the result of post-war cleanup,
salvage or explosives training operations. In 1984, munitions were visibly scattered in the wreckage.
These were reported as “4 inches in diameter and 15.6 inches long” (101 mm by 396 mm long) (PBEC
1985:10-11). Recent examination of the wreckage site reveals there are still munitions left on the site.
The wreck lies on a flat, slightly undulating sandy seabed interspersed with coral outcrops. The vessel is
not heavily concreted and has a layer of encrustation that is low in profile and patchy in appearance due
to colonizing species variation. Some larger living corals (isolated) have established on the vessel
structure but this is not extensive and reflects the minimal growth of coral on the boulders scattered
about the surrounding seabed. Fish, observed grazing on algal growth, etc. principally on the upper port
side of the bow, inhibit the establishment and extent of growth of potential colonizing marine biota. As
with all sites in the lagoon freshwater runoff and associated pollution may be influencing marine
growth.
It appears that the wreck is not subjected to regular burial/exposure cycles. Limited sediment coverage
has occurred on some hull structure lying on the seabed but due to the relatively high profile of the
shipwreck remains, this negates extensive burial. The establishment of hard corals indicates that
significant accumulation of sediment does not readily occur. Maximum exposure height of the main
structure is approximately 3m. Relatively strong currents are experienced on this site and scouring
under ship structure is a likely consequence. The site lies in the comparatively protected lagoon area but
due to the extensive profile of the vessel above the seabed it is conceivable that storms would have an
impact on the ship’s structure. There are a few areas of active corrosion evident on the site, indicated by
the presence of the typical red/brown “rust” spots on the surface of the vessel but these are sporadic
compared to the size of the wreck remains.
A second site (Possible Auxiliary Submarine Chaser 2), identified by SEARCH, Inc. in 2008 may include the
remains of sections of this site. A closer inspection of the second site in 2010 and again in 2012 revealed
similar types of hull construction techniques (i.e. welded hull plating). However there were also sections
of wreckage that did not match the first site’s construction type (i.e. riveted hull plating). The sections of
similar hull plating may indicate that the vessel was cut up and/or blown up just aft of amidships during
post-war operations and dumped on the opposite end of the channel. Alternatively, the second site
could be the location of a second auxiliary submarine chaser wreck known to have sunk in the lagoon.
30
Preservation and Deterioration
There appeared to be no obvious relationship between water depth and corrosion rates at this site
according to Richards and Carpenter (2012). Isolated iron features are corroding at a faster rate than
elements that are in physical connection over a larger surface area (i.e. the hull structure). Furthermore,
articulated hull structure and an anchor lying in direct contact with the collapsed starboard hull plates
that are lying flat on the seabed and in a more protected area had a lower corrosion rate than isolated
remains. The starboard side is corroding at a slightly higher rate compared to the port side, which would
be a reflection of the increased damage, disarticulation and collapse on this side of the vessel.
This site is covered with relatively thin aerobic concretions and more hard corals present compared to
the shallower wrecks, such as the LVTs and the tanks. This suggests there is a significant amount of
water movement on this site despite the greater relative water depth (10m). Generally the thicker the
concretion and corrosion layer, the lower the rate of deterioration but only if the concretion layer
remains essentially undisturbed (i.e. no damage occurs through human/natural interference).
The auxiliary submarine chaser may be corroding at a slightly faster rate than the Japanese merchant
ship in part because of natural phenomena (i.e. current or water movement or cyclonic activity) or
human intervention (i.e. salvage, explosive damage during WWII).
Unidentified Steamship
An unidentified steamship is located on the southwest side of Saipan, inside Tanapag Harbor at a
maximum depth of about 10m (UTM
). This site was recorded in 2008 by SEARCH,
Inc. (2008a:69) and has not been archaeologically documented; it was only cursorily investigated in
2012.
The ship has not been positively identified and is constructed primarily of steel (Figure 14). There is
considerable damage and although many pieces are disarticulated they lay in close proximity. Various
sections are bent, twisted and distorted suggesting war damage rather than a consequence of impact
when it sank or collapse due to corrosion. A dislodged ship’s boiler is basically intact as are some storage
tanks. The engine and propeller were not observed and may have been salvaged. Overall impressions of
the site were limited by poor visibility. Generally the metal structures appear to be in a strong and
robust condition. There were no areas of active corrosion evident.
The wreck lies on a flat, slightly undulating sandy seabed interspersed with coral outcrops of varying
dimensions. The vessel is not heavily concreted and has a layer of encrustation that is low in profile and
patchy in appearance due to colonizing species variation. Some larger living corals (isolated) are
established on the vessel structure, but this is not extensive and reflects the minimal growth of coral on
the boulders scattered about the surrounding seabed. At the time of the survey there were significant
quantities of suspended material in the water column. This limits light penetration and sediment
deposition from the water column would deter some forms of marine growth. The source of the
increase in water turbidity was not determined but may be derived from land runoff. Relatively few fish
31
were observed at this site. As with all sites in the lagoon freshwater runoff and associated pollution may
be affecting marine growth.
It appears that the unidentified steamship is not subject to regular burial and exposure cycles, at least to
any great extent. Sediment has partially covered the lower profile areas, such as collapsed hull/deck
plates, and the mast/king post structures lying on the seabed were approximately one third buried in
sediment. Maximum exposure height of the main structure was approximately 4m. However, due to the
high profile of the remains and the establishment of hard corals in some areas it is likely that significant
accumulation of sediment does not readily occur on this site.
Figure 14. Unidentified steamship – boilers (Carpenter 2012).
Preservation and Deterioration
Based on the results from Richards and Carpenter (2012) there is a small but statistically valid decrease
in the corrosion of this site, which suggests that it is deteriorating at a slower rate than both the
auxiliary submarine chaser and the Japanese merchant ship. This would further suggest that human
interference (i.e. recreational diving activities) is having some impact on the deterioration rate of the
Japanese merchant ship and auxiliary submarine chaser sites. However, the local environment may also
be contributing to this decrease in corrosion rate on the unidentified steamship site.
Aichi E13A “JAKE”
The Aichi E13A aircraft wreck is located in Tanapag Lagoon approximately 1200 ft (365 m) south of the
eastern edge of Mañagaha Island and 900 ft (275 m) west-northwest of a large coral reef patch exposed
32
at low tide (UTM
). This site was first investigated by PBEC and recorded as CNMI
Historic Property Register Site 9 in the 1980s (1984:14). It was also visited and photographed by NPS
volunteer divers William Cooper and Dennis Blankenbacker (Miculka et al. 1984:2), was included as Site
number 5 in Carrell’s report (1991:502-508) and archeologically investigated in 2010 (Figure 15). This
aircraft lies within a Marine Conservation Area and is frequented by snorkelers and is popular with dive
instructors for novice diver training.
The aircraft remains measure 36 ft (10.9 m) in length from bow to stern with a wingspan of 46 ft (14 m)
and it is constructed most probably of duralumin (an aluminum alloy). The aircraft is relatively intact
lying inverted on the seabed. It is listing to port with the end of the port wing buried in the surrounding
sediment and the starboard wing rising above the seabed with no support. One of the two floats is
absent the other lies in close proximity to the port wing. The aircraft structures and components,
although damaged, remain in relatively good condition and still retain strength and resilience as attested
by the unsupported starboard wing. A piece of landing gear is located a short distance from the tail of
the aircraft but does not appear to be associated with the aircraft (McKinnon and Carrell 2011:55).
Overall damage to the aircraft structure is not readily distinguishable between crash, storm or potential
war damage, however it is thought that the aircraft was intentionally scuttled (McKinnon and Carrell
2011: 57).
Figure 15. Overview photograph of Aichi E13A JAKE (Carpenter 2012).
The aircraft remains are located on a relatively flat, undulating seabed with sporadic large coral
outcrops. The port wing and the end of the tail are partially buried in fine sediment and localized,
seasonal exposure/reburial cycles occurs in this area as changes to sediment levels are noted with each
33
visit. However, most of the remains are exposed and total burial seems unlikely to occur. The maximum
exposure height of the main structure is approximately 1.5 m above the seabed. A thin mucilaginous
layer and algal forms cover the aluminum surfaces with coral growth evident on various parts of the
aircraft, especially near the engine area where the presence of ferrous components encourage more
secondary colonization. The algal growth on the under-surfaces is denser than the growth on the upper
surfaces. A greater variety of colonizing species are also apparent on the under-surfaces. A steady and
generally light current affecting the site did not visibly move sediment. The site is relatively shallow
(6m) and may be affected by turbulent seas generated by storms and cyclones.
Preservation and Deterioration
According to Richards and Carpenter (2012) the metal composition of the wings, aft fuselage, propellers
and boss head are very similar. The forward part of the fuselage is corroding at a slightly slower rate
than other portions of the articulated remains. This is expected as this area of the fuselage is in direct
electrical contact with what appears to be machinery associated with the engine, resulting in more
contact with iron, copper and other less reactive metals that lowers the overall corrosion rate in this
area. The disarticulated float is corroding at a slightly higher rate than the other areas suggesting that
the aluminium alloy composition of the float is different from the rest of the aircraft and contains less
copper and more aluminium making the corrosion potential higher.
Some sections of the Jake, such as the wings and the float, showed discreet areas of perforation but the
extent was significantly less than that observed on the TBM Avenger. This suggests that the
environment at the Avenger site is considerably more aggressive than that experienced by the Jake. This
difference in damage can be attributed, in part, to the fact that the Avenger is much shallower and sits
on a reef platform that is subject to much greater overall water movement.
Kawanishi H8K “EMILY”
The Kawanishi H8K (Type 2 Large Flying Boat) aircraft (Figure 16) was recorded in 1984 as CNMI Historic
Property Register Site 1 with a recommendation that it be made a part of an underwater park (PBEC
1984:8). It was visited and photographed by NPS volunteer divers William Cooper and Dennis
Blankenbaker (Miculka et al. 1984:2) and was reported as site number 4 in Carrell’s report (1991:502).
During a remote sensing survey of Tanapag Lagoon it was again relocated (UTM
)
and identified in 2008 (SEARCH, Inc. 2008b:73) and was intensively mapped in 2010 (McKinnon and
Carrell 2011).
The main section of wreckage consisting of wing and engine nacelles rests inverted on the seabed. Other
components including the four engines and propellers, bow gun turret, cockpit, painted sections of the
fuselage, and smaller pieces are scattered over a large area. Most of the aircraft structures and
components, although damaged and disconnected, remain in relatively good condition and still retain
strength and resilience. However, extensive corrosion is evident on the nacelles. Because damage to the
structure is extensive and the site is highly disarticulated and scattered, it suggests a catastrophic
wrecking event (McKinnon and Carrel 2011:70).
34
This aircraft is a popular dive site but there is little evidence of anchor damage. Changes to the site from
visitation are clear, however. The cockpit was repositioned so divers can sit in the pilot’s seat and many
smaller artifacts and components were moved from their original positions and are piled up at the
nearby Korean and Japanese monuments.
The remains are located on a relatively flat, undulating seabed with sporadic coral outcrops in the
vicinity. Some lower profile sections, such as the wings are covered in a thin layer of fine sediment.
There appears to be evidence of localized, seasonal exposure/reburial cycles on the site, but most of the
remains are exposed and total burial seems unlikely to occur. The maximum height of the main
structure is approximately 1.5 m above the seabed. A thin mucilaginous layer and algal forms cover the
aluminum surfaces and coral growth is evident on various parts of the aircraft, especially near areas
where the presence of ferrous components encourages more secondary colonization. A steady and
generally light current affecting the site did not visibly move sediment. The plane is in relatively shallow
water (9m) and may be affected by turbulent seas generated by storms and cyclones.
Figure 16. Kawanishi H8K (EMILY) (Carpenter 2012).
Preservation and Deterioration
The results from Richards and Carpenter (2012) indicated that the surface of the aluminium alloy
sections of the Emily wing are in good condition with the exception of the nacelles where discrete areas
of pitting and perforation of the residual metal are obvious. The metal composition of these areas is
very similar, so similar corrosion rates are expected. The cockpit, the plane section of the aft nacelle, the
propeller and boss of engine 3 had similar rates, while the boss and propeller of engines 1 and 2 are
deteriorating at a slightly slower rate. Because engines 1 and 2 are considerably more intact and their
35
bosses and propellers are connected to engine components, this has afforded them some additional
protection from corrosion.
Martin PBM Mariner
An unidentified aircraft wreck (Figure 17) was first recorded by PBEC as CNMI Historic Property Site
Number 7 (1984:12-13). It was visited again in 1990 and reported on by Carrell (1991:508). Preliminary
investigations in 1984 postulated that it was a Japanese type 99 2EFB “Cherry.” In July 2009 and 2010
the site was revisited and extensively mapped leading to its identification as a Martin PBM Mariner
(McKinnon and Carrell 2011). The wreck is located in Tanapag Lagoon approximately 600 m east of
Mañagaha Island and 100 m north of an exposed patch reef (UTM
). This aircraft
lies within a Marine Conservation Area.
The aircraft is constructed most probably of duralumin (an aluminum alloy) and is 24.33 m in length,
8.38 m high and had a 35.97 m wingspan. The main component is sitting inverted on the seabed at a
depth of approximately 7 m (23 ft) and consists largely of the wings with twin engine compartments
(minus engines and propellers), gun turrets, tail sections and a portion of cockpit, etc. Most of the
aircraft structures and components, although damaged and disconnected, remain in reasonably good
condition retaining strength and resilience. Because damage is extensive and the site highly
disarticulated and scattered over a relatively large area, it suggests a catastrophic wrecking event
(McKinnon and Carrell 2011:85).
This site is frequented by divers and there is evidence of recent anchor damage. Many smaller artifacts
and components have been moved from their original positions and piled up in one area on the site.
Local informants told PBEC researchers that a radio and other electronic instruments removed from the
site had U.S. markings on them (PBEC 1984:12). It is also possible that some form of salvage occurred as
the engines and propellers are missing.
The remains are located on a relatively flat, undulating seabed with sporadic large coral outcrops. Some
lower profile sections, such as the wings, were covered in a thin layer of fine sediment. There appears to
be evidence of localized, seasonal exposure/reburial cycles, however, most of the remains are exposed
and total burial seems unlikely to occur. The maximum exposure height of the main structure is
approximately 1.5 m above the seabed. A thin mucilaginous layer and algal forms cover the aluminum
surfaces, and coral growth is evident on various parts of the aircraft especially near areas where the
presence of ferrous components encourages more secondary colonization. A steady and generally light
current affecting the site did not visibly move sediment. Because the site is relatively shallow (7 m) it
may be affected by turbulent seas generated by storms and cyclones.
36
Figure 17. Overview shot of central portions of Martin PBM Mariner site; note dihedral wing (D.
McHenry, June 2010).
Preservation and Deterioration
The results from Richards and Carpenter (2012) revealed that the surfaces of the aluminium alloy
sections of the Mariner, such as the wings, floats, engine cowlings, nacelles, etc. were quite corroded
and there was significant pitting and perforation of the residual metal. The remains of the Mariner are
more deteriorated than the Jake even though they lie in a similar environment at similar depths. This
can be attributed in part to increased stress and metal fatigue caused by the wrecking event and
because the Mariner is more disarticulated and spread over a wider area with less opportunity for the
beneficial effects of proximity to other remains to help preservation.
TBM Avenger
A wrecked TBM Avenger (Figure 18) is located just inside the barrier reef near the north edge of the
main channel entrance to Tanapag Harbor in approximately 7-10 ft (2.5-3 m) of water (UTM
). The site was recorded by PBEC as CNMI Historic Property Site Number 13 (PBEC
1984:16-17), visited and photographed by NPS volunteer divers William Cooper and Dennis
Blankenbacker in 1990, reported by Carrell (1991:508), and extensively surveyed in 2010 (McKinnon and
Carrell 2011). Identification is based on the wing size, width, fuselage and landing gear configuration
(Carrell 1991:508). The aircraft lies within a designated Marine Conservation Area.
The aircraft is most probably constructed of duralumin (an aluminum alloy) and was 12.19 m in length,
5.00 m high and had a 16.51 m wingspan. The aircraft is inverted on the seabed with a concentrated site
distribution consisting of the central portion of the wing. The wing measures approximately 50 ft (15 m)
in length and is 8 ft (2.5 m) in width at the fuselage. The aircraft remains are mostly submerged,
however the hydraulic landing gear, which is in the fully extended position, is exposed to the
37
atmosphere at extreme low tides. The aircraft is missing its tail section, engine and propeller. A few
small sections of wreck are scattered within 65-130 ft (20-40 m). The sections include part of a radial
engine, a section of fuselage with an observation port and a turret ring. Also, what appears to be a radio
box is approximately 20 ft (6 m) aft of the wreckage.
Figure 18. Grumman TBM Avenger (Carpenter 2012).
The aircraft remains are located on the top of the barrier reef that creates Tanapag Harbor and are
subjected to considerable water movement due to this high energy environment. Large quantities of
dead coral are strewn over the seabed, which is comprised of coarse-grained calcareous sediment. The
surviving aircraft structure is slowly being integrated into the reef through the growth of corals. The
maximum exposure height of the main structure is approximately 1.5 m above the seabed. The
aluminum skin of the wings is corroded and has a number of irregular holes and smaller perforations.
Its reef top position implies that the Avenger remains are always exposed and overall sediment burial is
very unlikely. Sand is present inside and in front of the engine bay cavity which may scour in more
turbulent conditions. Localized, limited and partial exposure cycles may occur in this area. The
development of coral growth may potentially cover the aircraft remains with time however the limited
size and extent of hard coral growths on the aircraft structure after some 65 years of immersion is likely
to be a consequence of the more dynamic localized environment (including storm damage).
The smooth metal skins of aircraft generally seem to inhibit the establishment of larger forms of marine
biota unless there is a break in the surface or a ferrous metal is present. The exception is colonization of
the under-surfaces of areas, such as wings, where light levels may exist that are similar to those found in
38
the entrance to underwater caves, etc. These conditions suit the establishment of lower profile sponges
and small gorgonia (sea fans) that colonize these features. The protruding landing struts are essentially
devoid of marine growth because they are subjected to a relatively strong and constant current as water
streams over the reef into the lagoon. The much lower profile, and largely reef-shielded wing remains
are less affected by excessive water movement.
Preservation and Deterioration
Richards and Carpenter (2012) reported that the condition of the aluminum alloy is poor compared with
those aircraft wrecks located on sandy sediments in calmer areas of the lagoon. It is not possible to
distinguish the cause of damage to the site between crash, storm or visitation. The site is shallow (3m)
and its reef top position means that it must be affected by turbulent seas generated by storms and
cyclones. Local surfers use the protruding landing gear as boat moorings and bright, bare aluminum
surfaces are evident suggesting direct impact in these areas.
Consolidated PB2Y Coronado
A Consolidated PB2Y Coronado (Figure 19) is located on the southwest side of Saipan, inside Tanapag
Lagoon at a depth of about 7 m (UTM
). The aircraft remains were shown to
project staff by a captain of the local tourist submarine and positively identified as a Consolidated PB2Y
Coronado in 2012, a U.S. four engine maritime patrol flying boat, constructed most probably of
duralumin (an aluminum alloy) and was 24.16 m in length, 8.38 m high and had a 35.05 m wingspan.
The aircraft remains are disconnected and scattered over a very large area. Among the components
identified were a single detached engine, flight instruments with dials, aerial mast, cockpit canopy with
windscreen wipers, hatch covers, float support, a chair, concreted forks and a number of unidentifiable
hull sections. Damage is extensive and the site is disarticulated and scattered over a large area,
suggesting a catastrophic wrecking event. The presence of smaller artifacts (e.g. chair, forks, etc.) in situ
supports the fact that this site is not visited frequently and human interference has been minimal to
date.
The remains are located on a relatively flat, undulating seabed comprising primarily of fine calcareous
sediment. There are sporadic large coral outcrops in the vicinity and a very large reef formation lies to
the south east of the major site concentration. Some lower profile sections, such as the wings and tail
planes were covered in a thin layer of fine sediment. Most of the remains are exposed and total burial
seems unlikely to occur. The maximum exposure height of most remains is approximately 50cm with the
engine being the exception rising about 1.5 m above the seabed. A thin mucilaginous layer and algal
forms cover the aluminum surfaces and coral growth is evident on various parts of the aircraft,
especially near areas where the presence of ferrous components encourages more secondary
colonization. A steady and generally light current affecting the site did not visibly move sediment.
Because the site is in shallow water, it may be affected by turbulent seas generated by storms and
cyclones.
39
Preservation and Deterioration
Based on the findings of Richards and Carpenter (2012) the surfaces of the aluminium alloy sections of
the Coronado exhibited significant pitting and perforation of the residual metal. The average corrosion
potentials of the float strut, an area of unidentified wreckage, and the control panel suggest that these
have different metal compositions (i.e. higher aluminium contents) than other sections resulting in a
higher rate of corrosion. The Coronado is more deteriorated than the Jake and Emily even though they
are in similar environments at similar depths. This is due in part to the increased stress and metal fatigue
resulting from damage and because it is disarticulated and spread over a wider area, and therefore not
afforded the beneficial effects of proximity to other sections to slow rates of deterioration.
Figure 19. Consolidated PB2Y Coronado (Carpenter 2012).
40
Chapter 4: Threats and Impacts
Cultural Threats and Impacts
Cultural or human impacts to underwater sites are not that much different from the impacts on their
terrestrial counterparts. Over-visitation, looting, vandalism, removal or movement of artifacts, and
development affect all cultural heritage sites. The major difference between terrestrial and underwater
sites it that is inherently more difficult to identify, mitigate and monitor impacts to underwater sites that
are literally-out-of-sight and often out-of-mind to managers and to all but a select few visitors. This
often leads to an accumulation of significant impacts. There is no question that the underwater sites in
Saipan have suffered as a result of the fundamental difference in their “visibility.”
Most often it is development with dredging, filling and construction that impacts sites. Because the
lagoon is shallow and the shipping channel has existed and been routinely dredged for many years, few
new developmental impacts affect these sites. This type of impact has already happened, most of which
was immediately after the successful capture by US forces, subsequent cleanup, and post-war salvage.
Going forward, new development can be regulated and potential impacts monitored and mitigated.
In what may be a unique problem, the majority of impacts to Saipan’s submerged sites are a direct
result of visitation, specifically anchor or mooring damage, looting, moving artifacts, and acts of
vandalism. This presents a challenge for managers who have limited staff, time, and funding.
Anchor and Mooring Damage
Within the Mañagaha Marine Conservation Area (Figure 20) where there are restrictions on anchoring,
steps to prevent damage have been underway since the heritage trail was developed in 2009. The CRM
office is in the process of installing mooring buoys at the more heavily visited sites including the
Kawanishi H8K “EMILY” site (which now has two moorings) and repairing and replacing moorings on the
Japanese merchant ship. Plans are currently in the works for installing more moorings on heritage trail
sites within the conservation area.
However, for those sites outside of the Marine Conservation Area there is no mandate or support for
installing moorings so anchor damage is a greater risk. Conversations with a local boat driver disclosed
the Avenger’s landing gear is regularly used as a boat mooring for local surfers (Sheldon Preston
personal communication, 2010). The effects of mooring are seen on the landing gear where exposed
metal is obvious (Figure 21). Continued use of the landing gear as a mooring will eventually cause severe
damage if the boats collide with the aircraft or break the landing gear during rough swell conditions.
41
Figure 20. Map of Mañagaha Marine Conservation Area (CRM).
Figure 21. Avenger landing gear, note shiny metal where concretion has been rubbed away due to
mooring (Carpenter 2012).
42
Looting and Moving Artifacts
Looting and the movement of artifacts on site are probably the most common and most destructive
impacts. By their very nature modern, war-related sites have a considerable amount of associated small
portable objects. For many years divers have been removing artifacts or simply rearranging them on
site. Because this activity impacts the historical and archeological context or fabric of a site, it can make
identification more difficult and also affects the information that can be learned from the way in which
the site was created (i.e. crashing, sinking, and dumping).
Of the nine sites on the trail, four have had some form of looting or movement of artifacts. At the
Daihatsu 1 site the steering wheel for the craft was propped up on the stern deck and glass bottles, not
associated with the site, are regularly re-arranged on the deck. The Japanese merchant ship site aka
Shoan Maru includes a Korean monument on which 50-caliber bullets have been placed and rearranged
into patterns (Figure 22). It is uncertain where the bullets actually originated and if they are from the
wreck or another site.
Figure 22. Korean monument with .50 caliber rounds (Gauvin 2010).
Divers are having a major impact on the Kawanishi H8K (EMILY). Impacts in the cockpit area were
photographically documented during the February and June 2010 field seasons (McKinnon and Carrell
2011) (Figure 23). The cockpit control panel was shifted and subsequently balanced on its mount giving
the appearance of its original position. Furthermore, the steering column was moved to the opposite
side of the cockpit chair. Consultation with dive operators confirms that tourist divers like to take
photographs while seated in the cockpit. This behavior is detrimental to the preservation of the site and
will eventually destroy these unique features.
43
Figure 23. Cockpit configuration changes over time on Kawanishi H8K (Bell 2010).
A ladder-like metal object was propped up against the southern side of the aircraft wing. Smaller
artifacts located on the Kawanishi site have been moved from their original positions including stacking
ordnance and gas cylinders around the Japanese monument (Figure 24). It is possible that these artifacts
were moved to enhance the memorialization of the site. While such movement of artifacts on a historic
site destroys contextual evidence, from a socio-cultural perspective it reflects the different norms by
which cultural groups memorialize sites and commemorate their visits.
Figure 24. Japanese monument surrounded by gas cylinders and other moveable objects (Seymour
2012).
The present locations of the four engines on the Kawanishi may also be an example of a different sort of
cultural impact. There is some question whether the engines are in situ or were moved to their current
44
locations. Engine 2, in particular, has a questionable position as it is standing on its edge on a coral head
in a fashion that implies it may have been placed intentionally for the purposes of aesthetics and a
potential photographic setting.
While there is no evidence of any systematic salvage at the Kawanishi, there are indications of
opportunistic salvage. A local dive shop owner is in possession of “identification plates” that were
reportedly removed from the plane (McKinnon and Carrell 2011) (Figure 25). Upon further research, the
plate was identified as from the plane’s wireless radio. Separately, the Historic Preservation Office was
informed of a piece of aircraft deposited on a beach, suspected to have come from the Kawanishi
(Figure 26). After careful review of photographs it was determined to be a portion of the nose area.
There is no way to accurately determine the extent of or damage caused by such activities.
Figure 25. Wireless radio Identification plate recovered from the Kawanishi H8K sites (masadivesaipan.com, accessed June 2010).
(N.B.: The bottom row of characters is Matsushita Musen Kabushikigaisha. Matsushita is the
predecessor company to Panasonic, musen means wireless, and kabushikigaisha is a company
classification. The row above the company name is the manufacture date beginning with the year (02),
month (8) August, and production or model number (19). The year 02 is most likely a reference to the
imperial year system; the zero refers to the production year 2600 (1940), 2602 then is 1942.)
Figure 26. A portion of a Kawanisihi H8K reported to the HPO (Rogers 2010).
45
With the implementation of the underwater trail, the Martin PBM Mariner is being visited more
frequently, which is having an impact on its preservation and integrity. Ordnance and smaller artifacts
are being moved from their original positions; this was documented in February 2010 and their new
locations in June 2010 (Figure 27). They were gathered into one area in a manner similar to that
observed at the Japanese Memorial at the Kawanishi H8K site. Fifty-caliber rounds were also located on
site with gunpowder spilling from the casings; it is possible that this was caused by divers breaking the
casings open. A leather shoe sole was also moved to this location and piled with the rounds.
Figure 27. Artifact pile created by divers and altered over time (Bell 2010).
Because the plane is missing its engines and propellers, and there is no chance they could have
disintegrated; salvage or removal is a strong possibility. Interestingly, there are two four-blade
propellers at a nearby “site” that was created by the local submarine tour company (Figure 28). While
these propellers may belong to the Mariner site or some other aircraft wreck, the artifact pile represents
another type of cultural impact – faux site creation.
The artifacts at the “site” include: ammunition boxes, 50-caliber rounds, two propellers, hatches,
multiple aircraft seats and other unidentifiable artifacts. It is uncertain when this pile was created but it
is now touted to submarine tourists as an “aircraft” wreck. This type of misinformation to tourists is
disappointing and now that the artifacts have been displaced, their context will never be recovered and
historical and archeological data has been lost forever.
46
Figure 28. Faux airplane wreck created for a submarine tour with artifacts gathered from sites in the
vicinity. Note four-blade propeller (foreground) and gun in background (Carpenter 2012).
Acts of Vandalism
Vandalism, whether intentional or unintentional, also impacts submerged sites. Local tour boats
frequent the Sherman tanks and “banana” boats pull passengers by for a closer look. Tour operators
were observed demonstrating how to climb on the tanks and/or swing off the gun barrels, a dangerous
and destructive activity. It is suspected that similar behavior at Tank 3 resulted in a portion of the end of
the barrel breaking. Certainly there will come a time when the barrels have degraded and can no longer
sustain the weight of people jumping or swinging off them.
Graffiti has been etched into the mucilaginous layer on the aluminum surface of the Kawanishi H8K
aircraft on the wing of the aircraft and the gun turret (Figure 29). These graffiti areas were not noticed
during the February 2010 field season but were found in June 2010. The etching on the bow turret is
indiscernible; however, initials appear to be etched on the starboard wing. The letters or characters are
not distinguishable. They likely represent the initials of the inscriber and may have been etched to
personally memorialize one’s attendance at the site. There are several places on the island where initials
have been carved into objects. For example, at Suicide Cliff there are large cacti whose pads have been
used to etch initials of visitors. Particularly interesting is the fact that the majority of the initials are of
Asian languages including Japanese and Korean.
47
Figure 29. Graffiti etched into the Kawanishi H8K (Bell 2010)
Tourism Services Impacts
Certain tourist services have a direct impact on sites. The Sherman tanks are subjected to an enormous
amount of “foot traffic” when Jet Skis and banana boats pass nearby. These vehicles typically create
wakes that wash over the tanks causing a cyclical pattern of wetting and drying. This affects the
immediate site environment by increasing oxidization levels in the water that in turn increase corrosion.
Tank 3 shows signs of increased corrosion rates that are most likely due to its proximity to two large
resorts and a Jet Ski course. The newly investigated LVT 2 site showed signs of a recent impact where a
boat or Jet Ski hit the vehicle.
Unsightly rubbish, while not a serious impact, is found at sites. Because the Sherman tanks are located
just offshore from several large resorts and locally popular picnic beaches, rubbish including plastic bags,
beer and soda cans, plastic forks and fishing line accumulate. Not only can trash present hazards to
snorkelers and divers, but they certainly have an impact on marine life. Turtles and fish may ingest
pieces of plastic; an adverse impact that is well documented elsewhere.
Another tourism service impact that affects both the environment (i.e. marine organisms) and cultural
heritage is the operation of the local tourist submarine. The Japanese merchant ship is on the tour and
as it moves towards the shipwreck the submarine disperses large quantities of fish feed including rice to
attract fish to the wreck. No information is available on the impacts of repeated discharges of rice or
other non-marine organics into the water in the vicinity of the wrecks. Does this lead to higher rates of
pollution and therefore deterioration? More concerning are the submarine’s thrusters that blow onto
the shipwreck as it makes its turn. Because the thrusters are powerful enough to move portions of the
iron plating up and down, this will almost certainly lead to increased deterioration in those areas and
significant impacts.
48
Memorialization
The process of memorialization affects the sites in Saipan through the addition of outside material,
aggregation of moveable objects, potential damage to buried artifacts, and altering the overall integrity
and “feeling” of a site. Two monuments on the Kawanishi wreck site are dedicated to those lost during
the battle. The first and largest monument is located north of the port wing and was placed there by
Challenge! Earth Exploration, a television adventures series that previously aired on the Korean
Broadcasting System (KBS). The larger panels of the monument state in both Korean and English, “Spirits
sacrificed in the Pacific War, rest in peace, KBS Challenge! Earth Exploration, Inmolt Engineering Co.
Ltd.”
One side of the square monument lists the director, producer and others involved in the television
program’s placement of the monument. The other side has a series of memorial poems and statements
(Jack London personal communication, 2010). One poem on the side of the monument dedicates the
memorial, “to spirits who hired to the compulsory military service and died during the Pacific War,” and
an additional poem relates, “Anger, tears and grunge.” As only one translation was received for the
poem, it is uncertain whether the word “grunge” is accurate or if it is in fact “grudge.” These remarks
clearly indicate a Korean connection with those lost during the Battle of Saipan and further emphasize
that Korean soldiers were forced into service. Although, it is uncertain why a Japanese aircraft wreck
was chosen as the placement site for a Korean monument.
The second, Japanese monument is much smaller than the Korean and small artifacts from around the
wreck site have been piled around it. The Japanese monument appears to be an epitaph for an
individual (Jun Kimura personal communication, 2010). As the first few letters are in a special writing
style, they are indiscernible; however the last four characters translate to “Underwater (seabed) War
Memorial.” The shape of the monument is similar to that of a wooden stupa used for modern Japanese
Buddhist style graves.
There is a third monument located on the Japanese merchant ship. Much larger than the two
monuments on the Kawanishi, this monument is off the starboard side of the bow and is dedicated to
Korean conscripts lost during the battle. Together all the monuments do not largely affect the historical
and archeological context of a site but when considered alongside the developments occurring on land,
which include significant increases in the numbers, more monuments might begin to impact the context.
Environmental Threats and Impacts
In general, the physical-chemical measurements of the local environment surrounding the wreck sites in
Saipan reported by Richards and Carpenter (2012) are typical for a shallow, near coastal, open
circulation, oxidizing marine environment, where corrosion rates are likely to be relatively high for both
ferrous (iron) alloy wrecks and aluminum alloy aircraft. All of the wrecks and the aircraft were mostly
exposed with only very thin layers of sediment covering some lower profile areas lying on the seabed,
which would be particularly mobile during periods of excessive water movement (i.e. storm and cyclonic
activity). Hence, natural protection via seasonal sediment burial would be very unlikely for any of the
wrecks identified to date.
49
The results from the examination of a number of locations on each of the iron alloy sites are revealing.
While each site and locales with in each site are unique they can be characterized as: actively corroding
and will continue to do so until all the iron is consumed, are in equilibrium where the corrosion has
slowed and formation of concretions have stabilized the underlying metal, or have moved into the
passive zone where little if any residual metal remains and only the concretion layer is left. A shift from
equilibrium to active corrosion can be triggered if the concretion layer is damaged through human or
natural interference. Because the vast majority of the sites are in shallow water, wind and wave driven
sand and debris from a strong storm or cyclonic activity has the potential trigger a return to active
corrosion.
When comparing sites is it important to look at the data in conjunction with the environmental and site
dynamics information. While all the tanks are the same type and in generally similar environments,
there is a statically significant increase in the corrosion rate of Tank 3 compared to Tank 1. This suggests
that the natural and cultural impacts of the local environment on Tank 3 are more aggressive than those
experienced by Tank 1. More importantly, as there appears to be more tourist activity associated with
Tank 3, it may be this increase in human interference that is causing the accelerated deterioration of
Tank 3 (Richards and Carpenter 2012).
It is difficult to say whether the LVT 2 is corroding at a faster rate than the LVT 1 as all average
measurements are within their respective statistical errors. However, considering the extent of
deterioration of the LVT 2 as compared to the LVT 1 it would appear that the natural and cultural
impacts on the LVT 2 would be greater than those experienced by the LVT 1 (Richards and Carpenter
2012).
Based on the results from the Daihatsu wrecks some differences in corrosion rate can be ascertained.
Daihatsu 2 may be corroding at a slightly faster rate than both Daihatsu 1 and 3. This is not unexpected
as it is known that isolated iron artifacts and steel hull structures that have been damaged either
through natural phenomena (e.g. cyclonic activity) or human intervention (e.g. salvage, explosive
damage during WWII) possess higher corrosion rates than those hull structures that are relatively intact
(i.e. Daihatsu 1 and Daihatsu 3). In addition, it appears that Daihatsu 1 and Daihatsu 3 are corroding at
relatively similar rates, despite the fact that Daihatsu 3 is a much shallower site, where it would be
expected that the corrosion rate would be slightly higher. This would seem to suggest that human
interference (i.e. recreational diving activities) is having some impact on the deterioration rate of the
deeper Daihatsu 1 site (Richards and Carpenter 2012).
It is difficult to determine any changes in corrosion behavior of the larger shipwrecks, Japanese
merchant ship, the auxiliary submarine chaser and the unidentified steamship as most average
measurements are within their respective statistical errors. However, based on the results, it appears
that there is a small but statistically valid decrease in the corrosion potential of the unidentified
steamship suggesting that this vessel is corroding at a slower rate than both the auxiliary submarine
chaser and the Japanese merchant ship. Further, the results from the auxiliary submarine chaser suggest
50
that it may be corroding at a slightly faster rate than Japanese merchant ship. This is not unexpected as
steel hull structures that have been extensively damaged (i.e. auxiliary submarine chaser) possess higher
corrosion rates than those hull structures that are relatively intact (i.e. Japanese merchant ship and the
unidentified steamship). This would seem to suggest that human interference (i.e. recreational diving
activities) is having some impact on the deterioration rate of the Japanese merchant ship and auxiliary
submarine chaser sites as the unidentified steamship site is not on the heritage trail. However, the local
environment (i.e. increase in turbidity) may also be contributing to this decrease in the corrosion rate on
the unidentified steamship site (Richards and Carpenter 2012).
The five aluminum alloy aircraft wrecks in Saipan present their own set of problems. Most of the aircraft
manufactured during WWII used a variety of aluminum alloys consisting mainly of aluminum but
including varying concentrations of minor alloying constituents (e.g. iron, copper, magnesium,
manganese, zinc and silicon) in order to change the functionality of the aluminum. One of the most
common alloying metals used was copper (e.g. Duralumin), which was added to aluminum to increase
its strength. However the presence of copper dramatically decreased the corrosion resistance of the
metal to seawater. The other issue that will increase the deterioration rates of the aircraft is galvanic
corrosion, where the more reactive aluminum alloys corrode faster effectively protecting the more
noble metals, such as iron and copper (Richards and Carpenter 2012).
All these issues combined make it extremely difficult to determine any differences in corrosion rates for
the five planes. However, because all aluminum alloys are corroding in a common oxidizing marine
environment in Tanapag Lagoon, the different values of the corrosion potentials may provide a guide to
the underlying differences in alloy composition of the aircraft. Based on the data collected, the metal
composition of the aluminum alloys for each aircraft, in order of decreasing concentrations of
incorporated copper (or other less reactive metals) is Avenger > Jake > Mariner ~ Coronado > Emily. That
is, the Avenger may have the highest concentration of copper in this group of aluminum alloys
measured while the EMILY will have the lowest. This has consequences for the corrosion rates of these
aircraft as higher concentrations of copper increases the rate of pitting and intergranular corrosion if the
aircraft are subjected to similar environmental conditions and other complicating factors, such as
increases in corrosion through stress and metal fatigue, are absent. Unfortunately, this is not the case
with these aircraft (i.e. the Avenger lies in a very aggressive, shallower marine environment and the
Coronado is extensively damaged with separate sections strewn over a very large area) highlighting the
problem with interpreting corrosion data based on only one set of corrosion parameter measurements
(Richards and Carpenter 2012).
It is obvious that there are problems with determining differences in corrosion behavior of wrecks based
on only one set of data measurements. Only through continued observation and collection of corrosion
measurements will it be possible to tease out the subtle differences in deterioration caused by local
environmental conditions versus the more obvious impacts resulting from human activity. This type of
information can inform decisions on which sites to actively discourage visitation versus sites that are
better able to withstand the bumps and bangs caused by divers. It can also inform decisions on where to
place warning markers for boat traffic and how to educate visitors to the fragile nature of these sites.
51
In Situ and Ex Situ Artifacts
Material recovered from WWII sites includes objects that were removed at the time of wrecking or postbattle during government sanctioned salvage and harbor-clearing works. However, artifacts removed by
private persons since WWII, if associated with US military craft of any type, are legally the property of
the US government.
Legislation aimed at protecting and preserving cultural heritage has a long history in the U.S. beginning
with the Antiquities Act of 1906, the Historic Sites Act of 1935, the Archaeological and Historic
Preservation Act of 1974, Archaeological Resources Protection Act of 1979, the Abandoned Shipwreck
Act of 1987, and the Sunken Military Craft Act of 2004, among others. Internationally the 1970 UNESCO
Convention on the Means of Prohibiting and Preventing the Illicit Import, Export and Transfer of
Ownership of Cultural Property, Law of the Sea Convention of 1982, and the UNESCO Convention on the
Protection of Underwater Cultural Heritage of 2001 each seek to protect cultural heritage from damage,
looting, and treasure salvage.
Under the property clause of the US Constitution, the 1982 United Nations Convention on the Law of the
Sea (Articles 95 and 96) and established principles of international maritime law and sovereign
immunity, the U.S. Department of the Navy retains custody in perpetuity of its ships and aircraft. These
laws state the right, title, and ownership of federal property is not lost to the US government due to the
passage of time, or by neglect or inaction.
This also applies to the remains of Japanese military craft. Under the same international laws, the
Japanese government retains control and ownership of those items unless they were specifically and
formally disposed of. In the case of WWII wrecks, the Treaty of Peace with Japan, signed 8 September
1951, provides in Chapter V, Article 14(a)2(I) that each of the Allied Powers “shall have the right to seize,
retain, liquidate or otherwise dispose of all property, rights and interests” of Japan, “which on the first
coming into force of the present Treaty were subject to its [the Allied Powers] jurisdiction” (DOS 1951).
This effectively gave ownership and control of Japanese sites to the US government. Because these
properties are not considered “abandoned” in the Abandoned Shipwreck Act of 1987 (43 U.S. C. 21012106) they did not transfer to the states with adoption of the Act.
It is difficult to determine the amount of artifact removal or looting that has taken place since the end of
WWII. Because WWII is within the recent past, cultural remains and artifacts associated with the war
were not generally recognized by the public as having historical significance. Many objects connected
with these sites were routinely collected by local divers as well as by visitors. Only with the passage of
time are these sites now seen as an important part of our shared heritage and worthy of protection.
On Saipan a great deal of material removed from sites in the past is held in personal, undocumented
collections on the island, and some may have left the island. There is no effort by the US government to
reclaim these objects nor is there an effort by the CNMI to do so. Rather the emphasis is on insuring
that the sites and objects that remain are protected and preserved for future generations. A limited
52
number of artifacts are on display at the American Memorial Park, however it is not known if any of
these are from underwater contexts. No material was recovered during the recent archeological site
inspections.
53
54
Chapter 5: Public Outreach
Introduction
This project is the next phase of a multi-year effort to raise awareness on the importance of protecting
WWII UCH on Saipan. Phase one began in 2009 with a grant from the ABPP GA-2255-09-028. Public
outreach and community collaboration efforts began with that grant and have continued uninterrupted
during this project.
By holding both formal and informal meetings with government agencies, tourism offices, dive shops,
volunteers, the public and stakeholders with a variety of perspectives made their needs known. In
response to stakeholder input, training in basic site documentation and maritime heritage was provided
in 2010. Stakeholder input also prompted the development an underwater heritage trail, full color
posters and dive guides in two languages—Japanese and English—reflecting the two major user groups
of these resources. Outside the confines of Saipan a website and project blog were created to reach
beyond the local tourism base and educate the broader public about this unique collection of WWII UCH
sites.
Because underwater trail, posters, dive guides, websites and blogs reach only a fraction of the public, a
17-minute interpretive film WWII Maritime Heritage Trail: Battle of Saipan, focusing on selected sites,
was developed. This interpretive film is shown at the NPS’s American Memorial Park visitor center.
Consultation, Public Meetings, Presentations, Press and Digital Media
Consultation
Beginning in 2009, three agencies on the island have played a crucial role in determining the success of
project efforts: the Historic Preservation Office (HPO), the Department of Environmental Quality (DEQ),
and the Coastal Resources Management Office (CRM). Initial meetings and consultations were held with
these agencies to assess their cooperation, interest, and involvement with the archaeological
documentation and underwater heritage trail projects (McKinnon and Carrell 2011). All were again
consulted for the current project and their direct involvements encouraged.
Because HPO is the regulatory agency that deals with heritage, we have worked hard to maintain an
open line of communication through regular email and phone contact. Staff was involved with the
conservation study, more fully described elsewhere and provided as Appendix A, and facilitated
communication with other agencies and the general public. CRM and DEQ staff met with us to provide
useful information on the efficacy of the ongoing mooring buoy project at selected sites on the
underwater heritage trail and to discuss preliminary results of the conservation study.
The Marianas Visitor Authority Office (MVA), a government-funded office focused on the development
of tourism, was contacted for consultation. By providing demographic information on tourist divers and
snorkelers, it quickly became apparent that interpretive products should be produced in Korean,
Russian, Japanese, Chamorro and Carolinian in addition to the planned English. They provided assistance
55
with dissemination of the posters in 2011 to locals and tourists and important public feedback that were
applied to the interpretive film.
The project team also held meetings with small groups including non-profit organizations. Two that
were particularly helpful are the Northern Marianas Council for the Humanities, a non-profit supported
by government funding, and the Pacific Marine Resources Institute, a non-profit with interests in
traditional Micronesian fishing. Vital support from these organizations was provided in the form of local
information and contacts with smaller user groups including diving and fishing organizations.
Consultation with diving and fishing groups included visits to local dive shops to receive feedback on
needs at the local level. Two important groups, Marianas Dive and Mariana Sports Club, Inc., were
consulted. Several members of the Marianas Dive group participated in training held in 2010 and
members of the Mariana Sports Club, Inc. provided valuable historical information and insight into the
history of the wrecks.
The NPS American Memorial Park on Saipan and War in the Pacific National Historical Park staff and
Superintendent were key consultants. They provided important insights into the needs of visitors and
questions often asked. They reviewed the posters and trail guides for accuracy and provided guidance
on the interpretive film. Because the film was designed from the outset to be shown in the park visitor
centers, their input was particularly important.
The collaboration of the NPS Submerged Resources Center (SRC) was crucial in bringing the film to
completion. As the only team of maritime archaeologists and filmmakers focusing on UCH they
participated directly in the filming and editing phases. Windward Media, our video producers, went to
Denver to meet and review the film with the SRC. The finished product is a testament to the close
collaboration between SRC and Windward Media.
Information, participation, and collaboration from all of these organizations materially informed the
development of the underwater heritage trail, the posters, dive guides and ultimately the interpretive
film.
Public Meetings and Presentations
Several public meetings and presentations were given during this multi-year project. The first public
meeting was held on Saipan in June 2010 at the American Memorial Park and included an audience of
over 100 people. The presentation consisted of a 45-minute presentation on the archeological and
historical research, the concept of the underwater heritage trail and the benefits to the community as
well as a question and answer session. The meeting was sponsored by the Council for the Humanities
and the PowerPoint slide was posted on their website for public viewing. A public meeting was held in
April 2011 at the American Memorial Park and included an audience of 57 people. The presentation was
an hour long and presented the final results of the trail including drafts of the underwater guides and
posters. The meeting was sponsored by the Asia Pacific Academy of Science, Education, and
Environmental Management. Following this meeting there was a question and answer period.
56
More than 20 presentations have been given at professional societies and organizations, including the
following:
2014
2013
2013
2012
2012
2012
2011
2011
2011
2011
2011
2011
2011
2011
2011
2011
2010
2010
Site Formation Processes of Sunken Aircraft: A Case Study of Four WWII Aircraft In Saipan’s
Tanapag Lagoon. Paper presented at the Society for Historical Archaeology Conference. Quebec,
Canada.
A Sea Story of Fluidity in the Mariana Islands. Paper presented at Sea Stories: Maritime
Landscapes, Cultures and Histories Conference. Sydney, New South Wales.
Community Archaeology Approaches in the Commonwealth of the Northern Mariana Islands.
Paper presented at the Society for American Archaeology Meeting. Honolulu, Hawaii.
Before 3D, there was 2D; Collaborative efforts in creating an interpretive film for underwater
heritage in Saipan, CNMI. Seminar given at the Department of Archaeology Seminar Series,
Flinders University. Adelaide, South Australia.
The Economic Benefits of Protecting Underwater Cultural Heritage. Invited paper at the 2012
UNESCO Asia-Pacific Regional Meeting for the Protection of Underwater Cultural Heritage. Koh
Kong, Cambodia.
Heritage that Hurts: Interpreting Battlefield Sites in Maritime Archaeology. Paper presented at
the Society for Historical Archaeology Conference. Baltimore, Maryland.
Saipan’s Underwater Heritage. Presentation given to the Rotary Club. Kissimmee, Florida.
Interpreting Underwater Battlefield Sites for the Public – Inclusion, Negotiation and
Communication. Invited paper presented at the Pacific War: 1941-45 Heritage, Legacies, and
Culture Conference. Melbourne, Victoria.
Inclusion and negotiation: Interpreting underwater battlefield sites for the public. Paper
presented at the Asia-Pacific Regional Conference on Underwater Cultural Heritage. Manila,
Philippines.
The Potential for research on Spanish cultural heritage in the Commonwealth of the Northern
Mariana Islands. Paper presented at the Asia-Pacific Regional Conference on Underwater
Cultural Heritage. Manila, Philippines.
Recording the Indigenous Maritime Cultural Landscape and Seascape in Saipan. Poster
presented at the Asia-Pacific Regional Conference on Underwater Cultural Heritage. Manila,
Philippines.
Management and engagement: using maritime heritage trails to interpret and protect
submerged WWII heritage from the Battle of Saipan. Paper presented at IKUWA4. Croatia.
The WWII maritime heritage trail – Battle of Saipan project: lessons learned. Paper presented at
the Australasian Institute of Maritime Archaeology Conference. Brisbane, Queensland.
Fair winds and following seas: community (maritime) archaeology. Seminar given to the Flinders
Institute for Research in the Humanities. Adelaide, South Australia.
Recent underwater archaeological research and “discoveries” in the CNMI. Presentation given to
the Asia Pacific Academy of Science, Education and Environmental Management. Saipan, CNMI.
A WWII underwater heritage trail: developing an underwater program in Saipan, CNMI. Paper
presented at the Society for Historical Archaeology Conference. Austin, Texas.
The task of reinterpreting: using maritime heritage trails to interpret submerged WWII heritage
from the Battle of Saipan. Paper presented at the Maritime Heritage Conference. Baltimore,
Maryland.
Saipan’s underwater heritage; developing an underwater WWII heritage trail. Presentation
sponsored by the Northern Mariana Islands Council for the Humanities. Saipan, CNMI.
57
2010
2009
From training to tourism; developing a WWII maritime heritage trail in Saipan. Seminar given at
the Department of Anthropology Brown Bag Series, Florida State University. Tallahassee,
Florida.
Recent archaeological investigations in the Commonwealth of the Northern Mariana Islands.
Seminar given at the Department of Archaeology Seminar Series, Flinders University. Adelaide,
South Australia.
Press and Digital Media
A vital link to the community on Saipan is through written, radio and television press. A number of press
announcements were released locally from 2009 to the present.
NMI’s First Public 3D Presentation (Marianas Variety May 2013)
Underwater Heritage Trail 3D Documentary available in June (Marianas Variety May 2013)
NMI to Have First 3D Public Presentation (Marianas Variety April 2013)
Radio Interview on Interpretive Film and Conservation research project (KKMP (1440 AM
and 92.1 FM March 2013)
3D Film on Underwater Heritage Trail to Premier (Marianas Variety, 22 January 2013)
Choose your own adventure: Saipan’s New WWII Maritime Heritage Trail Battle of Saipan is
now open for business (Mariana’s Variety, 21 September 2011)
Radio Interview with “Your Humanities Half-Hour” (Power 99FM, sponsored by Humanities
Council, April 2011)
Television Interview on Underwater Heritage Trail products (KSPN Channel 2 Sport Program,
April 2011)
Television Interview on Underwater Heritage Trail (John Gonzales Live Show, KSPN Channel2, April 2011) (presented in English and Chamorro)
Heritage Tourism Tipped as CNMI Money Spinner (Radio New Zealand International, April
2011)
Commonwealth Can Develop Heritage Tourism (Saipan Tribune, April 2011)
Commonwealth Should Organize "Heritage Tours" (Pacific News Center, April 2011)
Heritage Awareness Seminar Today (Marianas Variety, April 2011)
Archaeologist Discusses Prospects for NMI Heritage Tourism (Marianas Variety, April 2011)
Lecture to preview underwater WWII heritage trail (Marianas Variety, 23 June 2010)
Underwater heritage trail in the works for Saipan lagoon (Saipan Tribune, 30 June 2010)
58
Television interview with two archeology students on project (KSPN Channel 2, July 2010)
Television interview with J. McKinnon on project (KSPN Channel 2, February 2010)
Rusting Relics Still Have Tales to Tell (Flinders Journal, September 2009)
NMI should tap heritage tourism potential (Marianas Business Journal, 3-16 August 2009)
Study of Saipan war relics planned (Saipan Tribune, 30 June 2009)
SDE to develop underwater heritage trail in CNMI (Marianas Variety, 23 June 2009)
$49970K grant to fund underwater mapping of Saipan lagoon (Saipan Tribune, 30 June
2009)
Website
A website, http://www.pacificmaritimeheritagetrail.com/ , was created using WordPress to
information about the trail, photographs and a location where the dive guides and posters can be
downloaded.
Facebook
A WWII Maritime Heritage Trail: Battle of Saipan Facebook group was created for the trail,
https://www.facebook.com/#!/groups/120863607992582?ap=1 . The idea behind creating a
group was mentioned by a local diver who wished to have a space to post photographs and his
experiences of diving on sites somewhere. The group is open access which means any person can
view and join the group. It includes copies of the dive guides and posters and has already
generated a good amount of interest. As of March 2014 there are 114 members in the group. This
group will be maintained by Ships of Discovery Facebook members and local divers who have
volunteered to be administrators.
Wikipedia
A Wikipedia page, http://en.wikipedia.org/wiki/Maritime Heritage Trail - Battle of Saipan, has
been created with basic information about the heritage trail. Because Wikipedia is an open source
public space, the general public can edit the entry and include factual information about the trail
and the heritage sites.
Interpretive Materials and Film
Two types of public outreach products were created as part of a larger plan to aid in the preservation
and protection of UCH sites that were already being impacted through visitation: printed “hand out”
materials that visitors could take with them or download from the internet and print and an interpretive
film. The printed materials, a poster and underwater dive guide, were produced under grant GA-225509-028. The film is the public outreach component of this grant.
Poster and Dive Guide
With the establishment of the WWII underwater heritage trail in Saipan (more fully described in
McKinnon and Carrell 2011) we hoped it would solidify the concept that these resources were more
59
than just a collection of random “left-overs” from a time and event that have a direct connection only to
the oldest of Saipan’s residents. Today’s post-WWII generations have grown up seeing remnants of
ships, barges, landing craft, and tanks poking above the water from the time they were children. Their
very commonness made them easy to ignore and their significance easy to overlook.
Anecdotal evidence from other areas where similar trails exist suggests that the development and
promotion of underwater heritage trails helps to foster an appreciation for local heritage. Saipan’s
situation is different from other areas that have underwater trails in one important aspect; the vast
majority of users are tourists not locals. The interaction between tourism service providers (locals) and
the users (tourists) meant that in order to engender an appreciation in the tourists we needed to
educate the service providers.
After much public consultation it was clear that posters and re-usable dive guides were the preferred
products. The combination meant that both the non-diving and diving public were targeted. The design
of the posters, large format, color, and double-sided outlined each site’s history and importance, legal
protection and proper etiquette for visiting (Figure 30, and Figure 31). Each was designed to be
attractive and include quality photographs. For the diving public, water-proof laminated site guides
(Figure 32) including a brief description and a drawing to identify key features were produced in both
English and Japanese.
A total of 750 posters and 500 waterproof dive guides of nine sites were printed distributed. The
National Park Service American Memorial visitor center was the primary distribution point for the
posters along with the Mariana’s Visitor Bureau, HPO and the Humanities Council. The dive guides were
assembled into 55 sets and distributed to 10 of dive shops.
The limitations of this approach are obvious. Once all the posters have been given away, once the all the
dive guides are “used up,” their impact and effect on behavior rapidly diminishes. However, all artwork
for the guides and posters were given to several agencies including HPO, MVA, NPS and Humanities
Council so that they may be printed. Additionally, they were posted for download on the trail website.
60
Figure 30. Front of Japanese shipwrecks poster.
61
Figure 31. Back of Japanese shipwrecks poster.
62
Figure 32. Aichi 313A dive guide.
The Film - WWII Maritime Heritage Trail: Battle of Saipan
Why a film? It is an accepted truism that film can consistently reach more people over time than an
inherently finite supply of books, pamphlets, posters, and static museum exhibits. Depending upon
where a film is shown, for example in the United Kingdom, films about archaeology shown on TV
regularly receive 3-5 million viewers per airing (Clack 2006:87). In comparison, the British Museum had
slightly fewer than 5 .5 million visitors in all of 2005 (Clack 2006:87). The implication is clear: in a single
hour a film can reach nearly as many people as the largest heritage museums in the world can reach in
an entire year. The power of film to reach a diverse audience and to reinforce its message is evident. It
also has the advantage of being unlimited in its use, reuse, and venue. With the rise of YouTube, it can
now be streamed worldwide. Film can do all of this at a fraction of the cost of print media per target
audience member. So the question isn’t “why a film?” but “why not a film?”
The 17-minute interpretive film produced for this project had two objectives: to educate the local
population and tourists about UCH and to encourage its appreciation and preservation. The film was
designed to take diving and non-diving visitors on a tour of the WWII Underwater Heritage Trail
developed under the 2009 ABPP Grant. It tells the story of the Battle of Saipan through the underwater
heritage sites that are scattered across Saipan’s seabed including aircraft, tanks, landing vehicles, and
63
ships. Special care was taken to include all those involved and affected by the Battle and the larger war
in general. The video also embraces a strong conservation and preservation message.
Western styles of cultural preservation tend to focus on tangible heritage while Micronesians display a
preference for non-tangible heritage (e.g. traditional skills and knowledge) (O’Neill and Spennemann
2001:46). According to Spennemann, some have argued that WWII remains are left to deteriorate by the
Indigenous Micronesian populations because they do not care about them or at least were not
concerned about them in the past. But why should they care about them? With few exceptions, the
Pacific islanders did not actively choose to be involved in the War. According to Spennemann (1992:15),
It happened around them; it happened against them. Their islands were bombed and
burned; their gardens burned by napalm or destroyed by tanks plowing through them; their
villages shelled by naval vessels and canoes sunk by aircraft; the islanders themselves were
commandeered for forced labor, experienced food shortages and starvation.
For the descendants of those who lost their lives on Saipan during WWII there is a built-in affinity for the
locations where these events occurred, but not necessarily the “debris” resulting from the events. The
challenge of the film was to transcend these single views into a shared story and a shared history.
As the standard of living has increased in developing countries, visitors from Russia, Japan, Korea, and
China have increased exponentially. Special tours arranged by organizations with ties in those countries
bring in busloads of tourists. Many of these groups include the National Park Service American Memorial
park visitor center on their list of stops. There are also tour organizations that cater to the diving public
and they too bring their customers to the visitor center. Visitors by far have the greatest impact on the
underwater sites, and visitors, whether they remain on land or venture into the sea, will take home the
story that the sites represent.
A film that provides accurate information to these target audiences and does so in a respectful manner
can take the single perspective and turn it into one that is shared. It can encourage “ownership” across
time, distance and cultures. In an effort to broaden and deepen the impact of the film it is also subtitled
in Japanese, who are the majority of visitors. By using the NPS American Memorial Park visitor center as
a primary point of distribution the park staff can customize the visitor experience depending on the
origin of the group. There is also a version that is subtitled in English for the hearing impaired.
A secondary point of distribution is the school system. This may prove to be the best and longest-lived
impact and where the concept of preserving sites that exemplify a shared history will be the most
influential. Teachers can use it as one element in a suite of tools to teach the next generations about
their history and their stories.
Finally, we live in a digital age where content is available 24-7. Nearly everyone expects that anything
can be accessed with the touch of a few keys. The third point of distribution is YouTube, which has
become the de facto source for video uploads and searches. To access a global audience Ships of
64
Discovery set up a dedicated YouTube channel (http://www.youtube.com/user/ShipsOfDiscovery) to
showcase this film, among others.
In addition, copies of the film were provided to the following:
Office of the Governor
Historic Preservation Office
Coastal Resources Management
Mariana’s Visitor Authority
Northern Marianas Council for the Humanities
Saipan Chamber of Commerce
Pacific Development, Inc.
War in the Pacific National Park (Guam)
Prior to finalizing the film, it was shown to a number of different target audiences: university students,
family members, friends, resource managers, park staff, tour operators, and members of other ethnic
backgrounds. From the professional audience the most often received comment was that the
conservation/preservation message was clear and not over-bearing and the overall tone and feeling of
the film was positive and respectful. From non-professional friends, family and others, comments (and
an audible sigh) occurred when damage through vandalism was mentioned and shown. They generally
expressed a deep concern that this would lead to the destruction of the site and a loss to future
generations. This feeling was the same whether the individual was a diver or non-diver. It did not matter
whether the individual ever anticipated visiting the site. From park staff the response was very positive
and only recommended adding Chinese subtitles. Subtitling in Korean and Russian were also suggested.
Under the current grant, the funds to accomplish those translations and film editing were unavailable.
The true test of the success or failure of the film will be the response of the visitors who see it at the
American Memorial Park visitor center. National Parks have a long history of providing intelligent,
informative, unbiased information to educate and enlighten the public. It is hoped that this film will
support that mission and lead to a greater sense of shared ownership for these and other WWII heritage
sites. Only time (and visitor feedback) will tell.
WWII in the Pacific, a momentous event in world history from a Western and Eastern perspective, is
simply a brief interlude from the Indigenous Pacific islanders’ point of view (Spennemann 1992:15). This
perspective, though pragmatic, has the potential to hinder the effective preservation of non-Indigenous
heritage resources. The rise of global tourism and more particularly eco- and heritage-tourism is forcing
a change in outlook. For a small country such as the CNMI, tourism has arguably become its most
important economic driver. Heritage managers in the CNMI face numerous challenges in balancing site
protection with public interpretation. The fragility, vulnerability, and significance of the resources make
the sites more susceptible to visitor impacts. A film can educate and inspire, but it cannot take the place
of committed stakeholders and government agencies to take on the hard tasks of management and
protection.
65
66
Chapter 6: Recommendations and Relevant Issues
O’Neill and Spennemann (2001:46) argue that efficacious preservation of cultural resources is
dependent upon several factors: political will, community interest, and availability of resources. Saipan
struggles in each of these areas. As Saipan’s economy continues to weaken the impact on the agencies
that are charged with managing, protecting, and interpreting UCH and the environment has had their
budgets and personnel reduced. The HPO has been without a Director since 2010 and a qualified staff
archaeologist since 2011. This presents particular challenges in the development of both community and
agency action planning and implementation.
An upsurge in the CNMI’s heritage tourism industry will undoubtedly stimulate the local economy,
attracting visitors and drawing money to the islands. However, the connection between economic gain
and heritage preservation is a precarious one in that heritage sites are vulnerable resources that may be
harmed by tourism activities. According to Carrell, “developing tourism operations of the CNMI could
cause heavy visitation to these sites by scuba divers. There is already a commercial tour submarine on
Saipan that offers tours of some underwater sites. There have been reports of this tour submarine
damaging some of the sites” (1991:335). This warning, given nearly 25 years ago, simply reinforces the
difficulty in creating a viable environment for heritage protection with limited resources.
A framework for managing UCH that both promotes and protects Saipan’s submerged heritage is timely
and necessary. Each issue was identified based on discussions with managing agencies and the dive
community tempered with knowledge of the sites, their historical and archeological context, the
environmental and cultural impacts affecting the sites, and the social, economic and political conditions
of Saipan. Because of overlapping responsibilities, limited personnel, and lack of an overarching agency
with both the authority and resources to push forward a comprehensive management strategy, these
can only be recommendations.
The recommendations fall into four broad categories: policies and procedures, programmatic, site
specific and public outreach. Included within policies and procedures are legislative initiatives, capacitysharing and strategic planning and inter-agency cooperative agreements. Programmatic
recommendations focus on those areas that are mandated by various legislative requirements. Site
specific recommendations include direct and indirect site monitoring, while public outreach is selfexplanatory. Each recommendation is followed by a discussion of underlying issues that constrain or
impact implementation, and then an action item with a proposed time frame. Table 2 at the end of this
section is provided as a quick reference summary to the recommendations, actions and time frames.
The time frames are all dependent upon having adequate staffing at the HPO, including hiring a qualified
archaeologist as soon as possible. The recommendations require the leadership of HPO staff to initiate
consultation and strategic planning as a foundation for action and implementation. Lack of adequate
staff, and a qualified archaeologist, at the HPO is the biggest obstacle to the long-term preservation of
UCH on Saipan. The current team is simply not able to take on this long-term effort with their current
level of staffing.
67
Management of Underwater Cultural Heritage Sites
Recommendation. Ensure that the HPO has adequate qualified and trained staff to manage cultural
heritage sites.
Issue: Lack of Staff. An obstacle in the CNMI is that there are no effective means to protect UCH
because of a lack of trained staff in all of the managing agencies. This is particularly problematic at the
HPO. As of this writing there is only one archeological technician at the HPO certified to dive and that
has undergone training in underwater archeology (having taken the Flinders University training in
underwater archeology in 2009). This makes it impossible to manage, monitor, assess and enforce
legislation. The archaeologist position at the HPO has been vacant since 2011.
This status quo is contradictory to Part II, Guideline 5 of the Abandoned Shipwreck Act 1987 that states
“The agencies responsible for the management of State-Owned waters should have (or have access to)
adequate professional staff, office and laboratory facilities, vessels, diving and underwater survey
equipment to carry out assigned responsibilities.” As a result, the community currently relies on offisland resources and expertise to assist them with the management of their submerged archeological
heritage through projects such as this.
Action. Hire a qualified archeologist and archeological technicians (a minimum of three), and existing
staff be trained in SCUBA and the ability to conduct archeological assessments underwater.
Time Frame. As soon as possible
Consultation and Strategic Planning
Recommendation. Consultation and strategic planning among the key government bodies in the
management of submerged heritage is encouraged. The fundamental principles of site protection as
stipulated in the National Historic Preservation Act of 1966, the Archaeological Resources Protection Act
of 1979, the Abandoned Shipwreck Act of 1987, the Sunken Military Craft Act of 2005, and the CNMI
Historic Preservation Act of 1982.
Issue: Cooperation. The sites discussed in this report are all located within CNMI waters and many, but
not all, are located within the Mañagaha Marine Conservation Area. This means that there are
overlapping jurisdictions and responsibilities in some areas and gaps in management and capabilities in
others.
The HPO has the overall administrative responsibility for cultural heritage in the CNMI. This obligation
extends off shore to all UCH, whether they have been identified or are as yet to be investigated.
However, the HPO does not have any on-the-water capacity to monitor sites or enforce violations.
To accomplish the legislative purpose of the Mañagaha Marine Conservation Act, the CNMI Department
of Lands and Natural Resources (DLNR) was delegated the exclusive authority to manage the Mañagaha
Marine Conservation Area (MMCA), as well as other marine conservation areas in the CNMI (Section 6 of
PL 12-12). The role of the DLNR in regard to the management of UCH is to promote public access to the
68
sites while protecting the physical remains. The 2005 management plan for the conservation area states
that DLNR has a role in providing logistic and financial support for placing moorings on sites to reduce
impacts by anchoring. They have on-the-water capability to monitor sites and the ability to enforce laws
and cite violators. But they do not have anyone with a cultural heritage background.
Coastal Resources Management (CRM) was established on 11 February 1983, with the implementation
of Public Law 3-47 within the Office of the Governor. The CRM program was established in order to
promote the conservation and wise development of coastal resources. CRM is responsible for general
permitting activities that impact coastal resources in Saipan and in particular permits for dive boats and
dive tour operations. They have on-the-water capability to monitor sites and enforce violations outside
the MMCA. Similarly, the Department of Environmental Quality (DEQ) mainly concerned with water
quality and pollution and the US Fish and Wildlife Service (US FWS) all have on-the-water capabilities
and enforcement obligations. But none of these agencies have staff with a cultural heritage background.
In order to coordinate their natural resources efforts these agencies formed the Marine Monitoring
Team (MMT) that is charged with providing statistically sound and relevant scientific information
necessary for the management of reef and fish resources. It is comprised of marine biologists and
environmental technicians that collect information on coral species diversity, colony populations,
benthic percent cover, and fish and macro-invertebrate numbers.
Action. Establish an HPO-DNLR-CRM-DEQ-FWS working group (Working Group) to discuss this
management plan, to organize effective means to monitor and enforce protection of UCH through
capacity-sharing, and to address specific threats to both natural and cultural resources. Where
necessary, create additional language for policies and procedures in each agency that support the
protection of submerged heritage.
This could be a UCH-specific Working Group that follows the MMT model. However, given limited staffs
and time, the HPO is urged to partner with the MMT.
Time Frame. 2014 Join the MMT/form a UCH Working Group to lay the framework for capacity sharing
Time Frame. 2014-2015 Identify additional language in policies and procedures
Legislation and Effective Enforcement
Recommendation. Review CNMI legislative mandates to bring them up to international standards.
Ensure effective protection under existing legislation. WWII sites are currently protected under the
National Historic Preservation Act of 1966, Archaeological Resources Protection Act of 1979, Sunken
Military Craft Act of 2005, and the CNMI Historic Preservation Act of 1982.
Issue: Legislation. Presently the CNMI does not use the Annex Rules contained in the UNESCO
Protection on the Protection of the Underwater Cultural Heritage 2001 as a basis for best practice to
conduct under water archaeological investigations or UCH management. Good management requires
balancing legislation and capabilities. This is not a call for harsher restrictions but an approach that
applies existing legislation most efficiently by layering methods of protection and enforcement.
69
Action. Review CNMI’s legislation as it relates to the protection of UCH and update as necessary. The
Annex Rules contained in the UNESCO Protection on the Protection of the Underwater Cultural Heritage
2001 should be used as a model in formulating this process. This falls under the purview of the HPO, but
will necessitate the full cooperation of the CNMI legislature and other relevant agencies and
departments. Ideally, the CNMI would endorse the UNESCO Convention and adopt the Annex Rules as
best practice.
Time Frame. 2018-2019
Action. Working Group identifies means to layer protection using existing legislation and agency
mandates to manage cultural and natural resources. This could involve something as simple as
recognizing the location of the sites within the Conservation Area or National Landmark and applying
that legislative framework.
Time Frame. 2015-2016
Issue: Effective Enforcement. Because there is only one scuba-certified staff member at the HPO who is
currently able to participate in site inspections and the HPO does not have law enforcement training,
there is no effective enforcement to ensure protection of the submerged sites.
Action. Partner with and enlist other agency staff whose job description already includes enforcement
of natural or environmental legislation. The MMT, which includes the CRM, DLNR, DFW, and DEQ, have
in place the boat assets and, with heritage training, the staff that could include visits to UCH sites and
incorporate them in their routine biodiversity and monitoring studies.
Time Frame. 2015
Action. Seek funding for UCH training for MMT members
Time Frame. 2016
Action. Working Group establishes intra-agency agreements in which HPO staff accompany other
agency enforcement officers during inspections. Section 6 of Public Law 12-12 states that “...the
Department [of Lands and Natural Resources] may coordinate and assist other Commonwealth or
Federal agencies in performing their emergency or other agency functions within marine conservation
areas, if the exercise of such functions is deemed prudent or necessary by the Department, or the
performance of such functions is clearly permitted by law within marine conservation areas.” There is a
precedent set in law that allows for collaborative efforts in management of marine resources.
Time Frame. 2015-2016
Complete National Register Nominations
Recommendation. The nomination of significant sites to the National Register of Historic Places is
essential in identifying and demonstrating their importance to the local, national and international
community; this programmatic requirement should be met as soon as possible
70
Issue: Register nominations. No National Historic Register Nominations exist for submerged WWII sites
in CNMI.
Action. HPO undertake nominations or seek grant funding for preparation of National Register
nominations of the submerged WWII sites. Information from Carrell 2009 and McKinnon and Carrell
2011 can be used in the development of nominations. At a minimum the sites included on the heritage
trail should be nominated.
Time Frame. 2016-2017
Site Database and At-Risk Artifacts
Recommendation. HPO should develop and maintain a submerged sites database that includes a means
to record and inventory at-risk artifacts removed from sites or donated to the CNMI. HPO develop a
form for documenting submerged sites.
Issue: Database and recording form. The HPO does not have a submerged sites database or a site
recording form specific to recording submerged sites. By creating a GIS database and form the agency
will be able to better manage existing sites and update databases with new sites and current
information regarding the condition of sites. Further a database will assist with assessments in the event
of development applications.
Action. HPO create a GIS database and form for recording submerged sites.
Time Frame. 2016
Issue: Artifacts. Since the post-battle period of government sanctioned salvage, no archeological project
has raised artifacts from WWII submerged sites. In recent surveys of the wrecks (see McKinnon and
Carrell 2011) a variety of personal items and moveable items were noted. These are vulnerable to the
environment and looting, and should be assessed and removed from the sites if risks are posed.
Action. At minimum photographic documentation of moveable artifacts and inclusion in a database
should be completed. This could help track items if removed from the sites. Enlisting the help of the
previously trained cadre of divers to photograph these items is a means to encourage stakeholder
ownership of the sites and a means to monitor ongoing movement on the sites or removal.
Time Frame: 2015-2016
Action. A comprehensive evaluation of sites that have not been heavily visited or still have moveable
items should be undertaken by HPO in conjunction with Naval Historical Center staff. If artifacts are
removed, it is recommended that professionals properly conserve them. The Naval Historical Center has
indicated their support for this process.
Time Frame: 2016-2017
71
Programmatic Research and Inventory
Recommendation. HPO in conjunction with other agencies, organizations, and stakeholders continue
programmatic investigation of submerged sites through historical research, survey, site identification,
site documentation, and the collection of oral histories. This research is necessary for comprehensive
management of UCH.
Issue: Incomplete Information on WWII UCH. Additional historical research into the location,
identification and details of WWII submerged wrecks is crucial to completing the history of the Battle of
Saipan. For known sites, the research should be specific and geared toward answering questions of
identification and circumstances of loss. For sites yet to be identified, research should be more general
to obtain information on post-invasion cleanup, cold war demolition, and recent salvage, channel
clearing and dredging.
Action. HPO seek grant funds to continue programmatic historical research and identification of UCH.
Time Frame. 2017 and beyond
Issue: Further survey inside the lagoon. A total of 1,543 potential archeological targets were identified
during the SEARCH, Inc. 2008 remote sensing surveys of Saipan’s western lagoon. Only a small portion of
those targets have been tested and identified. There is a need to conduct further research on these
anomalies to determine whether they are UCH.
Action. HPO seek grant funds to continue programmatic identification and evaluation of previously
located submerged resources. HPO can also partner with and seek cooperation of U.S. Navy, U.S. NPS,
Japanese government, CRM, DEQ, universities, recreational divers and community groups to undertake
this work.
Time Frame. 2016 and beyond
Issue: Further survey in potential areas. A number of WWII submerged sites have been reported that
fall outside of the western lagoons and that are in need of baseline remote sensing survey and
preliminary investigation. There are a few known dumps around the island where U.S. forces simply
threw equipment off cliffs into the water. One in particular is just outside of the northern edge of the
lagoon and said to be a collection of at least five LVTs. Another known dump is located just north of
Bonsai Cliff below a concrete pad on the cliff edge. These and other known or reported dumps should be
investigated and documented archeologically.
Lau Lau Bay on the east side of Saipan was a significant area to WWII operations. Although it was not an
invasion beach, Japanese forces used the bay as a deep-water anchorage. The bay is 731m (2,400 ft) at
the deepest point (PBEC 1984:S3). It is also the suspected location of a B-29 crash site. This area has not
been archeologically surveyed.
72
An area that is likely to have scattered remains is the deep water off the western edges of the fringing
coral reef. The 2008 SEARCH, Inc. survey extended just beyond the reef and identified a number of
anomalies. This is an area where large naval vessels or private vessels commissioned by the Navy
currently anchor and it is also the U.S. staging area for the invasion. The area has great historical
significance and has not been surveyed. Because of its ongoing use as an anchorage, any potential sites
are currently under threat. Vessels in deeper water have a better chance of being less impacted by
cultural and natural factors and therefore could be significant in terms of what they have to offer
archeologically.
Action. HPO seeks copies of any previously collected relevant bathymetric mapping and survey data for
Lau Lau Bay to determine if further survey with remote sensing equipment would be useful. It is possible
that bathymetric mapping and survey has been conducted by other government agencies that could
lead to identification of UCH sites.
Time Frame. 2015
Action. HPO seek grant funding and develop partnerships to undertake programmatic inventory through
remote sensing surveys of areas outside the western lagoon, Lau Lau Bay, and in areas where other
significant or threatened sites may be located.
Time Frame. 2017 and beyond
Issue: Oral histories. Oral histories can provide a more nuanced understanding of historical events
particularly in cultures with traditions in oral histories. Individual experiences are often not found in
government documents and the only access to those is through the collection of oral histories and
review of diaries and personal letters or memoirs. Oral histories can provide a local narrative and
understanding of the WWII wrecks that is undocumented. Thus collection of oral histories should be
considered in order to record personal accounts and narratives of the battle and understand local values
with regard to submerged WWII sites.
Action. HPO seek grant funding and partner with relevant CNMI agencies, organizations, community
and regional stakeholders to collect oral histories.
Time Frame. 2017-2018
Monitor Material Remains to Identify Natural Impacts
Recommendation. Implement a long-term monitoring and conservation program to identify natural
impacts to UCH and develop of a strategic plan to mitigate adverse impacts.
Issue: Corrosion Surveys. The submerged shipwrecks and aircraft wrecks located in Saipan are a
significant part of WWII history and are one of the main tourist attractions in Saipan. It is important that
an appropriate monitoring and conservation plan is implemented to ensure the future preservation of
these sites. The program should include regular monitoring and data collection to determine the status
of the natural and cultural features on the wrecks. This is best carried out using a systematic approach
to data collection combined with visual inspections. A baseline corrosion study was completed as part of
73
this project and the full report is provided in Appendix A. The optimal information to be gathered under
a systematic data collection program is outlined in the On-Site Corrosion Survey Data Sheet included in
that report.
Action. The Working Group identifies which agency has the capacity to conduct regular site inspections
at the identified sites. The most important aspect of the regular site inspections is photographic
documentation of any changes that occur. This will allow meaningful comparisons to be made in the
future to ascertain if any significant changes to a particular site have occurred. These could be carried
out as part of other duties to monitor the health of natural or marine resources and thereby save money
and time. The report provided in Appendix A includes the locations of each of the corrosion data
collection points. These locations should be photographed and the photographs labeled and filed with
appropriate members of the Working Group.
Time Frame. 2014-2015
Action. The Working Group establishes an inspection schedule that would allow for two visits to each
site per year. Additional inspections are urged following any severe storm or cyclonic activity so any
changes in the integrity of the site are noted by direct comparison with earlier surveys.
Time Frame. 2014-2015
Action. The Working Group seeks funding for another full corrosion and environmental survey in 3-5
years. In this way, from comparisons of the regular site inspection results and the additional corrosion
parameter data for each wreck site, it will be possible to ascertain if there is indeed any effect from
diving tourism on the sites and if it is at all comparable to the detrimental effects afforded by natural
occurrences, such as seasonal storm and cyclonic activity. Finally, using a combination of information
gathered from these surveys it will be possible to prioritize these submerged sites with respect to their
overall in situ management requirements and the most appropriate management plans determined and
applied to each site.
Time Frame. 2016-2017
Recognize UCH as Sites of Natural Significance
Recommendation. Undertake additional research on the marine environment and associated marine
life at each of the sites investigated.
Issue: WWII wrecks as sites of natural significance. A wreck can create a unique local environment for
fauna and flora to thrive, and this has a bearing on defining the site's significance and issues of research
and interpretation. Biological research was conducted in conjunction with the development of the
heritage trail (2009-2010) by a researcher from Sydney University. Fowler conducted two years of fish
assemblage studies on the sites included on the trail. His research determined that “well- established
vessel-reefs are capable of approximating fish abundances and assemblage parameters on natural coral
reefs” (Fowler and Booth 2012).
This research indicates that WWII sites are of natural and environmental significance and should be
74
protected as such. There are no provisions within Public Law 12-12, which regulates the Mañagaha
Marine Conservation Area, to identify submerged sites as aquatic reserves or equivalent. However
access to the sites, if they are environmentally sensitive, maybe be controlled under provisions of the
Mañagaha Marine Conservation Area management plan.
Action. Working Group with MMT develops a long term research program to gather more biological
information at all of the WWII sites identified to date to 1) assess their value as sites of natural resource
significance, 2) monitor the impacts of visitor use on the marine biota, and 3) make recommendations
for visitor use. The results of this study would better inform the management of the UCH, add to the
corpus of information on these sites, and support the mandate of the MMT.
Time Frame. 2015-2016-2017
Protect Material Remains from Cultural Threats
Recommendation. All WWII shipwrecks, aircraft wrecks and vehicles underwater are protected by the
National Historic Preservation Act of 1966, Archaeological Resources Protection Act of 1979, Sunken
Military Craft Act of 2005, and the CNMI Historic Preservation Act of 1982. Under this legislation it is
illegal to interfere with, damage or remove an historic site or related items. Raising awareness and
community outreach using a variety of media and partnerships is strongly urged.
Issue: Salvage and looting. Many WWII archeological sites have been partially salvaged during
sanctioned post-battle operations (e.g. Japanese merchant ship, Possible auxiliary submarine chaser)
More relevant to management issues is the question of whether some sites were salvaged in the recent
past after sanctioned government salvage took place. Particularly the iron shipwrecks are vulnerable
from those that may salvage for scrap metal value. No records or instances have come to light
concerning this type of activity. However, there is knowledge of souvenir hunting or looting and artifact
movement that has occurred regularly on these sites (see Chapter 4). All features on all sites are
arguably vulnerable from this threat. Those features under greatest threat, because of their individual
appeal, are personal objects or small, moveable items such as bullets or serial number plates and
recognizable features of the machinery such as handles and gauges. Though not much has been located
in terms of personal objects, these may still be buried on site.
Action. The HPO and/or Working Group develops partnerships with the MVA, National Park Service
American Memorial Park, tourism service providers, dive shops and organizations, and schools to spread
the word about heritage protection. A heritage preservation awareness day, poster contest for school
children, and radio interviews are all viable options. This message can incorporate protection of natural
heritage; protecting one often protects the other.
Time Frame. 2015-2016
Action. The HPO and/or Working Group enlist a community leader or community groups to develop an
education program that raises people's awareness of the significance of historic wrecks. A public
outreach program that reviews protective legislation as well as communicating the historical, cultural
75
and environmental significance of the wrecks would be most effective and should be aimed at all age
groups including both local and tourist populations.
Time Frame. 2015-2016
Issue: Accidental interference while diving. Damage could be caused by divers who are unaware of
appropriate wreck diving practices. Divers could handle, move or accidentally damage artifacts because
they do not know that interference is illegal under the law. Divers may also accidentally touch fragile
material with fins, tanks or their bodies. Therefore divers need to be made aware of the appropriate
diving practices expected when visiting sites. This includes a policy of “look but do not touch” and a
request that divers pay attention to their buoyancy, so as to not accidentally damage material. This is
the same message given with regards to natural resources and so it should be easy to relate this concept
to cultural resources.
Action. Because this issue impacts both natural and cultural resources, the Working Group should
partner with the community and relevant stakeholders to develop a brochure or educational outreach
program aimed at dive operators/shops that outlines the importance of promoting appropriate behavior
and buoyancy on wrecks (See Figure 33 for example).
Time Frame. 2016-2017
76
Figure 33. Diving Shipwreck brochure produced by South Australia Heritage.
Issue: Anchoring on the site. Visitors drop their anchors onto or drag their anchors across the remains of
wrecks to moor over them. Anchors dragged across or dropped onto sites cause damage to the remains.
Therefore anchoring on site is interpreted as interference and damage to an historic wreck, which is
illegal under the multiple laws that protect these sites. In addition to anchoring, “tying off” to the
exposed remains of a wreck also causes interference and damage and is therefore illegal. The
destructive effect of anchoring or mooring directly onto a wreck has been documented on several sites
in Saipan.
Action. Because this issue impacts both natural and cultural resources, the Working Group should
partner with the community and stakeholders to expand the existing mooring system outside the MMA
to include the more heavily visited or fragile sites.
Time Frame. 2016-2017
Action. The Working Group hold public meetings to increase education for boaters and captains of
charter vessels about the damages they cause and how their actions are illegal, potentially through a
brochure or sticker required on their vessel (See Figure 34 for example).
Time Frame. 2015-2016
77
Action. HPO coordinate with the agency responsible for issuing boat permits and ask they include
language in permits or licenses about the laws regarding anchoring and mooring on historic sites to
prevent future disturbance.
Time Frame. 2016-2017
78
Figure 34. Anchoring brochure produced by South Australia Heritage Branch.
Issue: Boat maneuvering. Some wrecks are near the surface and break the water at low tides. A boat
being maneuvered around the site may accidentally collide with the remains and cause major structural
damage. Therefore a boat collision with the remains of any historic wreck would constitute interference
and damage, both acts illegal under the current legislation.
Action. The Working Group addresses this issue in public meetings and enlists the help of the
community and stakeholders to exercise all possible caution while maneuvering around sites or in
transit. Moorings or markers would also contribute to raising awareness about their location.
Time Frame: 2015-2016
Action. HPO to work with relevant agency to insure that all sites on the heritage trail as well as any
known sites that pose a hazard to navigation (e.g. LVT2) included on NOAA charts.
Time Frame: 2016-2017
Issue: Vandalism. Vandalism has occurred on some sites and may be the result of intentional or
unintentional behavior. However, any action, intentional or unintentional, that damages or interferes
with a site is illegal. Scratching names into the metal fabric of a site causes the metal to enter into an
active state of corrosion until such time that it can reach equilibrium again. Further, as has been
demonstrated in the in situ corrosion survey, many of these sites are still actively corroding or in some
case all of the metallic fabric has been lost. This creates a situation in which the vessels are fragile and
79
should be handled with care. Thus climbing on the sites or holding on to the sites could damage or
interfere with the site, which is illegal under current legislation.
Action: The Working Group holds public meetings and partners with stakeholders to educate users.
Vandalism also impacts natural resources so this is an excellent opportunity to promote an antivandalism message. Partner with educators and stakeholders to develop and promote educational
outreach.
Time Frame: 2015-2016 and beyond
Issue: Interference. Tour operators and resorts undertake or promote activities that impact the wrecks
negatively. Some sites are too closely visited by small boats, Jet Skis or submarines and are on a regular
track or course, which creates a wake or disturbance of the immediate water and the wreck itself.
Through the creation of a wake or movement of parts of the hull of a wreck, changes are made to the
environment and the structure, which increases corrosion due to oxidization or disturbance of the metal
fabric. This disturbance and interference damages the sites and is illegal.
Action: The Working Group in cooperation with stakeholders create “no disturbance/no wake” zones
around sites that are vulnerable, particularly the Sherman tanks. Regulate the tracks or courses that
boats, Jet Skis or submarines make around or nearby sites through the permits required. For example,
change the Jet Ski course for those sites nearby or require the submarine to make its turn away rather
than towards the Japanese merchant ship so that its thrusters do not affect the wreck. Partner with
stakeholders to establish protocols and alternatives.
Time Frame. 2016-2017
Monitor Impacts of Development
Recommendation. The HPO, the Department of Public Lands (DPL) and DEQ work to insure that future
development will not adversely impact UCH.
Issue: Development. The DPL is responsible for all lands in the CNMI, including underwater, and is the
permitting agency for all land use. Lessees currently include the Hyatt, Shimizu Corporation, Mobil Oil,
Shell Oil, and Pacific Telecom Inc. Such leases fuel a significant part of the local economy. There are at
present no known plans for development in the vicinity of the WWII wrecks in Saipan’s lagoon. However
this does not preclude development from affecting the sites in the future. There are sites located within
the intertidal zone on the beaches that could be affected by coastal erosion caused by development or
upgrading of beaches for recreation. Additionally, those sites that are located nearby or within the
existing working navigational channel may be impacted by future dredging plans or vessel traffic. Thus
knowledge of where the sites are and what development plans are projected is vital to protecting these
resources.
Action. Develop a mechanism to involve the HPO in all lease permits that include submerged or
shoreline lands. This type of cooperation already exists for on-shore lands regarding the potential for
human remains. This should be built upon to include WWII UCH.
80
Time Frame: 2015-2016
Action. The Working Group includes in its inspections of sites for natural and/or cultural impacts a
checklist to identify impacts attributable to development.
Time Frame: 2016
Monitor Visitation to Sites
Recommendation. Visitation to WWII sites included on the maritime heritage trail should be
encouraged and visitor numbers tracked to provide information on potential impacts.
Issue: Visitor Numbers. The majority of visitors to the underwater WWII sites are divers and the vast
majority of those do so on chartered vessels or as part of a dive tour. To date, there are no data
collected and provided to the HPO that give visitation numbers to UCH sites. Because the numbers of
users has a direct bearing on the potential for adverse impacts, particularly on fragile sites, this type of
information is needed to inform long term management. Dive charter operations are registered and
typically they require that divers complete liability paperwork, so visitor use can be tracked.
Action. The HPO and CRM in cooperation with stakeholders update the reporting system to include
visitor numbers to UCH sites. This should not be a separate report, but simply an expansion of existing
CRM requirements to report visits to natural areas. In all discussions with stakeholders, it must be
emphasized that this not a means of controlling who visits, but only a non-obtrusive method of
monitoring site visitation. The goal is to insure long term preservation of the sites for visitor use. The
benefits of this approach are:
• HPO staff will be able to gather information without having to visit the sites,
• it is an opportunity for the HPO to distribute interpretation and site access literature and for
visitors to access other forms of interpretation on the wrecks,
• registration, as a means of collecting visitation information, is an effective yet non-intrusive
method, and
• a formal registration process will likely positively affect the behavior of visitors when on the
sites.
Time Frame: 2015-2016
Strengthen Relationship with the Dive Community
Recommendation: The HPO develops a stronger public presence with the dive community to increase
awareness about the importance of protecting submerged heritage.
Issue: Relationship with dive community. Traditionally there has been little communication between
local divers and CNMI HPO. Local divers report being frustrated when trying to report new sites to HPO
and receive information about their history or significance. This disconnect is a result of HPO’s inability
through training, equipment and funding to be involved in the management of submerged sites.
81
Action. HPO should partner with stakeholders and local divers, dive groups and dives shops in the
location and identification of new sites. Those who know the waters best are those who use it regularly.
Identify a community based leader or point person outside the HPO who can keep up momentum. There
is already a small cadre of previously trained divers (McKinnon and Carrell 2011) who could form the
basis for this relationship. Education programs, meetings, information sessions and the development of
a site-reporting program are all options for increasing communication.
Time Frame: 2016 and beyond
Increase Availability of Interpretive Materials
Recommendation. Educational outreach is the best method to guarantee appreciation and long term
preservation of all cultural heritage sites. The dissemination of information about WWII wrecks should
be widespread and cater to the general public, as well as tourists. Interpretation should promote
awareness of the wrecks’ significance and of the need to preserve them.
Issue: Interpretive materials. Dive guides and posters were produced on the maritime heritage trail in
2010-11. They include site plans, information about the history of the vessels, and site locations. They
also state that the sites are protected and what this means in regards to site access. Distribution of this
material began in early 2011 and all materials have been distributed. An interpretive film was produced
in 2012 that takes the diver and non-diver on a virtual tour of the WWII sites. The film is shown at the
American Memorial Park visitor center, but is not readily available elsewhere.
Action. The HPO should partner with the Marianas Visitors Authority and other stakeholders as
appropriate to raise funds to print more dive guides and/or posters. The posters could be sold in a
variety of venues including dive shops and tourist centers. The sales price to the venues should be
sufficient to replenish the funds to print more. Dive shops are one logical venue for resale to their
customers. The original production files are not copyright protected and are readily available for this
use.
Time Frame. 2015-2016 and beyond
Action. The HPO should partner with the Mariana’s Visitor Authority and other stakeholders to raise
funds to duplicate the existing World War II Maritime Heritage Trail: Battle of Saipan interpretive film.
The film could be sold in a variety of venues including dive shops and tourist centers. The sales price to
the venues should be sufficient to replenish the funds to duplicate more copies. The film is not copyright
protected are readily available for this purpose.
Time Frame. 2015-2016 and beyond
Action. The HPO should partner with the Mariana’s Visitor Authority and other stakeholders to raise
funds to obtain Chinese, Korean, and/or Russian language translations and re-editing of the film to
include those language subtitles. This would increase visitor outreach and public impact.
Time Frame. 2017-2018
82
Issue: Display. The American Memorial Park (AMME) has a WWII display available to visitors to the Park
but does not include any information on the underwater sites.
Action. Work with AMME to provide supplemental information on the WWII UCH of Saipan.
Information could be taken directly from the interpretive material, which has already been produced.
Photographs and site plans are not copyright protected and are readily available for this purpose.
Time Frame: 2017-2018
Table 2. Quick reference to recommendations, action items and time frame
Recommendation
Action
Time Frame
HPO hire a qualified archaeologist and
additional staff to manage cultural
heritage sites on land and underwater
ASAP
Ensure adequate professional staffing
at HPO
Consultation and Strategic Planning
Policies and procedures Create Working Group to lay
framework for capacity sharing
Policies and procedures Identify and create language in existing
polices/procedures to support
protection of submerged sites
Review CNMI legislation and ensure
effective protection and enforcement
Policies and procedures Bring CNMI legislation up to
international standards
Policies and procedures Identify means to use existing
legislation to layer protection
Policies and procedures Hire qualified archeologist and
archeological technicians
Policies and procedures Partner with and enlist other agency
staff to help with enforcement and
protection
Policies and procedures Create intra-agency agreements to
accompany other agency officers
during routine inspections of natural
resources
Complete National Register
Nominations
Programmatic HPO seek funding or complete
nominations in house
Develop and maintain submerged Site
database
Programmatic Create GIS database and form for
submerged sites and artifacts
Site specific Photographic documentation of
83
2014
2014-2015
2018-2019
2015-2016
ASAP
2015
2015-2016
2016-2017
2016
2015-2016
Recommendation
Action
Time Frame
artifacts at risk
Site specific Re-evaluation of sites with at risk
artifacts
Continue programmatic investigation
of submerged sites
Programmatic HPO seek grant funds to continue
historical research to identify
submerged sites
Programmatic HPO seek grant funds to continue
identification and evaluation of
previously located submerged sites
Programmatic HPO seek grant funds to survey other
areas for as yet to be located
submerged sites
Programmatic HPO seek copies of prior bathymetric
mapping and survey for evaluation
Programmatic HPO seek grant funding to collect oral
histories
Implement long-term monitoring and
conservation to identify natural
impacts to sites
Policies and procedures Working Group identifies CNMI agency
to conduct regular site inspections
Policies and procedures Working Group establishes inspection
schedule
Programmatic HPO and/or Working Group seeks
grant funding for second full corrosion
study
Conduct additional natural resources
research on each site
Site activity Working Group assess value of sites
for natural resources significance
Site activity Monitor impacts of visitor use on
marine biota
Protect sites from cultural threats
Public outreach HPO/Working Group develop
stakeholder partnerships to conduct
public outreach
Public outreach HPO/Working Group enlist a
community leader to develop public
education program
Public outreach HPO/Working Group with stakeholders
develop a public outreach program
aimed at divers to protect sites from
visitor use damage
Site activity HPO/Working Group with stakeholders
84
2016-2017
2017 and beyond
2016 and beyond
2017 and beyond
2015
2017-2018
2014-2015
2014-2015
2016-2017
2015-2016
2016 and beyond
2015-2016
2015-2016
2016-2017
2016-2017
Recommendation
Action
Time Frame
to expand existing mooring system
Public outreach Working Group and stakeholders hold
public meetings to increase education
for boaters
Policies and Procedures HPO work with agency responsible for
boating permits to add language in
permits about anchoring and mooring
on historic sites
Public outreach Working Group with stakeholders hold
public meetings to educate boaters
about mooring around submerged
sites
Policies and Procedures HPO ensure that any sites that pose a
hazard to navigation are included on
NOAA charts
Public Outreach HPO/Working Group hold public
meetings to educate visitors about
vandalism
Site activity Working Group with stakeholders
establish “no wake” zones around
vulnerable sites
Protect sites from impacts of future
development
Site activity Working Group includes checklist to
identify impacts to sites attributable to
development
Policies and procedures Develop mechanism to involve HPO in
lease permits that include submerged
or shoreline lands
Monitor visitation to sites
Policies and procedures HPO/CRM with stakeholders update
existing reporting system to include
visitor numbers at submerged sites
Strengthen relationship with dive
community
Site activity HPO partner with stakeholders and
local dive community to help locate
and identify new sites
Increase availability of interpretive
materials
Public outreach HPO partner with MVA to duplicate
and distribute film as a sale item in
dive shops and tourist centers
Public outreach HPO partner with MVA and
stakeholders to reprint dive guides and
posters
85
2015-2016
2016-2017
2015-2016
2016-2017
2015-2016
2016-2017
2016
2015-2016
2015-2016
2016 and beyond
2015-2016 and
beyond
2015-2016 and
beyond
Recommendation
Action
Time Frame
Public outreach HPO partner with MVA to raise funds
to add Chinese, Korean and/or Russian
subtitles to interpretive film
Public outreach HPO work with AMME to provide
supplemental information about
submerged sites
86
2017-2018
2017-2018
Chapter 7: Conclusion
The Mariana Islands and the Pacific region in general have their own unique set of challenges and issues
that are quite different from other parts of the world. It is important, therefore, that other “models” are
not applied wholesale to this region. A method that considers the challenges and difficulties the
Marianas is facing and a program that is suited to deal with these challenges and resolve them into the
future is the most likely successful long term approach. As Anita Smith stated in Contested Heritages in
the Pacific Islands (nd), “Communities and governments in the region are keen to engage with
international conservation programs not only because they are interested in protecting their heritage
and resources but also as they provide a source of income, training and avenue for communication with
the global community. The challenge is for processes of heritage protection and national legislation to
govern and enforce this protection to be based in and evolve from traditional systems of governance
and cultural practices rather than imposed from the outside.”
The recommendations in this report were written with the intention of assisting the CNMI HPO and
other interested agencies in managing their submerged WWII heritage. The information has the
potential to empower the local community with knowledge about their sites and how they might initiate
further investigation and research into such sites and protect them from further degradation through
cultural and natural factors.
While this report specifically focuses on WWII submerged heritage, there are many more sites from
previous and later time periods that are significant to the local history. It is hoped that this report will
serve as a blueprint to guide the HPO in managing and preserving all sites. Inclusive and collaborative
efforts should be made with the regulatory agencies that deal with the marine environment, the
community and stakeholders. The interests of the U.S. and Japanese government are at the core of this
heritage, but a local perspective that incorporates Chamorro, Carolinian, Filipino and Korean values
should be actively pursued.
The report includes a range of recommendations and actions. All are dependent on a number of issues
local to the island including: staff, training, equipment, funding, cooperation, and priorities. Additional
assistance with corrosion surveys using specialized equipment may be needed in the future, although
basic corrosion survey using cameras can be conducted by the HPO or its partners using methods
outlined in Appendix A. Nevertheless, when corrosion or archeological surveys are needed, there are
several partners identified within this document who might assist with these needs including agencies,
museums and universities. Additionally, it might be useful to seek the assistance of other Micronesian
HPOs who have trained staff, including Guam that has been actively researching and surveying its
submerged heritage.
A community-centered stakeholder-invested approach to public outreach is particularly important. The
film offers an opportunity to reach a wide audience with a positive message of preservation and
appreciation that goes beyond one group or ethnic identity. It can be used as a springboard to invite and
encourage community involvement. The old adage applies here-- many hands make the work light.
87
Sharing the work and spreading the benefits among the people of Saipan will enrich all of us, whether
we live there or not, whether we will ever see the WWII underwater heritage sites first hand or only
through vicarious means.
These sites may represent the stories of WWII on Saipan but they are our shared history.
88
Further Contact
The authors and researchers of this report would be pleased to answer any questions that might arise
from its content and would welcome any feedback or suggestions. Should there be a need for additional
training or survey, the authors would be happy to assist. The contact details of each author are listed
below in case of further questions or contact.
Jennifer McKinnon, Ships of Discovery and East Carolina University, Program in Maritime Studies,
Admiral Eller House, Room 103, Greenville, NC 27858-4353, +252-328-6788, mckinnonje@ecu.edu
Toni Carrell, Ships of Discovery, 39 Condesa Road, Santa Fe, NM 87508, +505-466-2240,
tlcarrell@shipsofdiscovery.org
Vicki Richards, Western Australian Museum, 45-47 Cliff St., Fremantle, WA 6160 AUSTRALIA, +61 (8)
9431 8472, Vicki.richards@museum.wa.gov.au.
Jon Carpenter, Western Australian Museum, 45-47 Cliff St., Fremantle, WA 6160 AUSTRALIA, +61 (8)
9431 8472, Jon.Carpenter@museum.wa.gov.au.
89
90
References Cited
Amesbury, Judith, Darlene Moore and Rosalind Hunter-Anderson
1996 Cultural Adaptations and Late Holocene Sea Level Changes in the Marianas: Recent Excavations at
Chalan Piao, Saipan, Micronesia. In Indo-Pacific Prehistory: The Chiang Mai Papers, edited by Ian C.
Glover and Peter Bellwood, pp. 53-69. Volume 2. Australian National University, Canberra, ACT.
Avenger
n.d. Naval Aircraft. Electronic document, http://www.history.navy.mil/branches/hist-ac/tbm-3s.pdf,
accessed 4 March 2011.
Bulgrin, Lon E.
2005 The Tudela Site: Fire and Steel Over Saipan, 15 June 1944. The Journal of Conflict Archaeology 1:117.
Carrell, Toni L. (editor)
1991 Micronesia Submerged Cultural Resources Assessment. National Park Service, Santa Fe, New
Mexico.
2009 Maritime History and Archaeology of the Commonwealth of the Northern Mariana Islands.
Submitted to the CNMI Department of Community and Cultural Affairs, Division of Historic Preservation.
Ships of Exploration and Discovery Research, Inc., New Mexico.
Clack, T.
2006 Living Archaeology: Conflict and Media. In Remapping the Field: New Approaches in Conflict
Archaeology, pp. 87-93. John Schofield, Axel Kausmeier and Louise Purbrick editors. WestkreuzDruckerei Ahrens KG, Berlin.
Crowl, Philip A.
1960 US Army in World War II: The War in the Pacific: Campaign in the Marianas. Office of the Chief of
Military History, Department of the Army, Washington, DC.
Denfield, D. Colt
1992 Japanese World War II Fortifications and Other Military Structures in the Central Pacific. Northern
Mariana Islands Division of Historic Preservation, Saipan. Division of Historic Preservation,
Commonwealth of the Northern Mariana Islands.
Department of Community and Cultural Affairs, Division of Historic Preservation
2004 Pacific Preservation: The Commonwealth of the Northern Mariana Islands 2004-2008 State Historic
Preservation Plan. Department of Community and Cultural Affairs, Saipan.
Denton, G. R. W., Concepcion, L. P., Galt Siegrist H., Vann, D.T., Wood, H.R., and Beardon, B.G.
2001 Contaminant Assessment of Surface Sediments from Tanapag Lagoon, Saipan. Unpublished
Technical Report No. 93, University of Guam, Water and Environmental Research Institute of the
Western Pacific, Guam.
91
Fowler, A.M. and Booth, D.J.
2012 How well do sunken vessels approximate fish assemblages on coral reefs? Conservation
implications of vessel-reef deployments. Marine Biology 159:2787-2796.
Grove, E.
1976 World War II Tanks. 2nd ed. Excalibur Books, New York, NY.
Lodge, O.R.
1954 The Recapture of Guam. Reprinted 1998. Awani Press, Fredericksburg, Texas.
Lord, Christine and Colin Plank
2003 Surveys of Abandoned Vessels: Guam and the Commonwealth of the Northern Mariana Islands.
Submitted to the National Oceanic and Atmospheric Administration, National Ocean Service, Office of
Response and Restoration by Research Planning, Inc. Columbia, South Carolina.
M4 Sherman Tank
2010 M4 Sherman Tank. www.2worldwar2.com/sherman.htm, accessed September 2010.
MacLeod, Ian D., and Vicki Richards
2011 In situ conservation surveys of iron shipwrecks in Chuuk Lagoon and the impact of human
intervention, AICCM Bulletin, Vol. 32:106-122, Australian Institute for the Conservation of Cultural
Material, Sydney.
McKinnon, Jennifer F. and Toni L. Carrell
2011 Saipan WWII Invasion Beaches Underwater Heritage Trail. Ships of Exploration and Discovery
Research, Corpus Christi, Texas.
Miculka, James E., Carter, Kevin J., and Nobuo Ichihara
1984 Submerged Cultural Resources Assessment Saipan Lagoon 03 – 05 December 1984. War in the
Pacific National Historical Park, Guam.
Miculka, James E., Wood, Edward, Cooper, William, and Dennis Blackenbaker
1990 Submerged Cultural Resources Assessment: Saipan. War in the Pacific National Historical Park,
Guam.
Moore, David
2002 The Battle of Saipan—The Final Curtain. Electronic document, http://www.battle
ofsaipan.com/seabee.htm, accessed October 29, 2008.
O’Neill, J.G and D.H.R. Spennemann
2001 German Colonial Heritage in Micronesia. Cultural Resource Management 24(1):46-47.
Pacific Basin Environmental Consultants (PBEC)
1984 Underwater Survey of Tanapag Lagoon for Historic Properties. Submitted to the Community and
Cultural Affairs, Historic Preservation Office by Pacific Basin Environmental Consultants, Guam.
92
1984 Biological and Physical Survey of Bahia LauLau, Saipan Final Report. Prepared for Commonwealth
of the Northern Mariana Islands, Planning/Energy Office under a Grant from the U.S. Department of
Commerce, Office of Coastal Zone Management, Coastal Energy Impact Program.
Richards, Vicki and Jonathan Carpenter
2012 Conservation Survey and Management Program: Saipan WWII Underwater Archaeological Wreck
Sites. Unpublished report on file Ships of Exploration and Discovery Research, Inc., Santa Fe, NM.
Rogers, R. F.
1995 Destiny’s Landfall: A History of Guam. University of Hawaii Press, Honolulu, Hawaii.
Rottman, G. L.
2004a Saipan and Tinian 1944: Piercing the Japanese Empire. Osprey Publishing, Oxford, United
Kingdom.
2004b US World War II Amphibious Tactics; Army and Marine Core, Pacific Theater. Osprey Publishing,
Oxford, UK.
Russell, S.
1984 From Arabwal to Ashes: A Brief History of Garapan Village; 1818 to 1945. Micronesian
Archaeological Survey Report Number 19, Department of Education, Commonwealth of the Northern
Mariana Islands.
SEARCH, Inc.
2008a Archaeological Survey of Lagoons of on the West Coast of Saipan, Commonwealth of the Northern
Mariana Islands. Submitted to the CNMI Department of Community and Cultural Affairs, Division of
Historic Preservation. Southeastern Archaeological Research, Inc., FL.
2008b Archaeological Survey of Tanapag Lagoon Saipan, Commonwealth of the Northern Mariana
Islands. Submitted to the CNMI Department of Community and Cultural Affairs, Division of Historic
Preservation. Southeastern Archaeological Research, Inc., FL.
Smith, A.
n.d. Contested Heritages in the Pacific Islands. Accessed online February 2012.
Spennemann , D.H.R.
1992 Apocalypse Now? The Fate of World War II Sites on Central Pacific Islands. Cultural Resource
Management Bulletin 15(2):14-16.
U.S. Department of State
1951 Treaty of Peace with Japan, signed September 8, 1951
93
94
Appendix A
CONSERVATION SURVEY AND
MANAGEMENT PROGRAM
SAIPAN WW II UNDERWATER
ARCHAEOLOGICAL WRECK SITES
Vicki Richards and Jonathan Carpenter
2012
Appendix A
CONSERVATION SURVEY AND MANAGEMENT PROGRAM
SAIPAN WW II UNDERWATER ARCHAEOLOGICAL WRECK SITES
Prepared for
Ships of Exploration and Discovery Research Inc.
Grant Agreement No. GA-2255-11-018
By
Vicki Richards and Jonathan Carpenter
2012
The contents, opinions, conclusions and recommendations expressed in this report are those of the authors and do not
necessarily reflect the views or policies of the Department of the Interior.
Front Cover: Sherman Tank, Garapan Lagoon, Saipan, CNMI (Carpenter 2012)
This Page: Auxillary Submarine Chaser, Tanapag Lagoon, Saipan, CNMI (Carpenter 2012)
Appendix A
CONTENTS
ACKNOWLEDGEMENTS .................................................... i
LIST OF FIGURES .............................................................. ii
LIST OF TABLES................................................................ v
LIST OF APPENDICES..................................................... vii
EXECUTIVE SUMMARY ................................................................. 1
1
INTRODUCTION .................................................................... 3
1.1
1.2
BACKGROUND ....................................................................................... 3
SCOPE OF THE WORK .......................................................................... 3
2
METHODOLOGY ................................................................... 4
3
CONSERVATION ASSESSMENTS – IRON
ALLOY WRECKS .................................................................. 5
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
M4 SHERMAN TANK 1 – TANK 1 .......................................................... 5
M4 SHERMAN TANK 3 – TANK 3 ........................................................ 11
LANDING VEHICLE TRACKED 1 – LVT1 ........................................... 16
LANDING VEHICLE TRACKED 2 – LVT2 ........................................... 22
DAIHATSU LANDING CRAFT 1 – DAI1 ............................................... 26
DAIHATSU LANDING CRAFT 2 – DAI2 ............................................... 32
DAIHATSU LANDING CRAFT 3 – DAI3 ............................................... 36
JAPANESE FREIGHTER – JFR .......................................................... 41
AUXILIARY SUBMARINE CHASER – ASC ........................................ 46
STEAMSHIP – SS ................................................................................ 51
4
CONSERVATION ASSESSMENTS – ALUMINIUM ALLOY
AIRCRAFT WRECKS ................................................................... 56
4.1
4.2
4.3
4.4
4.5
GRUMMAN TBM AVENGER – AVR ..................................................... 56
AICHI E13A – JAKE .............................................................................. 62
MARTIN PBM MARINER – MNR .......................................................... 67
KAWANISHI H8K – EMILY ................................................................... 72
CONSOLIDATED PB2Y CORONADO – CRDO ................................... 77
Appendix A
5
CONCLUSIONS ................................................................... 82
6
RECOMMENDATIONS ........................................................ 85
7
REFERENCES ..................................................................... 88
8
APPENDICES ...................................................................... 89
2
Appendix A
ACKNOWLEDGEMENTS
The authors wish to sincerely thank Jennifer McKinnon and Toni Carrell, Ships of Exploration and
Discovery Research Inc, for providing this unique opportunity to work with some extremely dedicated
and highly professional individuals in an unequivocally special part of the world. Special thanks must
go to the “A-Team” - Jason Raupp, Peter Harvey, Cosmos Coroneos, John “John John” D. San
Nicolas, Jennifer McKinnon and Toni Carrell for their expert assistance in carrying out the corrosion
surveys of 15 wreck sites in less than four days. Thanks must also go to the captains and crew of the
dive boat, who made the diving an absolute pleasure. Thanks must also go to Genevieve Cabrera who
shared with us her extensive knowledge of the history and cultural heritage of Saipan. Finally we
would like to thank all of the people that have supported this project over the years in so many ways,
whether we met you personally or not and hope that our small contribution will help, in some small
way, to better preserve the UCH of Saipan and benefit the local community as a whole.
i
Appendix A
LIST OF FIGURES
Figure 1. M4 Sherman Tank 1 (Tank 1) – port side view (Carpenter 2012) ............................ 5
Figure 2. Change in dissolved oxygen content with increasing water depth measured
on the M4 Sherman Tank 1 site .................................................................................................. 6
Figure 3. Location of the M4 Sherman Tanks, Saipan, CNMI (Richards 2012 after Google Earth
2012) 7
-6
Figure 4. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the intercepts
of the areas measured on the M4 Sherman Tank 1 .................................................................. 8
Figure 5. Schematic plan and profile views of the M4 Sherman Tank 1 indicating the corrosion
parameter measurement positions (Richards 2012 after Hanks 2010 in
McKinnon and Carrell 2011:107) ................................................................................................ 9
Figure 6. M4 Sherman Tank 3 (Tank 3) – stern view (Carpenter 2012) ................................. 11
Figure 7. Change in dissolved oxygen content with increasing water depth measured
on the M4 Sherman Tank 3 site (Richards 2012) .................................................................... 12
Figure 8. Schematic plan and profile views of the M4 Sherman Tank 3 indicating the corrosion
parameter measurement positions (Richards 2012 after Hanks 2010) ................................ 14
-6
Figure 9. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the intercepts
of the areas measured on the M4 Sherman Tank 3 (Richards 2012) .................................... 15
Figure 10. Landing Vehicle Tracked 1 (LVT1) – port side view (Carpenter 2012) ............... 16
Figure 11. Change in dissolved oxygen content with increasing water depth measured
on the LVT1 site (Richards 2012) ............................................................................................. 17
Figure 12. Location of the LVT(A)-4 or LVT1 in Tanapag Harbour, Saipan, CNMI
(Richards 2012 after Google Earth 2012) ................................................................................ 18
Figure 13. Schematic plan and profile views of the LVT1 indicating the corrosion
parameter measurement positions (Richards 2012 after Arnold 2010 in McKinnon and Carrell
2011:118) .................................................................................................................................... 19
-6
Figure 14. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the LVT1 (Richards 2012) .......................................... 20
Figure 15. Landing Vehicle Tracked 2 (LVT2) – front view (Carpenter 2012) ...................... 22
Figure 16. Image indicating the corrosion parameter measurement positions 1, 2 and 5
on the LVT2 (Richards after Carpenter 2012) ......................................................................... 24
Figure 17. Image indicating the corrosion parameter measurement positions 3 and 4 on
the LVT2 (Richards after Carpenter 2012) ............................................................................... 24
-6
Figure 18. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the LVT2 (Richards 2012) .......................................... 25
Figure 19. Daihatsu Landing Craft (DAI1) – stern to bow view (Carpenter 2012) ............... 26
ii
Appendix A
Figure 20. Change in dissolved oxygen content with increasing water depth measured
on the DAI1 and DAI2 sites (Richards 2012) ........................................................................... 27
Figure 21. Location of the Landing Vehicle Tracked 1 (LVT1 or LVT(A)-4), Daihatsu
Landing Craft 1 and 2 (DAI1 and DAI2), Japanese Freighter (Freighter) and the possible
Auxiliary Submarine Chaser (Sub Chaser) in Tanapag Harbour, Saipan, CNMI (Richards 2012
after Google Earth 2012) ........................................................................................................... 29
Figure 22. Schematic plan view of the DAI1 indicating the corrosion parameter measurement
positions (Richards 2012 after McAllister and Yamafume 2011 in
McKinnon and Carrell 2011:94) ................................................................................................ 30
-6
Figure 23. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the DAI1 (Richards 2012) ........................................... 30
Figure 24. Daihatsu Landing Craft 2 (DAI2) – port side view (Carpenter 2012)................... 32
Figure 25. Image indicating the corrosion parameter measurement positions on the
DAI2 (Richards after Carpenter 2012) ...................................................................................... 33
-6
Figure 26. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the DAI2 (Richards 2012) ........................................... 34
Figure 27. Daihatsu Landing Craft 3 (DAI3) – starboard side view (Carpenter 2012) ......... 36
Figure 28. Change in dissolved oxygen content with increasing water depth measured
on the DAI3 site (Richards 2012) .............................................................................................. 37
Figure 29. Image indicating the corrosion parameter measurement positions on the
DAI3 (Richards after Carpenter 2012) ...................................................................................... 39
-6
Figure 30. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the DAI3 (Richards 2012) ........................................... 39
Figure 31. Japanese Freighter (JFR) – bow view (Carpenter 2012) ...................................... 41
Figure 32. Change in dissolved oxygen content with increasing water depth measured
on the JFR site (Richards 2012) ............................................................................................... 42
-6
Figure 33. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the JFR site (Richards 2012) ..................................... 44
Figure 34. Auxiliary Submarine Chaser (ASC) – bow view (Carpenter 2012) ...................... 46
Figure 35. Change in dissolved oxygen content with increasing water depth measured
on the ASC site (Richards 2012) .............................................................................................. 47
Figure 36. Photomosaic indicating the corrosion parameter measurement positions
on the ASC (Richards after McKinnon and Carrell 2011:49) ................................................. 49
-6
Figure 37. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the ASC site (Richards 2012) .................................... 50
Figure 38. Steamship (SS) – boilers (Carpenter 2012) ........................................................... 51
Figure 39. Change in dissolved oxygen content with increasing water depth measured
on the SS site (Richards 2012) ................................................................................................. 52
iii
Appendix A
-6
Figure 40. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the SS site (Richards 2012) ....................................... 54
Figure 41. Grumman TBM Avenger (AVR) (Carpenter 2012) ................................................. 56
Figure 42. Change in dissolved oxygen content with increasing water depth measured
on the AVR site (Richards 2012) .............................................................................................. 57
Figure 43. Location of the aircraft wrecks, Saipan, CNMI (Richards 2012 after Google
Earth 2012) ................................................................................................................................. 58
Figure 44. Schematic plan of the TBM Avenger (AVR) indicating the corrosion
parameter measurement positions (Richards 2012 after Bell 2010 in McKinnon and
Carrell 2011:89) .......................................................................................................................... 60
-6
Figure 45. Pourbiax diagram for aluminium (3.7 x 10 M) in aerobic seawater at 25°C indicating
the intercepts of the areas measured on the AVR site (Richards 2012) .............................. 60
Figure 46. Aichi E13A (JAKE) (Carpenter 2012) ..................................................................... 62
Figure 47. Change in dissolved oxygen content with increasing water depth measured
on the JAKE site (Richards 2012) ............................................................................................ 64
Figure 48. Schematic plan of the Aichi E13A (JAKE) indicating the corrosion
parameter measurement positions (Richards 2012 after Bell 2010 in McKinnon and
Carrell 2011:56) .......................................................................................................................... 65
-6
Figure 49. Pourbiax diagram for aluminium (3.7 x 10 M) in aerobic seawater at 25°C indicating
the intercepts of the areas measured on the JAKE site (Richards 2012) ............................ 66
Figure 50. Martin PBM Mariner (MNR) (Carpenter 2012) ........................................................ 67
Figure 51. Change in dissolved oxygen content with increasing water depth measured
on the MNR site (Richards 2012) .............................................................................................. 68
Figure 52. Schematic plan of the Martin PBM Mariner (MNR) indicating the corrosion parameter
measurement positions (Richards 2012 after Bell 2010 in McKinnon and
Carrell 2011:78) .......................................................................................................................... 70
-6
Figure 53. Pourbiax diagram for aluminium (3.7 x 10 M) in aerobic seawater at 25°C indicating
the intercepts of the areas measured on the MNR site (Richards 2012) ............................. 71
Figure 54. Kawanishi H8K (EMILY) (Carpenter 2012) ............................................................. 72
Figure 55. Change in dissolved oxygen content with increasing water depth measured
on the EMILY site (Richards 2012) ........................................................................................... 73
Figure 56. Schematic plan of the Kawanishi H8K (EMILY) indicating the corrosion
parameter measurement positions (Richards 2012 after Bell 2010 in McKinnon and
Carrell 2011:63) .......................................................................................................................... 75
-6
Figure 57. Pourbiax diagram for aluminium (3.7 x 10 M) in aerobic seawater at 25°C indicating
the intercepts of the areas measured on the EMILY site (Richards 2012) ........................... 76
Figure 58. Coronado (CRDO) (Carpenter 2012) ...................................................................... 77
Figure 59. Change in dissolved oxygen content with increasing water depth measured
on the CRDO site (Richards 2012) ........................................................................................... 78
iv
Appendix A
Figure 60. Schematic plan of the Consolidated PB2Y Coronado (CRDO) indicating the corrosion
parameter measurement positions (Richards 2012 after Harvey and
Raupp 2012) ............................................................................................................................... 80
-6
Figure 61. Pourbiax diagram for aluminium (3.7 x 10 M) in aerobic seawater at 25°C indicating
the intercepts of the areas measured on the CRDO site (Richards 2012) ........................... 81
v
Appendix A
LIST OF TABLES
Table 1. Tidal variations for Saipan during the survey period (20–24 February 2012) ......... 5
Table 2. Dissolved oxygen content, salinity and temperature of the seawater on the
M4 Sherman Tank 1 site .............................................................................................................. 6
Table 3. Corrosion parameter measurements of the M4 Sherman Tank 1 ............................ 8
Table 4. Dissolved oxygen content, salinity and temperature of the seawater on the
M4 Sherman Tank 3 site ............................................................................................................ 12
Table 5. Corrosion parameter measurements of the M4 Sherman Tank 3 .......................... 13
Table 6. Dissolved oxygen content, salinity and temperature of the seawater on the
LVT1 site ..................................................................................................................................... 17
Table 7. Corrosion parameter measurements of the LVT1 ................................................... 19
Table 8. Corrosion parameter measurements on the LVT2 .................................................. 23
Table 9. Dissolved oxygen content, salinity and temperature of the seawater on the
DAI1 and DAI2 sites ................................................................................................................... 27
Table 10. Corrosion parameter measurements on the DAI1 ................................................. 29
Table 11. Corrosion parameter measurements on the DAI2 ................................................. 33
Table 12. Dissolved oxygen content, salinity and temperature of the seawater on the
DAI3 site ..................................................................................................................................... 37
Table 13. Corrosion parameter measurements on the DAI3 ................................................. 38
Table 14. Dissolved oxygen content, salinity and temperature of the seawater on the
JFR site ....................................................................................................................................... 42
Table 15. Corrosion parameter measurements on the JFR .................................................. 44
Table 16. Dissolved oxygen content, salinity and temperature of the seawater on the
ASC site ...................................................................................................................................... 47
Table 17. Corrosion parameter measurements on the ASC .................................................. 49
Table 18. Dissolved oxygen content, salinity and temperature of the seawater on the
SS site ........................................................................................................................................ 52
Table 19. Corrosion parameter measurements on the SS .................................................... 54
Table 20. Dissolved oxygen content, salinity and temperature of the seawater on the
AVR site ...................................................................................................................................... 57
Table 21. Corrosion parameter measurements on the AVR .................................................. 59
Table 22. Dissolved oxygen content, salinity and temperature of the seawater on the
JAKE site .................................................................................................................................... 63
vi
Appendix A
Table 23. Corrosion parameter measurements on the JAKE................................................ 65
Table 24. Dissolved oxygen content, salinity and temperature of the seawater on the
MNR site ..................................................................................................................................... 68
Table 25. Corrosion parameter measurements on the MNR ................................................. 70
Table 26. Dissolved oxygen content, salinity and temperature of the seawater on the
EMILY site ................................................................................................................................... 73
Table 27. Corrosion parameter measurements on the EMILY .............................................. 75
Table 28. Dissolved oxygen content, salinity and temperature of the seawater on the
CRDO site ................................................................................................................................... 78
Table 29. Corrosion parameter measurements on the CRDO............................................... 79
Table 30. Average corrosion parameter measurements for all iron alloy wrecks
measured in Saipan, CNMI from 20-24 February 2012........................................................... 82
Table 31. Average corrosion parameter measurements for all aluminium alloy aircraft wrecks
measured in Saipan, CNMI from 20-24 February 2012........................................................... 84
vii
Appendix A
LIST OF APPENDICES
Appendix A. On-Site Conservation Survey Data Sheet ......................................................... 89
Appendix B. On-Site Corrosion Survey Data Sheet ............................................................... 91
viii
Appendix A
EXECUTIVE SUMMARY
Ships of Exploration and Discovery Research Inc. (Ships) requested the services of a diving
conservation scientist to conduct pre-disturbance conservation surveys on selected World War II
underwater archaeological wreck sites in Saipan, Commonwealth of the North Marianas Islands and
provide recommendations for the development of a management plan for long-term preservation which
will be written by Ships archaeologists. This management plan will be developed by Ships
archaeologists in cooperation with the Historic Preservation Office, CNMI (HPO) and will outline an
appropriate framework for the monitoring, management and conservation of World War II resources in
the CNMI. This plan will be delivered to HPO for future reference and use.
Based on the scope of work and the eligibility requirements Dr Jennifer McKinnon awarded the
contract to Vicki Richards, Conservation Scientist, Department of Materials Conservation (DMC),
Western Australian Museum (WAM) in collaboration with Jonathan Carpenter, Senior Conservator,
DMC, WAM. Conservation surveys of 15 individual wreck sites were carried out by Vicki Richards and
Jonathan Carpenter, assisted by a small archaeological team, over a one week field season on
Saipan, CNMI in February 2012. This report will describe the results obtained from these conservation
surveys, discuss the data with respect to site stability and suggest some recommendations for future
monitoring and long-term preservation, which may assist in the development of a holistic management
program for these UCH sites.
In general, the physico-chemical measurements (pH, Eredox, dissolved oxygen, salinity, temperature,
etc) of the local environment surrounding the wreck sites in Saipan are typical for a shallow, near
coastal, open circulation, oxidising marine environment, where corrosion rates are likely to be relatively
high for both ferrous alloy wrecks and aluminium alloy aircraft. All of the wrecks and the aircraft were
mostly exposed with only very thin layers of sediment covering some lower profile areas lying on the
seabed, which would be particularly mobile during periods of excessive water movement (i.e. storm
and cyclonic activity). Hence, natural protection via seasonal sediment burial would be very unlikely for
any of the wrecks surveyed in 2012.
The corrosion parameters of a number of different areas on each of the ten iron alloy wrecks and the
five aluminium alloy aircraft were measured during the survey period from 20-24 February 2012.
Based on the corrosion parameter data and the environmental and historical information some
conclusions were drawn about the differences in corrosion behaviour of the fifteen wreck sites.
The natural and cultural impacts of the local environment on the M4 Sherman Tank 3 (Tank 3) are
more aggressive than those experienced by the M4 Sherman Tank 1 (Tank 1). More importantly, as
there appears to be more tourist activity associated with Tank 3, it may be this increase in human
interference that is causing the accelerated deterioration of Tank 3. It is difficult to say whether the
Landing Vehicle Tracked 2 (LVT2) is corroding at a faster rate than the Landing Tracked Vehicle 1
(LVT1), however, considering the extent of deterioration of the LVT2 as compared to the LVT1 it would
appear that the natural and cultural impacts on the LVT2 would be greater than those experienced by
the LVT1. It appears that the Daihatsu Landing Craft 2 (DAI2) is corroding at a slightly faster rate than
both the Daihatsu Landing Craft 1 (DAI1) and the Daihatsu Landing Craft 3 (DAI3). In addition, it
appears that DAI1 and DAI3 are corroding at relatively similar rates, despite the fact that DAI3 is a
much shallower site, where it would be expected that the corrosion rate would be slightly higher. This
would seem to suggest that human interference (i.e. recreational diving activities) is having some
impact on the deterioration rate of the deeper DAI1 site. It appears that the small but statistically valid
decrease in the corrosion potential (Ecorr) of the unidentified Steamship (SS) suggests that this vessel
is corroding at a slower rate than both the Auxiliary Submarine Chaser (ASC) and the Japanese
Freighter (JFR) and the small increase in the average E corr of the ASC suggests that it may be
corroding at a slightly faster rate than JFR. This would seem to suggest that human interference (i.e.
recreational diving activities) is having some impact on the deterioration rate of the JFR and ASC sites
as the SS site is not on the diving heritage trail. However, the local environment (i.e. increase in
turbidity) may also be contributing to this decrease in the corrosion rate on the SS site.
1
Appendix A
It is difficult to determine any differences in corrosion rates based on the corrosion parameter data for
the aluminium alloy aircraft wrecks as there are no statistically valid differences between any of the
average corrosion parameter measurements as all fall within the maxima/minima range calculated
from the standard deviations for each set of data points. However, since all aluminium alloys are
corroding in a common oxidising marine environment in Tanapag Lagoon, the different values of the
corrosion potentials may provide a guide to the underlying differences in alloy composition of the
aircraft. The metal composition of the aluminium alloys for each aircraft, in order of decreasing
concentrations of incorporated copper is the Avenger > Jake > Mariner ~ Coronado > Emily. That is
the Avenger may have the highest concentration of copper in this group of aluminium alloys measured
whilst the Emily will have the lowest based on this data set. This may have consequences for the
corrosion rates of these aircraft as higher concentrations of copper will increase the rate of pitting and
intergranular corrosion if the aircraft are subjected to similar environmental conditions and other
complicating factors, such as increases in corrosion through stress and metal fatigue, are absent.
Since this is not the case with these aircraft (i.e. the Avenger lies in a very aggressive, shallower
marine environment and the Coronado is extensively damaged with separate sections strewn over a
very large area) this highlights the problem with interpreting corrosion rates based on only one set of
corrosion parameter measurements for aluminium alloy wrecks.
In conclusion, a holistic approach must be taken using all the data obtained including the
environmental and historical information in order to understand the corrosion processes occurring on a
wreck site. Hence, continued observation of the sites and further corrosion measurements in the future
may assist in corroborating or refuting the aforementioned inferences.
It is recommended that site inspections of these fifteen wreck sites using the guidelines provided by
Richards and Carpenter, are undertaken at regular intervals and after any severe storm or cyclonic
activity so any changes in the integrity of the sites are noted. The more surveys carried out the better
as it will provide more information regarding the rate of deterioration and the inherent stability of a site,
which will assist in recognising which sites are a priority for future implementation of appropriate in situ
conservation management strategies. In addition, it is recommended that another full corrosion and
environmental survey using the underwater corrosion survey equipment is performed in another few
years. In this way, from comparisons of the regular site inspection results and the additional corrosion
parameter data for each wreck site, it will be possible to ascertain if there is indeed any effect from
diving tourism on the sites and if it is at all comparable to the detrimental effects afforded by natural
occurrences, such as seasonal storm and cyclonic activity. Finally, using a combination of information
gathered from these surveys it will be possible to prioritise these submerged sites with respect to their
overall in situ management requirements and the most appropriate management plans determined
and applied to each site.
2
Appendix A
1
INTRODUCTION
1.1
BACKGROUND
Ships of Exploration and Discovery Research Inc. (Ships) requested the services of a diving
conservation scientist to conduct pre-disturbance conservation surveys on selected World War II
underwater archaeological wreck sites in Saipan, Commonwealth of the North Marianas Islands and
provide recommendations for the development of a management plan for long-term preservation which
will be written by Ships archaeologists. This management plan will be developed by Ships
archaeologists in cooperation with the Historic Preservation Office, CNMI (HPO) and will outline an
appropriate framework for the monitoring, management and conservation of World War II resources in
the CNMI. This plan will be delivered to HPO for future reference and use.
1.2
SCOPE OF THE WORK
Working with Ships archaeologists, the contractor will:
1.
2.
3.
4.
5.
Review previous archaeological survey results and consult with Ships archaeologists in
selecting wreck sites for conservation surveys.
Conduct pre-disturbance conservation surveys on selected wreck sites including the
collection of environmental, chemical and cultural data.
Provide on-site training for Historic Preservation Office and Coastal Resources Management
staff on conducting conservation surveys.
Conduct conservation assessments on selected sites with regards to site condition,
environmental, chemical and cultural factors affecting the site, corrosion activities of metals
and degradation of organic materials, and overall stability of wreck sites.
Prepare recommendations for the development of an appropriate conservation management
program for long-term preservation.
The eligibility requirements for the contractor:
1.
2.
3.
4.
5.
Extensive experience in in situ conservation surveys.
Ability to meet RFP specifications based on the Scope of Work.
Demonstrated understanding of and working on archaeological research projects.
Previous experience in the production of conservation management programs.
Previous experience in training managers in conservation survey programs.
Preference will be given to producers who have experience working with underwater archaeologists
and on WWII underwater wreck sites.
Based on the scope of work and the eligibility requirements Dr Jennifer McKinnon awarded the
contract to Vicki Richards, Conservation Scientist, Department of Materials Conservation (DMC),
Western Australian Museum (WAM) in collaboration with Jonathan Carpenter, Senior Conservator,
DMC, WAM.
Conservation surveys of 15 individual wreck sites were carried out by Vicki Richards and Jonathan
Carpenter, assisted by a small archaeological team, over a one week field season on Saipan, CNMI in
February 2012. This report will describe the results obtained from these conservation surveys, discuss
the data with respect to site stability and suggest some recommendations for future monitoring and
long-term preservation, which may assist in the development of a holistic management program for
these UCH sites.
3
Appendix A
2
METHODOLOGY
A series of corrosion parameter measurements [pH; corrosion potential; total depth of penetration
(concretion + corrosion); water depth] were conducted on each of the fifteen wrecks (10 iron alloy
based wrecks and 5 aluminium alloy aircraft) to determine the underlying nature of the corrosion
processes. The surface pH measurements were effected by a VWR epoxy body, flat surface pH
electrode connected to a Cyberscan 200 pH meter and the corrosion potentials measured via a
platinum electrode connected to a high impedence Finest digital multimeter set to read at 2V direct
current. Both meters were housed in a custom-built plexiglass waterproof housing. In order to obtain
reproducible results, it was essential that the measurements were taken by a two person dive team
(one assistant diver; one operating diver): the assistant diver drilled and filled the holes while the
operating diver conducted the measurements and took the positional photographs. However, during
this fieldwork season, in order expedite the measuring process, a third diver filled the drill holes with
epoxy while a fourth diver took the photographs.
Contact was made with the underlying residual metal by drilling through the marine growth with an airpowered pneumatic drill equipped with a masonry tungsten-tipped bit. This type of drill bit drilled
through the concretion and corrosion product layer but did not penetrate into the sound residual metal.
When the drill could not penetrate further (the metal surface had been reached), the drill bit was
removed by the assistant diver and the operating diver immediately inserted the flat-surface glass pH
electrode into the drill hole. The minimum pH of the microenvironment created by the corroding metal
was recorded.
Following the pH measurement, the platinum electrode was inserted into the same drill hole and the
corrosion potential (in volts) was recorded. Good electrical contact was made with the underlying metal
when the voltage only changed by ±0.001V measured against a flow-through silver/silver chloride/
seawater reference electrode attached to the underwater housing lying immediately adjacent to the
area of measurement. If the voltage reading was not stable the assistant diver drilled another hole
immediately adjacent to the previous position and the entire measurement process was repeated (i.e.
pH, corrosion potential, etc).
The total depth of penetration (concretion + corrosion) was then measured with a plastic vernier. If
possible, the depth of corrosion was also measured. This is where the protective encapsulating layer
of marine concretion ceased and the original outer surface of the metal began measured to the bottom
of the drill hole. However this interface is extremely difficult to discern under most circumstances.
The water depth to the drill hole was then measured with a digital dive computer and a series of
photographs taken to indentify the measurement position on-site. The drill hole was then filled with an
underwater curing two part epoxy sealant (e.g. Selleys Knead-It).
The measurement of pH on the aluminium alloy aircraft surfaces was difficult owing to the very thin
layer, often less than 1 mm, of marine growth and corrosion products. The thin nature of this surface
deposit and the inherent softness of aluminium alloys meant that the use of any form of drill was
inappropriate. The assistant diver used the flat end of a diving knife to scrape the surface and
immediately the operating diver placed the pH electrode against the exposed shiny metal surface to
record the underlying acidity. These experimental difficulties meant that the pH values on the aircraft
were generally very conservative, i.e. generally the pH will be lower than was reported. The corrosion
potential, water depth and photographs were then taken but obviously the depth of penetration was not
measured and there was no requirement to fill any drill holes with epoxy.
The temperature, salinity and dissolved oxygen content of the seawater column were measured on
each site at 0.5m intervals to the seabed surface with the appropriate underwater sensors connected
to a TPS 90DC microprocessor, which was located on the dive boat.
Finally an on-site conservation survey data sheet was completed for every site (Appendix A).
4
Appendix A
3
CONSERVATION ASSESSMENTS – IRON ALLOY WRECKS
3.1
M4 SHERMAN TANK 1 – TANK 1
Figure 1. M4 Sherman Tank 1 (Tank 1) – port side view (Carpenter 2012).
Date of Inspection
20 February 2012
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). In the morning (20/6/2012) the winds were ENE at 13 to 18 knots which
tended more easterly in the afternoon, increasing to 15 to 21 knots. Seas were relatively consistent
over the entire day with breezy whitecapping conditions and moderate choppy seas with small, short
period wind waves (morning - ENE 1.5m at 10 seconds; afternoon – NE 1.7m at 10 seconds). The
tides were semi-diurnal over the survey period and are reported in Table 1 (http://buoyweather.com).
Table 1. Tidal variations for Saipan during the survey period (20–24 February 2012).
Day
Mon 20 Feb 2012
Tues 21 Feb 2012
Wed 22 Feb 2012
Thurs 23 Feb 2012
Fri 24 Feb 2012
Lower low water
-0.07m @ 1236
-0.05m @ 0119
-0.02m @ 0157
0.02m @ 0231
0.07m @ 0302
Higher high water
0.60m @ 0743
0.60m @ 0813
0.60m @ 0840
0.59m @ 0904
0.58m @ 0927
Higher low water
0.32m @ 1316
0.27m @ 1356
0.23m @ 1432
0.20m @ 1507
0.16m @ 1541
Lower high water
0.55m @ 1818
0.55m @ 1910
0.54m @ 1956
0.53m @ 2051
0.51m @ 2124
5
Appendix A
indicates that there is very little variation in the dissolved oxygen content with increasing water depth
over such a shallow depth range, which is not unexpected. Hence, this trend coupled with the other
physico-chemical measurements, are typical for a shallow, near coastal, open circulation, oxidising
marine environment, where corrosion rates are likely to be relatively high.
Wreck Site
Tank 1 is located on the south western side of Saipan, inside the barrier reef about 180m off shore
from Susupe Beach (GPS
) at a depth of about 2m (Figure 3). The tank
identified as a M4A2 Dry model, constructed principally of rolled and cast homogenous steel, is 5.84m
in length, 2.62m wide and 2.74m in height (Grove 1976:130-131). Tank 1 is orientated with its bow
pointing towards the shore on a bearing of 133° and the 75mm gun fixed on a bearing of 197°
(McKinnon and Carrell 2011:106). The tank is semi-submerged and at low tide all components above
the upper hull including the turret and gun are exposed to the atmosphere (Figure 1). Bright high nickel
welds are evident on the upper hull edges.
Figure 3. Location of the M4 Sherman Tanks, Saipan, CNMI (Richards 2012 after Google Earth
2012).
The surrounding seabed is relatively flat, comprising of calcareous sediment interspersed with large
patches of seagrass. The tank lies in a shallow depression with gently sloping edges and the lower
2
section of the track and roller assembly is mostly buried. A circular area about 12m surrounding the
tank is free of seagrass but algal forms are present on the seabed and on the submerged parts of the
hull. High nutrient levels in the lagoon may be contributing to this extensive algal growth (Denton et al.
2001).
7
Appendix A
The main body of the vehicle is mostly intact but other smaller components are missing. This loss may
be due to corrosion and/or cultural impacts, such as salvage. There are many areas of active corrosion
evident on the site, indicated by the presence of the typical red/brown “rust” spots on the surfaces of
the tank (Figure 1). There are also signs of accelerated corrosion on the upper sections of the tank
(flaking, spalling and cracking of the metal surfaces and gun barrel) that are cyclically exposed to the
atmosphere. Children were observed playing on the tank, walking along the main gun barrel and
jumping or diving into the water. In time, this practice may become a health and safety issue due to the
extensive corrosion exhibited by these exposed sections and the ever increasing probability that these
areas may collapse or fragment. However, divers and snorkelers are actively encouraged to visit the
site through the WWII Maritime Heritage Trail – Battle of Saipan, provided they follow the visitation
guidelines and do not interfere with the site (i.e. disturb or attempt to remove any components).
Corrosion Survey
The corrosion parameters of eleven different areas on Tank 1 were measured over a 70 minute dive
on 20 February 2012. The results are presented in Table 3 and the on-site positions shown in Figure
5. In order to compare the corrosion data collected from the different positions measured on Tank 1
and ascertain the thermodynamically stable state of the iron, the corrosion potentials (Ecorr) and the pH
of the residual iron alloy surfaces were plotted on the iron Pourbaix diagram in aerobic seawater at
25°C (Figure 4). The exception being the exposed weld as this particular Pourbaix diagram is not
applicable due to the high nickel content which will significantly change the corrosion mechanism. The
temperature of the seawater on-site was 26°C, however this 1°C increase does not significantly affect
the nature or equilibria of the chemical species described in this diagram.
Table 3. Corrosion parameter measurements on the M4 Sherman Tank 1.
Position
Number
1
2
3
4
5
6
7
8
9
10
11
Description
stbd aft vertical volute spring on the
suspension bogie (immersed)
stbd side of upper hull (immersed)
stbd top surface of upper hull (splash
zone)
stern midline (immersed)
port aft vertical volute spring on the
suspension bogie (immersed)
port side of upper hull (immersed)
port upper track (immersed)
port top surface of upper hull (splash
zone)
bow midline (immersed)
stbd side turret (splash zone)
stbd side exposed high nickel weld
(splash zone)
pH
Ecorr vs
NHE (V)
Water
Depth
(m)
-0.305
dtotal = depth
of concretion
+ corrosion
(mm)
23
6.95
6.51
6.40
-0.307
-0.306
39
5
0.9
0.7
6.87
6.89
-0.304
-0.304
24
24
1.0
1.6
6.53
6.73
6.51
-0.304
-0.299
-0.303
14
23
29
0.8
1.2
0.5
7.38
8.06
nd
-0.306
ns
-0.304
9
9
na
0.9
0.3
0.5
1.7
ns = not stable
nd = not determined
na = not applicable
8
Appendix A
(a)
Fe3+
0.8
active
0.6
FeO.OH
0.4
passive
0.2
Eh(v)
0.0
Fe2+
active
(b)
2, 6 & 8
5 1
-0.2
3
-0.4
7
4
9
Fe3O4
passive
-0.6
-0.8
Fe
-1.0
immune
0
1
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 4. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the intercepts
of the areas measured on the M4 Sherman Tank 1 (Richards 2012).
9
Appendix A
Figure 5. Schematic plan and profile views of the M4 Sherman Tank 1 indicating the corrosion
parameter measurement positions (Richards 2012 after Hanks 2010 in McKinnon and Carrell
2011:107).
10
Appendix A
Generally, the areas on the tank that are constantly immersed in this oxidising marine environment
were covered with relatively thick aerobic concretions (>20mm). Those areas that are either in the
splash zone or subjected to wetting/drying cycles, such as the turret, had significantly reduced
concretion layers thicknesses (<10mm). It should be noted that there was very little secondary marine
growth on the concretion with the exception of algae and some seaweed species, which suggests that
there is a significant amount of water movement on this shallow site and possible sediment
impingement during periods of rough sea conditions, which would significantly reduce colonisation
rates and increase corrosion rates. In addition, the deleterious effect on the growth by human
interference, such as walking on the upper surfaces of the tank, cannot be underestimated and will
also lead to thinner concretions subsequently increasing corrosion rates in these areas. Iron is not
biologically toxic and increases the growth rate of encrusting organisms but the exposed welds on the
edges of the upper section of the tank (11) possessed no concretion. These welds are high in nickel,
which inhibits concretion formation.
From the Pourbaix diagram (Figure 5), the intercepts of all points (1 – 9) measured on the iron alloy
sections of tank 1 (no stable corrosion potential could be obtained on the gun turret 10) lie in the active
corrosion region, where ferrous ions are the thermodynamically stable chemical species and corrosion
will continue until all iron is consumed. Generally, with film free corrosion mechanisms, such as occurs
on concreted iron artefacts, an increase in the corrosion potential (tending more positive) indicates an
increase in the corrosion of the metal. However, the average corrosion potential of the nine
measurement points was -0.305 ± 0.003V. This 3mV standard deviation is comparatively small and
within experimental error for the equipment and measuring procedure suggesting that the entire tank is
in electrical connection and the same film free corrosion mechanism applies to all areas on the tank.
Hence, it is not possible to determine any differences in corrosion between the positions based on the
Ecorr data.
It should be noted that Pourbaix diagrams are thermodynamic stability maps and therefore, do not
provide kinetic information with regard to corrosion rates. However, it is possible to calculate the
annualised corrosion rate if the depth of corrosion of the measurement point and the years of
immersion of the wreck is known. Unfortunately, on tank 1 which had been immersed for 68 years, it
was not possible to discern where the concretion layer ceased and the corrosion layer began, hence
the actual depth of corrosion could not be measured so it was not possible to estimate corrosion rates.
Since Ecorr data describes the electrochemical environment of the iron alloy that is electrically
connected to the measurement point (e.g. with tank 1 this is a very large surface area as all points are
in electrical connection) it is not as sensitive to changes in localised corrosion processes as the value
of the pH recorded at the same point, provided no damage has occurred to the protective concretion
layer. It has been shown that pH data is a useful guide to the corrosion rate, since the pH is controlled
2+
by the dynamic equilibrium (Equation 1) between the concentration of the Fe ions (represented as
FeCl2 in Equation 1) and their acidic hydrolysis products and is therefore, more sensitive to changes in
apparent corrosion rate (MacLeod and Richards 2011).
2FeCl2 + 2H2O [Fe(OH)2.FeCl2] + 2H + 2Cl ......................................................................... (1)
+
-
2+
So, generally as corrosion rate increases the concentration of Fe ions underneath the protective
layer of concretion increases, correspondingly the extent of hydrolysis increases producing more
hydrogen ions causing the pH to decrease (become more acidic).
On tank 1, the deeper positions below 1m (1, 4, 5 & 7) have an average pH value of 6.86 ± 0.09
whereas the shallower positions, above 1m (2, 3, 6 & 8) have a more acidic average pH of 6.49 ± 0.06.
This decrease of 0.37 pH units indicates that there has been a statistically significant increase in the
corrosion rate of the upper sections of the tank. This is not unexpected as the major site variable that
dominates the overall corrosion rate of iron is the amount of water movement and thus the flux of
oxygenated seawater to the concreted iron surface, which is directly affected by the water depth.
Hence, the shallower the position on the tank the greater the amount of water movement and oxygen
impingement to the concreted iron surface, consequently the corrosion rate will increase accordingly.
This increase in corrosion rate is further exacerbated by wetting/drying cycles that are experienced by
areas of the tank in the splash zone.
11
Appendix A
It has been previously reported (MacLeod et al. 2007) that the thickness of concretion is an important
factor in determining how effective the marine growth is in establishing separation of the anodic and
cathodic sites of the corrosion cell and this, in turn will be reflected in the pH values. On wreck sites
where there have been episodic deconcretion events, either caused by natural phenomena, such as
storms and cyclones, cyclic wetting/drying cycles and/or by human intervention, it takes some time for
the marine organisms to regrow and the rate of regrowth is dependent on a variety of interrelated
factors. Thus when measurement points are accessed there is a chance that the pH recorded is more
alkaline than the underlying long term corrosion rates would indicate. In simple terms, more recently
deconcreted and recolonised areas tend to present more alkaline pH values (e.g. turret 10 pH = 8.06;
dtotal = 9mm) whereas the fully matured sections possess more acidic values. In this instance, it is
important not to confuse alkaline pH values with low corrosion rates for without knowledge of the
corrosion thickness and the environmental history of the vessel it is not wise to apply simplistic
interpretation of the data as this can imply that the rate of corrosion is low whereas it is usually high in
these particular areas.
Interestingly, using this corrosion parameter data, the limit of the splash zone can be estimated to a
maximum depth of about 1m, even though the survey was carried out during high tide and the
shallower sections of the tank were fully immersed. This depth also corresponds to the average
maximum tidal range experienced on this site.
3.2
M4 SHERMAN TANK 3 – TANK 3
Figure 6. M4 Sherman Tank 3 (Tank 3) – stern view (Carpenter 2012).
Date of Inspection
24 February 2010
12
Appendix A
Similar to the Tank 1 site, the relatively small standard deviation between the measurements and a
decrease of only 0.33ppm over 1.5m indicates that there is very little variation in the dissolved oxygen
content with increasing water depth over such a shallow depth range, which is not unexpected. There
had been a 2ppK decrease in the salinity and a drop in pH (pH = 7.49) on the Tank 3 site, which may
indicate some fresh water contamination from land run off as it had rained quite heavily on previous
days prior to this survey. However, the measurements are still within range for a shallow, near coastal,
open circulation, oxidising marine environment, where corrosion rates are likely to be relatively high.
Wreck Site
Tank 3 is located approximately 1 km south of Tank 1 about 200m off shore from Chalan Kanoa
Beach, near Saipan World Resort and Saipan Grand Hotel (GPS
at a
depth of about 2m (Figure 3). The tank, identified as a M4A3 Wet model, constructed principally of
rolled and cast homogenous steel, is 5.91m in length, 2.62m wide and 2.74m in height (Grove
1976:130-131). Tank 1 is orientated with its bow pointing seaward on a bearing of 295° and the 75mm
gun fixed on a bearing of 60° (McKinnon and Carrell 2011:109-110). The tank is semi-submerged and
at low tide all components above the upper hull including the turret and gun are exposed to the
atmosphere (Figure 6). Bright high nickel welds are evident on the upper hull edges.
The surrounding seabed is relatively flat, comprising of calcareous sediment interspersed with large
patches of seagrass. The tank lies proud of the seabed and the lower track is visible. A circular area
2
about 8m surrounding the tank is free of seagrass but dead coral and algal forms are present on the
seabed. Extensive algal mats are present on the submerged parts of the hull. High nutrient levels in
the lagoon may be contributing to this extensive algal growth (Denton et al. 2001).
The main body of the vehicle is mostly intact but other smaller components are missing. Most obvious
is the loss of the engine cover and cowling and the gun barrel is broken. These remains are lying on
the seabed in close proximity to the tank. This loss may be due to corrosion and/or physical damage
by natural and/or human impacts. There are many areas of active corrosion evident on the site,
indicated by the presence of the typical red/brown “rust” spots on the surfaces of the tank (Figure 6).
There are also signs of accelerated corrosion on the upper sections of the tank (flaking, spalling and
cracking of the metal surfaces and the broken gun barrel) that are cyclically exposed to the
atmosphere. No human activity was observed on the site at the time of the survey but on a previous
survey in 2011, tour boats and ‘banana’ boats frequently passed near the site, tourists used jet skis on
a race course just north of the site and there was significantly more rubbish around the tank than
observed on the Tank 1 site (McKinnon and Carrell 2011:114). The fact that the site is located in close
proximity to two large hotels would account for this increase in human interference. However, divers
and snorkelers are still actively encouraged to visit the site provided they follow the visitation
guidelines and do not interfere with the site (i.e. disturb or attempt to remove any components).
Corrosion Survey
The corrosion parameters of eleven different areas on Tank 3 were measured over a 40 minute dive
on 24 February 2012. The results are presented in Table 5 and the on-site positions shown in Figure
8. In order to compare the corrosion data collected from the different positions measured on Tank 3
and ascertain the thermodynamically stable state of the iron, the corrosion potentials (E corr) and the pH
of the residual iron alloy surfaces were plotted on the iron Pourbaix diagram in aerobic seawater at
25°C (Figure 9). The exception being the exposed weld as this particular Pourbaix diagram is not
applicable due to the high nickel content which will significantly change the corrosion mechanism. The
temperature of the seawater on-site was 27°C, however this 2°C increase does not significantly affect
the nature of the chemical species described in this diagram.
Table 5. Corrosion parameter measurements on the M4 Sherman Tank 3.
Position
Number
1
Description
port mid vertical volute spring on the
suspension bogie (immersed)
pH
Ecorr vs
NHE (V)
6.41
-0.324
dtotal = depth
of concretion
+ corrosion
(mm)
5
Water
Depth
(m)
1.8
14
Appendix A
2
3
4
5
6
7
8
9
10
11
port upper track
port side of upper hull (splash zone)
port side turret (splash zone)
stern midline (immersed)
stbd mid vertical volute spring on the
suspension bogie (immersed)
stbd upper track (immersed)
stbd side of upper hull (splash zone)
stbd side turret (splash zone)
stbd side exposed high nickel weld
(splash zone)
bow midline (immersed)
6.01
7.09
7.49
5.89
6.61
-0.323
-0.321
-0.316
-0.323
-0.320
4
4
1
8
5
1.5
0.9
0.4
1.2
1.9
6.74
5.92
6.07
7.49
-0.319
-0.320
-0.318
-0.318
11
10
10
na
1.5
1.0
0.5
0.8
6.38
-0.316
1
1.6
na = not applicable
15
Appendix A
Figure 8. Schematic plan and profile views of the M4 Sherman Tank 3 indicating the corrosion
parameter measurement positions (Richards 2012 after Hanks 2010).
(a)
Fe3+
0.8
active
0.6
FeO.OH
0.4
passive
0.2
Eh(v)
0.0
Fe2+
active
(b)
2 91
6
-0.2
5
-0.4
3
8 11
7
Fe3O4
4
passive
-0.6
-0.8
Fe
-1.0
immune
0
1
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 9. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the intercepts
of the areas measured on the M4 Sherman Tank 3 (Richards 2012).
Generally, the areas measured on Tank 3, whether they are constantly immersed or lie in the splash
zone, were covered with relatively thin aerobic concretions (<11mm) in comparison to Tank 1. Unlike
16
Appendix A
Tank 1 there appears to be no relationship between the depth of immersion and the concretion
thickness. However, similar to Tank 1, there were very few secondary colonising organisms on the
concretion with the exception of algal forms and some seaweed species sporadically located on the
upper surfaces of the tank. Again this suggests that there is a significant amount of water movement
on this shallow site and possible sediment impingement during periods of rough sea conditions, which
would significantly reduce colonisation rates and increase corrosion rates. In addition, if the decrease
in salinity is due to fresh water ingress then this could also decrease concretion formation and reduce
the growth rate of many colonising marine organisms, which generally require very specific salinities
for reproduction and proliferation.
The exposed high nickel welds on the upper section of Tank 3 (10) were also devoid of concretion.
From the Pourbaix diagram (Figure 9), the intercepts of all points (1 – 9 & 11) measured on the iron
alloy sections of Tank 3 lie in the active corrosion region, where ferrous ions are the
thermodynamically stable chemical species and corrosion will continue until all iron is consumed.
Generally, with film free corrosion mechanisms, such as occurs on concreted iron artefacts, an
increase in the corrosion potential (tending more positive) indicates an increase in the corrosion of the
metal. However, the average corrosion potential of the ten measurement points was -0.320 ± 0.003V.
This 3mV standard deviation is comparatively small and within experimental error for the equipment
and measuring procedure suggesting that the entire tank is in electrical connection and the same film
free corrosion mechanism applies to all areas on the tank. The average corrosion potential for Tank 1
was -0.305 ± 0.003V, which is only 15mV more positive than Tank 3, hence, it is not possible to
determine any differences in corrosion behaviour between the measurement points on Tank 3 and
between Tank 1 and Tank 3 based on the Ecorr data.
Again, it was not possible to discern the interface between the concretion and the corrosion product
layers on Tank 3 so the depth of corrosion could not be measured, therefore it was not possible to
calculate the annualised corrosion rate for this tank.
As mentioned previously, the pH is often a more reliable indicator of changes in localised corrosion
rates, however the pHs measured on Tank 3 were very inconsistent in comparison to those measured
on Tank 1. Therefore, no statistically valid relationships between depth of immersion and changes in
pH could be observed for Tank 3. The areas that are constantly immersed (water depth >1m) will tend
to provide more consistent corrosion parameter results. Therefore, the average pH values of these
areas for Tank 1 (1, 4, 5 & 7) and for Tank 3 (1, 2, 5, 6, 7 & 11) were 6.86 ± 0.09 and 6.34 ± 0.33,
respectively. This decrease of 0.52 pH units for Tank 3 indicates that there has been a statistically
significant increase in the corrosion rate of Tank 3 as compared to Tank 1 and in conjunction with the
thinner dtotals measured on Tank 3, suggests that the natural and cultural impacts of the local
environment on Tank 3 are more aggressive than those experienced by Tank 1. More importantly, as
there appears to be more tourist activity associated with Tank 3, it may be this increase in human
interference that is causing the accelerated deterioration of Tank 3.
17
Appendix A
3.3
LANDING VEHICLE TRACKED 1 – LVT1
Figure 10. Landing Vehicle Tracked 1 (LVT1) – port side view (Carpenter 2012).
Date of Inspection
21 February 2010
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). On 21 February 2012 the winds were ENE at 17 to 24 knots. Seas were
choppy with a moderate long period swell (morning NNW 2m at 10 seconds; afternoon N 2m at 10
seconds). The tides were semi-diurnal over the survey period and are reported in Table 1
(http://buoyweather.com).
The in-water visibility was approximately 10m. The depth to the top and base of the wreck at 1431 was
0.5m and 2.7m, respectively. The pH of seawater usually falls within the range of 7.5 to 8.3. The redox
potential range of marine environments is -0.300 to 0.000V in reducing environments and 0.000 to
+0.250V in oxidising environments. The pH and redox potential of the seawater on-site at 2.6m was
8.27 and 0.200V respectively, indicating a normal, open circulation oxidising marine environment. The
change in dissolved oxygen content, salinity and temperature of the water column with depth
measured on 21 February 2012 is shown in Table 6.
18
Appendix A
Figure 12. Location of the LVT(A)-4 or LVT1 in Tanapag Harbour, Saipan, CNMI (Richards 2012
after Google Earth 2012).
The wreck lies on a flat, slightly undulating sandy seabed, comprising of calcareous sediment. There
are small and larger patches of reef surrounding the wreck but none are in direct contact with the
wreck. The stern of LVT1, which faces shoreward, was partially buried in the seabed and there
appeared to have been significant scouring around the bow (seaward side) of the vessel. The rear and
lower track were entirely buried and the upper surfaces were fully exposed to the marine environment.
Stormy sea conditions could result in sand movement that is likely to affect the extent of
burial/exposure, however it is not anticipated that the entire vehicle would ever become totally buried.
The wreck was covered in brown algal forms and some sporadic secondary colonisation was evident
(e.g. tunicates, soft and hard corals, seaweed, etc).
The LVT1 is mostly intact with major structural features and a number of field expedient modifications
still evident but many other components are missing, such as armour plating across the deck, tracks
and engine room, the guns and many of the controls. The turret has collapsed into the deck space of
the wreck. These losses may have been caused by corrosion but it is more likely that they were
salvaged, possibly during the disarming and disposal process outlined by the U.S. military (McKinnon
and Carrell 2011:123-124).
There are some areas of active corrosion evident on the site, indicated by the presence of the typical
red/brown “rust” spots on the surface of the vehicle (Figure 10) but these are minimal when compared
to the tanks. This LVT 1 is included on the WWII Maritime Heritage Trail – Battle of Saipan and divers
and snorkelers are actively encouraged to visit the site provided they follow the visitation guidelines
and do not interfere with the site (i.e. disturb or attempt to remove any components).
20
Appendix A
Corrosion Survey
The corrosion parameters of twelve different areas on the LVT1 were measured over a 40 minute dive
on 21 February 2012. The results are presented in Table 7 and the on-site positions shown in Figure
13. In order to compare the corrosion data collected from the different positions measured on the LVT1
and ascertain the thermodynamically stable state of the iron, the corrosion potentials (E corr) and the pH
of the residual iron alloy surfaces were plotted on the iron Pourbaix diagram in aerobic seawater at
25°C (Figure 14). The temperature of the seawater on-site was 28°C, however this 3°C increase does
not significantly affect the nature or equilibria of the chemical species described in this diagram. No pH
and Ecorr measurements were possible when there was total penetration of the steel structure.
Table 7. Corrosion parameter measurements on the LVT1.
Position
Number
1
2
3
4
5
6
7
8
9
10
11
12
Description
stbd side, aft bow,
near seabed
stbd side, aft bow,
1m above seabed
bow,
vertical
surface
upper track, port
side, near stern
port side, aft bow,
1m above seabed
port side, aft bow,
near seabed
bow, top surface
stern, top edge,
midships
upper track, stbd
side, near stern
turret, aft towards
stern
added aft gun
shield, port side
fwd gun mount,
stbd side
pH
Ecorr vs
NHE (V)
dtotal =
depth of
concretion
+ corrosion
(mm)
16
dc = depth
of
corrosion
(mm)
Annualised
corrosion
rate
-1
(mmy )
Water
Depth
(m)
7.50
-0.321
6
0.09
2.6
total penetration
23
7
0.10
1.6
5.46
-0.320
15
2
0.03
1.8
7.56
-0.321
20
1.9
total penetration
21
1.5
7.88
-0.321
2
2.5
total penetration
total penetration
nd
24
1.0
2.1
total penetration
25
2.1
5.44
-0.325
27
1.8
5.27
-0.326
11
1.0
5.66
-0.320
15
1.0
nd = not determined
21
Appendix A
Figure 13. Schematic plan and profile views of the LVT1 indicating the corrosion parameter
measurement positions (Richards 2012 after Arnold 2010 in McKinnon and Carrell 2011:118).
(a)
Fe3+
0.8
active
0.6
FeO.OH
0.4
passive
0.2
Eh(v)
0.0
Fe2+
active
(b)
-0.2
3 & 10
-0.4
11
1
12
6
Fe3O4
4
passive
-0.6
-0.8
Fe
-1.0
immune
0
1
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 14. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the LVT1 (Richards 2012).
Generally, the areas measured on LVT1 were covered with relatively thick aerobic concretions
(average dtotal = 18 ± 7mm). There were very few secondary colonising organisms on the concretion
suggesting that there is a significant amount of water movement on this shallow site and possible
sediment impingement during periods of rough sea conditions, which would significantly reduce
colonisation rates and increase corrosion rates.
22
Appendix A
From the Pourbaix diagram (Figure 14), the intercepts of all points (1, 3, 4, 6, 10-12) measured on
LVT1, with the exception of the positions where there was total penetration of the metal (2, 5, 7-9) and
no corrosion parameters could be recorded, lie in the active corrosion region, where ferrous ions are
the thermodynamically stable chemical species and corrosion will continue until all iron is consumed.
Generally, with film free corrosion mechanisms, such as occurs on concreted iron artefacts, an
increase in the corrosion potential (tending more positive) indicates an increase in the corrosion of the
metal. However, the average corrosion potential of the seven measurement points was -0.322 ±
0.002V. This 2mV standard deviation is comparatively small and within experimental error for the
equipment and measuring procedure suggesting that the entire vessel is in electrical connection and
the same film free corrosion mechanism applies to all areas on the LVT. This very small standard
deviation also means that it is not possible to determine any differences in corrosion behaviour
between the measurement points based on the Ecorr data.
As mentioned previously, the pH is often a more reliable indicator of changes in localised corrosion
rates. The average pH value for the more acidic areas (3, 10, 11 & 12) was 5.46 ± 0.16 and for the
more alkaline positions (1, 4 & 6) was 7.65 ± 0.20. There appeared to be no obvious relationship
between water depth and the average pH values. However, the large difference in average pH values
indicates that the more acidic positions are corroding at a faster rate than the more alkaline areas. It is
not unexpected that positions 1 and 6, which were located on the hull at the seabed surface at a water
depth of 2.5m, would be corroding at a slower rate than the shallower, more exposed positions. These
lower sections of the vessel near the sediment/seawater interface would be subjected to periodic burial
cycles, which would reduce the total amount of dissolved oxygen impingement to the concretion
surface, slow the cathodic reaction (reduction of oxygen at the concretion surface) of the corrosion
cell, consequently reducing the overall corrosion rate. In support, positions 2 and 5, which were
located on the hull, 1m directly above points 1 and 6 at a water depth of 1.5m and exposed at all
times, were totally corroded indicating a higher corrosion rate. Position 4, which was located on the top
of the track on the port side near the stern at a depth of 1.9m, also had a more alkaline pH indicating a
reduced corrosion rate. However, position 9 was located in a similar position but on the starboard side
and had a dtotal of 25mm with no residual metal remaining. The dtotal for position 4 was 20mm, which
was significantly thicker than the average dtotal of the other measured positions (12 ± 6mm) with the
exception of the turret (27mm), hence, the more alkaline pH on position 4 may not represent a slower
corrosion rate but be due to the track having almost no residual metal remaining and under these
conditions there is no longer the driving force to maintain the lower pH and E corr values inside the
concretion and the solution slowly equilibrates to those values of the local environment (higher Ecorr
and more alkaline pH).
The fact that there were five positions (2, 5, 7-9) where there was no residual metal and total
penetration with the drill occurred indicates extensive corrosion in these areas. However there was no
obvious relationship between the water depths, the orientation (i.e. vertical versus horizontal) and/or
the dtotal at these locations as compared to the positions where residual metal remained (1, 3, 4, 6, 1012). Therefore it is difficult to explain the reason for the total loss of metal at these particular positions.
One possibility is that the original metal thickness in these areas was less than the other areas where
residual metal remained. If positions 1 and 6 are discounted because they are near the sediment
surface and have a generally lower corrosion rate compared to the exposed positions on the vessel
and position 4 is included in the total penetration group because it is likely that very little residual metal
remains in the track then this leaves positions 3, 10, 11 and 12 which may have had thicker original
metal thicknesses. This assumption is not unfeasible as position 3, was the vertical surface of the bow,
10 was the turret, 11 was the added protective shielding on the turret and 12 was the forward gun
mount shield, where thicker metal would be expected affording better protection for the operators of
the machine. Some published specifications for the LVT(A)-4 support this inference with the upper
front of the hull being 13mm thick steel and the rest of the vessel (i.e. middle front, lower front, sides,
upper rear, lower rear and top) being 6.4mm thick and the turret gun shield being 38mm thick with the
front, sides and rear being 25mm (http://afvdb.50megs.com/index.html).
MacLeod (1998) showed that the depth of corrosion (d c) can be used to calculate the mean corrosion
rate. The data for this relationship is obtained by measuring the depth of the corroded layer, which is
then divided by the number of years of submersion of the metal to give an average annualised
corrosion rate. This depth of corrosion is normally determined on objects where there is no or very little
23
Appendix A
concretion, the concretion has been removed or where there exists a very clear demarcation between
the concretion and the corrosion phases. Presently, there are problems with the methodology involved
in measuring depths of corrosion and the interpretation of the associated data owing to the nonuniform nature of corrosion across large ferrous alloy objects so there is a need for caution in the
interpretation of corrosion depths. However, despite these anomalies, it is well known that the average
-1
long-term corrosion rate for isolated iron in aerobic seawater is approximately 0.11 mmy (La Que
1975).
Unlike the tanks, on some positions on LVT1 it was possible to discern where the concretion layer
ceased and the corrosion layer began so the depth of corrosion (d c) could be measured (Table 7). It
should be noted that the dc measurements are not extremely accurate and the subsequent calculated
annualised corrosion rates should only be treated as approximations. If we assume that the LVT1 was
sunk in 1944 then it has been immersed for 68 years at the time of this corrosion survey. The
calculated annualised corrosion rates for positions 1, 2 and 3 are presented in Table 7 with an average
-1
corrosion rate of 0.07 ± 0.04mmy , which is in the range for the standard long-term corrosion rate for
iron in flowing aerobic seawater. So based on this average corrosion rate it is not surprising that areas
of the LVT hull structure, which may have had original metal thicknesses of about 6mm, have corroded
completely in this open circulation, oxidising marine environment.
24
Appendix A
3.4
LANDING VEHICLE TRACKED 2 – LVT2
Figure 15. Landing Vehicle Tracked 2 (LVT2) – front view (Carpenter 2012).
Date of Inspection
22 February 2010
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). On 22 February 2012 the winds were ENE at 18 to 24 knots in the morning
tending ENE at 17 to 23 knots in the afternoon. Seas were choppy with a moderate short period swell
(morning E 2.2m at 9 seconds; afternoon ENE 2.1m at 9 seconds). The tides were semi-diurnal over
the survey period and are reported in Table 1 (http://buoyweather.com).
The in-water visibility was approximately 10m. The depth to the top and base of the wreck at 1501 was
0.5 and 1.5m, respectively. The pH of seawater usually falls within the range of 7.5 to 8.3. The redox
potential range of marine environments is -0.300 to 0.000V in reducing environments and 0.000 to
+0.250V in oxidising environments. The pH and redox potential of the seawater on-site at 1.5m was
8.12 and 0.122V respectively, indicating a normal, open circulation oxidising marine environment. The
LVT2 is in close proximity to Tank 1 so the change in dissolved oxygen content, salinity and
temperature of the water column with depth measured on the Tank 1 site will be used for the LVT2 site
(Table 2, Figure 2). The same interpretation and conclusions for the environmental conditions on the
LVT2 site will be similar to those for the Tank 1 site. Basically all measurements are typical for a
shallow, near coastal, open circulation, well-mixed oxidising marine environment, where corrosion
rates are likely to be relatively high.
Wreck Site
25
Appendix A
LVT2 is located in close proximity to Tank 1, which lies on the south western side of Saipan, inside the
barrier reef about 180m off shore from Susupe Beach (Figure 3). The LVT2 lies at a depth of about
1.5m dependent on the tide (Figure 15). The model has not been positively identified but it is likely to
be similar in design to the LVT1, which is a LVT(A)-4 constructed primarily of rolled homogenous steel.
The LVT2 is orientated parallel to the shoreline with its bow pointing NNE. The vessel is fully
submerged at all times.
The surrounding seabed consists primarily of calcareous sediment with dead coral interspersed
around the site and is relatively flat with short period undulating sand ripples caused by winnowing.
Occasional living coral can be observed on the surrounding seabed but there is very limited coralline
growth on the vehicle structure itself. The wreck was densely covered in brown algal forms. High
nutrient levels in the lagoon may be contributing to this extensive algal growth (Denton et al. 2001).
The vehicle has almost totally collapsed and a track sprocket has detached and lies on the port side
close to its previously installed position (Figure 15). The tracks themselves are also detached and lie,
exposed, in close proximity to the major vessel remains. The vehicle is incomplete with the main
components, such as the engine and superstructure missing. The remains were partially buried (lower
track wheel bogies were not visible) with the starboard side of the vehicle possessing more sediment
coverage. The starboard side of the vessel is extremely damaged and the loss of structure is quite
extensive with the bow region almost absent, which facilitates sediment ingress into the interior of the
vessel. Stormy sea conditions could result in sand movement that is likely to affect the extent of
burial/exposure, however it is not anticipated that the entire vehicle would ever become totally buried.
There are many areas of active corrosion evident on the site, indicated by the presence of the typical
red/brown “rust” spots on the metal surfaces and at the sediment interface.
The poor condition and collapsed state of the LVT2 could be due to a number of human factors
including WWII but storm damage through increased wave action, in such shallow water, has very
likely contributed to its gradual destruction. Recent impact damage was noted on the higher profile part
of the structure and is probably due to a small boat collision. The LVT2 is not included on the WWII
Maritime Heritage Trail – Battle of Saipan therefore visitation to the site by divers and snorkelers would
be less than to those wrecks that are listed on the heritage trail.
Corrosion Survey
The corrosion parameters of five different areas on the LVT2 were measured over a 28 minute dive on
22 February 2012. The results are presented in Table 8 and the on-site positions shown in Figures 16
and 17. In order to compare the corrosion data collected from the different positions measured on the
LVT2 and ascertain the thermodynamically stable state of the iron, the corrosion potentials (Ecorr) and
the pH of the residual iron alloy surfaces were plotted on the iron Pourbaix diagram in aerobic
seawater at 25°C (Figure 18). The temperature of the seawater on-site was 31°C, however this 6°C
increase does not significantly affect the nature or equilibria of the chemical species described in this
diagram. No pH and Ecorr measurements were possible when there was total penetration of the steel
structure.
Table 8. Corrosion parameter measurements on the LVT2.
Position
Number
1
2
3
4
5
Description
port side, aft bow, near seabed
port side, top surface
stern midships
stbd side, aft bow, near seabed
track, near bow
pH
Ecorr vs
NHE (V)
6.25
-0.273
total penetration
7.77
-0.327
7.36
-0.286
total penetration
dtotal = depth
of concretion
+ corrosion
(mm)
13
5
3
5
7
Water
Depth
(m)
1.5
0.8
1.0
1.2
1.4
Generally, the areas measured on LVT2 were covered with relatively thin aerobic concretions (average
dtotal = 7 ± 4mm). There were very few secondary colonising organisms on the concretion suggesting
that there is a significant amount of water movement on this shallow site and possible sediment
26
Appendix A
impingement during periods of rough sea conditions, which would significantly reduce colonisation
rates and increase corrosion rates.
Figure 16. Image indicating the corrosion parameter measurement positions 1, 2 and 5 on the
LVT2 (Richards after Carpenter 2012).
27
Appendix A
Figure 17. Image indicating the corrosion parameter measurement positions 3 and 4 on the
LVT2 (Richards after Carpenter 2012).
(a)
Fe3+
0.8
active
0.6
FeO.OH
0.4
passive
0.2
Eh(v)
0.0
Fe2+
active
(b)
1
-0.2
-0.4
3
4
Fe3O4
passive
-0.6
-0.8
Fe
-1.0
immune
0
1
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 18. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the LVT2 (Richards 2012).
28
Appendix A
From the Pourbaix diagram (Figure 18), the intercepts of all points (1, 3, 4) measured on LVT2, with
the exception of the positions where there was total penetration of the metal (2, 5) and no corrosion
parameters could be recorded, lie in the active corrosion region, where ferrous ions are the
thermodynamically stable chemical species and corrosion will continue until all iron is consumed.
Generally, with film free corrosion mechanisms, such as occurs on concreted iron artefacts, an
increase in the corrosion potential (tending more positive) indicates an increase in the corrosion of the
metal. The average corrosion potential of the three measurement points was -0.295 ± 0.028V. This
relatively larger 28mV standard deviation suggests that there is some break in electrical connectivity
between the remaining hull structure and there are some statistically valid differences in corrosion
behaviour between the areas measured on the LVT2 based on the Ecorr data. The average Ecorr for
positions 1 and 4 was -0.280 ± 0.009V, which is 0.047V more positive than position 3 (-0.327V)
suggesting that these former positions located on the port and starboard hull sides are more corroded
and there is very little residual metal remaining.
The pH is often a more reliable indicator of changes in localised corrosion rates but it is difficult to
comment based on only thee measurements for the LVT2. However, the average pH and d total values
were 7.13 ± 0.79 and 7 ± 4mm, respectively. The more alkaline average pH, lower d total and the more
positive average Ecorr (-0295 ± 0.028V) coupled with the total penetration of two positions (2 and 5)
strongly suggests that the LVT2 is extensively corroded with almost no residual metal remaining in the
hull structure. Under these conditions there is no longer the driving force to maintain the lower pH and
Ecorr values inside the concretion and the solution slowly equilibrates to those values of the local
environment (more positive Ecorr and more alkaline pH). Due to the collapsed nature of the LVT2 this is
not unexpected as it is known that isolated iron artefacts and steel hull structures that have been
damaged either through natural phenomena (e.g. cyclonic activity) or human intervention (e.g. boat
collisions, explosive damage during WWII) possess significantly higher corrosion rates than those hull
structures that are relatively intact (i.e. LVT1), where the current density of the corrosion process can
be spread over a much larger surface area effectively lowering the corrosion rate (MacLeod and
Richards 2011; Richards et al. 2011).
Again, it was not possible to discern the interface between the concretion and the corrosion product
layers on the LVT2 so the depth of corrosion could not be measured, therefore it was not possible to
calculate the annualised corrosion rate for this vessel. However, if the average long-term corrosion
-1
rate for isolated iron in aerobic seawater (0.11mmy ) is used as an approximation of the corrosion rate
on the LVT2 and it is assumed that the vessel was sunk in 1944 then it can be estimated that about
7.5mm of metal could be consumed over a 68 year submersion period. If the original metal thickness
of the hull structure was 6.4mm (e.g. sides, upper rear, lower rear) then it is not surprising that very
little residual metal remains on the LVT2.
Based on the corrosion parameter measurements it is difficult to say whether the LVT2 is corroding at
a faster rate than the LVT1 as all average measurements are within their respective statistical errors.
However, considering the extent of deterioration of the LVT2 as compared to the LVT1 it would appear
that the natural and cultural impacts on the LVT2 would be greater than those experienced by the
LVT1.
29
Appendix A
3.5
DAIHATSU LANDING CRAFT 1 – DAI1
Figure 19. Daihatsu Landing Craft 1 (DAI1) – stern to bow view (Carpenter 2012).
Date of Inspection
22 February 2010
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). On 22 February 2012 the winds were ENE at 18 to 24 knots in the morning
tending ENE at 17 to 23 knots in the afternoon. Seas were choppy with a moderate short period swell
(morning E 2.2m at 9 seconds; afternoon ENE 2.1m at 9 seconds). The tides were semi-diurnal over
the survey period and are reported in Table 1 (http://buoyweather.com). There was a slight current
(<0.5 knots) running in a NNE direction.
The in-water visibility was approximately 20m. The depth to the top and base of the wreck at 1051 was
9.9 and 11.3m, respectively. The pH of seawater usually falls within the range of 7.5 to 8.3. The redox
potential range of marine environments is -0.300 to 0.000V in reducing environments and 0.000 to
+0.250V in oxidising environments. The pH and redox potential of the seawater on-site at 11.2m was
8.06 and 0.198V respectively, indicating a normal, open circulation oxidising marine environment. The
change in dissolved oxygen content, salinity and temperature of the water column with depth
measured on 22 February 2012 is shown in Table 9.
30
Appendix A
Figure 20. Change in dissolved oxygen content with increasing water depth measured on the
DAI1 and DAI2 sites (Richards 2012).
For open circulation ocean environments, there is usually a surface maximum in the dissolved oxygen
concentration. This maximum is a direct result of absorption from the atmosphere interface, increased
water movement and photosynthetic activity by plants and cyanobacteria. Typically, after this surface
maximum the dissolved oxygen concentration of the water column will decrease with increasing depth.
Factors contributing to this trend are decreasing water movement, which leads to less oxygen
exchange with the atmosphere, decreasing photosynthetic activity due to less light penetration and
increasing aerobic respiration of plankton in the photosynthetic zone. Despite the variability in the
dissolved oxygen measurements, which would indicate a quite dynamic physical environment on this
site, the overall trend was decreasing dissolved oxygen content with increasing water depth. This trend
coupled with the other physico-chemical measurements, are typical for a shallow, near coastal,
dynamic, open circulation, oxidising marine environment, where corrosion rates are likely to be
relatively high.
Wreck Site
The Daihatsu Landing Craft 1 (DAI1) is located on the south western side of Saipan, inside Tanapag
Harbour (GPS
at a depth of about 11m (Figure 21). The wreck was
positively identified as a Daihatsu or 14m Japanese Landing Craft (McKinnon and Carrell 2011:93),
constructed primarily of welded steel and the dimensions are 14.58m in length, 3.35m in width with a
0.76m draught (http://pwencycl.kgbudge.com/D/a/Daihatsu class.htm). The vessel is fully submerged
at all times.
The wreck lies on a flat, slightly undulating sandy seabed, comprising of calcareous sediment. The
surrounding seabed is relatively devoid of marine biota. The wreck was covered in brown algal forms
and some sporadic secondary colonisation was evident (e.g. tunicates, soft and hard corals, seaweed,
etc). Isolated hard coralline growths have formed in places with one large coral formation on the upper
stern deck (Figure 19).
It appears that the wreck is not subjected to regular burial/exposure cycles. Sediment has partly
covered the lower profile areas but the establishment of some hard corals indicates that significant
accumulation of sediment does not readily occur. The site lies in the comparatively protected lagoon
area but the site has the potential to be affected by storm conditions.
The vessel is upright on the seabed and is reasonably intact except for the port side midships region
which has collapsed. A deck winch with the wire remains in position on the upper stern deck level. The
steering wheel has been displaced and is also lying on this upper deck level. Hull plates are holed in
many places and the armour shield lies on the seabed on the port side of the wheel house. The vessel
is partially buried around the stern and the upper half of the rudder is visible. Limited burial has
occurred with sand accumulation in the loading zone. The maximum height of the main structure rises
approximately 2-3m above the seabed.
There are some areas of active corrosion evident on the site, indicated by the presence of the typical
red/brown “rust” spots on the surface of the vehicle. This DAI1 is included on the WWII Maritime
Heritage Trail – Battle of Saipan and divers are actively encouraged to visit the site provided they
follow the visitation guidelines and do not interfere with the site (i.e. disturb or attempt to remove any
components).
Corrosion Survey
The corrosion parameters of ten different areas on the DAI1 were measured over a 56 minute dive on
22 February 2012. The results are presented in Table 10 and the on-site positions shown in Figure 22.
In order to compare the corrosion data collected from the different positions measured on the DAI1
and ascertain the thermodynamically stable state of the iron, the corrosion potentials (E corr) and the pH
of the residual iron alloy surfaces were plotted on the iron Pourbaix diagram in aerobic seawater at
25°C (Figure 23). The temperature of the seawater on-site was 28°C, however this 3°C increase does
not significantly affect the nature or equilibria of the chemical species described in this diagram. The
32
Appendix A
Ecorr measurement for position 5 on the outer surface of the starboard hull structure was unstable
indicating that electrical connection was not possible and there was no residual metal remaining in that
particular area.
Figure 21. Location of the Landing Vehicle Tracked 1 (LVT1 or LVT(A)-4), Daihatsu Landing
Craft 1 and 2 (DAI1 and DAI2), Japanese Freighter (Freighter) and the possible Auxiliary
Submarine Chaser (Sub Chaser) in Tanapag Harbour, Saipan, CNMI (Richards 2012 after
Google Earth 2012).
Table 10. Corrosion parameter measurements on the DAI1.
Position
Number
1
2
3
4
5
6
7
8
9
10
ns – not stable
Description
bow ramp midships, top edge
bow ramp midships, inner surface
port hull, outer surface
port hull, gunwale directly above 3
stbd hull, outer surface
stbd hull, gunwale directly above 5
stern midships, base, near drive shaft
stern midships, gunwale, top deck
windlass
armour shield
pH
Ecorr vs
NHE (V)
7.24
7.07
7.96
7.78
7.85
6.12
5.87
7.86
6.51
7.97
-0.333
-0.333
-0.344
-0.347
ns
-0.335
-0.317
-0.330
-0.325
-0.341
dtotal = depth of
concretion +
corrosion
(mm)
5
2
16
7
23
3
14
2
10
3
Water
Depth
(m)
10.1
10.7
11.3
11.3
10.9
10.6
11.4
9.9
9.9
11.3
33
Appendix A
Figure 22. Schematic plan view of the DAI1 indicating the corrosion parameter measurement
positions (Richards 2012 after McAllister and Yamafume 2011 in McKinnon and Carrell
2011:94).
(a)
Fe3+
0.8
active
0.6
FeO.OH
0.4
passive
0.2
Eh(v)
0.0
Fe2+
active
(b)
-0.2
9 1
7
-0.4
6
2
-0.6
8
10
3
Fe3O4
4
passive
-0.8
Fe
-1.0
immune
0
1
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 23. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the DAI1 (Richards 2012).
Generally, most areas measured on the DAI1 were covered with relatively thin aerobic concretions
(average 9 ± 7mm). There were hard corals and more varieties of colonising biota on this wreck
compared to the shallower wrecks, such as the LVTs and the tanks, however, the distribution was
sporadic suggesting there is still a significant amount of water movement on this site despite the
increase in water depth.
From the Pourbaix diagram (Figure 23), the intercepts of points 1, 2, 4, 6-9 measured on DAI1, lie in
the active corrosion region, where ferrous ions are the thermodynamically stable chemical species and
corrosion will continue until all iron is consumed. Positions 3 and 10 lie on the equilibrium line between
34
Appendix A
active corrosion and the passive region, which implies that the typical aerobic corrosion mechanism
2+
where the major stable chemical species is the ferrous ion (Fe ) is in equilibrium with the formation of
an insoluble corrosion product layer of magnetite (Fe 3O4). This is a very common corrosion state for
large steel ships where a large proportion of the vessel remains are still in electrical contact. No stable
voltage could be obtained for position 5 (starboard hull structure, outer surface), which indicates that
there was no residual metal remaining in that particular area. Generally, with film free corrosion
mechanisms, such as occurs on concreted iron artefacts, an increase in the corrosion potential
(tending more positive) indicates an increase in the corrosion of the metal. The average corrosion
potential of the nine measurement points was -0.334 ± 0.009V. This 9mV standard deviation is
comparatively small and within experimental error for the equipment and measuring procedure
suggesting that the entire vessel is in electrical connection and the same film free corrosion
mechanism applies to all areas on the DAI1. This very small standard deviation also means that it is
not possible to determine any differences in corrosion behaviour between the measurement points
based on the Ecorr data.
The pH is often a more reliable indicator of changes in localised corrosion rates. The average pH value
for the more acidic areas (6, 7 & 9) was 6.17 ± 0.32 and for the more alkaline positions (1 – 5, 8 & 10)
was 7.68 ± 0.36. There appeared to be no obvious relationship between water depth and the average
pH values as the difference in water depth from the shallowest to the deepest measurement positions
was only 1.5m. However, the large difference between the average pH values (1.51) indicates that the
more acidic positions are corroding at a faster rate than the more alkaline areas. The greater corrosion
rate for position 7, which is located on the base of the stern section, could be explained by the fact that
there may be some form of galvanic coupling occurring between the lower stern area and the engine
remains, the propeller shaft and the propellers, which will have different metal compositions (i.e.
copper alloy propellers). This form of galvanic corrosion would cause the stern section to corrode
preferentially to the other electrically connected parts, hereby increasing the corrosion rate in that
area. Isolated iron features tend to possess higher corrosion rates than those experienced on large
iron vessel remains, where the current density is dispersed over a much larger surface area. This
would explain the increase in the corrosion rate of the windlass, position 9. More difficult to explain is
the increase in the corrosion rate of position 6, which was on the starboard gunwale, midway along the
length of the wreck. However, position 5, which was directly below position 6, on the starboard hull
was totally corroded with no residual metal remaining. This particular area on the vessel stands upright
approximately 1.5m above the sediment level and is not protected by any other sections of the wreck.
Hence, the starboard side, as compared to the port side which lies almost level with the seabed and
the bow section which is protect inside the hull structure, would be subjected to the full force of the
current, which tends to run north-south across the wreck, increasing the amount of oxygen
impingement to the concreted metal surface, subsequently increasing the corrosion rate on the
starboard side. Most of the more alkaline areas (1 - 4 & 8) were measured on the hull remains, which
are in electrical connection, dispersing the current density over the entire hull and lowering the overall
corrosion rate.
It has been observed that the pH of corroding residual metal surfaces decrease linearly with increasing
total thickness of the corrosion product layer and the encapsulating concretion (MacLeod and Richards
2011). That is, generally the thicker the concretion, the lower the surface pH but only if the concretion
layer remains essentially undisturbed (i.e. no damage occurs through human/natural interference).
Generally, this relationship applies to this wreck, where the most corroded positions (7, 9 and 5)
possessed the thickest concretion and corrosion product layers (dtotal), which is in agreement with the
conclusions regarding corrosion rate differences based on the pH measurements.
It was not possible to discern the interface between the concretion and the corrosion product layers on
DAI1 so the depth of corrosion could not be measured, therefore it was not possible to calculate the
annualised corrosion rate for this wreck.
35
Appendix A
3.6
DAIHATSU LANDING CRAFT 2 – DAI2
Figure 24. Daihatsu Landing Craft 2 (DAI2) – port side view (Carpenter 2012).
Date of Inspection
22 February 2010
Environmental Conditions
The wreck site of the Daihatsu Landing Craft 2 (DAI2) lies approximately 45m to the southwest of DAI1
and the survey was carried out during the same dive as DAI1. Hence, the environmental conditions
were the same as for DAI1 (see Section 3.5). The depth to the top and base of the wreck at 1051 was
11.4 and 12.3m, respectively. The results are typical for a shallow, near coastal, open circulation,
oxidising marine environment, where corrosion rates are likely to be relatively high.
Wreck Site
The Daihatsu Landing Craft 2 (DAI2) is located about 45m southwest of DAI1, on the south western
side of Saipan, inside Tanapag Harbour (GPS
) at a depth of about
11m (Figure 21). The wreck was positively identified as a Daihatsu or 14m Japanese Landing Craft
(McKinnon and Carrell 2011:98), constructed primarily of welded steel and the dimensions are 14.58m
in length, 3.35m in width with a 0.76m draught (http://pwencycl.kgbudge.com/D/a/Daihatsu class.htm).
The vessel is fully submerged at all times.
The wreck lies on a flat, slightly undulating sandy seabed, comprising of calcareous sediment. The
surrounding seabed is relatively devoid of marine biota. The wreck was covered in brown algal forms
and some sporadic secondary colonisation was evident (e.g. tunicates, soft and hard corals, seaweed,
etc). Isolated hard coralline growths have formed in some places (Figure 24).
36
Appendix A
It appears that the wreck is not subjected to regular burial/exposure cycles. Sediment has partly
covered the lower profile areas but the establishment of some hard corals indicates that significant
accumulation of sediment does not readily occur. The site lies in the comparatively protected lagoon
area but the site has the potential to be affected by storm conditions. Limited burial has occurred with
sand accumulation in the loading zone and around the lower profile areas. The maximum height of the
main structure rises approximately 2-3m above the seabed.
The vessel is upright on the seabed and is in poor structural condition. DAI2 is considerably more
disarticulated than DAI1. Large sections lie separate, astern of the wreck and substantial remains have
collapsed on the port side of the vessel. The engine is missing and presumed salvaged but the rudder
and propeller shaft remain in situ.
There are some areas of active corrosion evident on the site, indicated by the presence of the typical
red/brown “rust” spots on the surface of the vehicle. This DAI2 is included on the WWII Maritime
Heritage Trail – Battle of Saipan and divers are actively encouraged to visit the site provided they
follow the visitation guidelines and do not interfere with the site (i.e. disturb or attempt to remove any
components).
Corrosion Survey
The corrosion parameters of five different areas on the DAI2 were measured over a 56 minute dive on
22 February 2012. The results are presented in Table 11 and the on-site positions shown in Figure 25.
In order to compare the corrosion data collected from the different positions measured on the DAI2
and ascertain the thermodynamically stable state of the iron, the corrosion potentials (E corr) and the pH
of the residual iron alloy surfaces were plotted on the iron Pourbaix diagram in aerobic seawater at
25°C (Figure 26). The temperature of the seawater on-site was 28°C, however this 3°C increase does
not significantly affect the nature or equilibria of the chemical species described in this diagram. Total
penetration occurred at position 1 on the base of the stern structure indicating that there was no
residual metal remaining in that particular area.
Table 11. Corrosion parameter measurements on the DAI2.
Position
Number
1
2
3
4
5
Description
stern midships, base, near drive shaft
stern midships, top deck
stbd hull, gunwale
port hull, gunwale
bow ramp midships, top edge
pH
Ecorr vs
NHE (V)
total penetration
6.09
-0.324
6.09
-0.327
6.72
-0.323
6.65
-0.327
dtotal = depth of
concretion +
corrosion
(mm)
15
3
9
28
4
Water
Depth
(m)
12.3
11.4
11.7
11.4
11.5
37
Appendix A
Figure 25. Image indicating the corrosion parameter measurement positions on the DAI2
(Richards after Carpenter 2012).
(a)
Fe3+
0.8
active
0.6
FeO.OH
0.4
passive
0.2
Eh(v)
0.0
Fe2+
active
(b)
-0.2
2&3
-0.4
5
Fe3O4
4
passive
-0.6
-0.8
Fe
-1.0
immune
0
1
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 26. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the DAI2 (Richards 2012).
Some areas measured on the DAI2 were covered with relatively thin aerobic concretions (average 2, 3
& 5 = 5 ± 3mm) while positions 1 and 4 had thicker concretions with an average dtotal of 22 ± 9mm.
There were hard corals and more varieties of colonising biota on this wreck compared to the shallower
38
Appendix A
wrecks, such as the LVTs and the tanks, however, the distribution was sporadic suggesting there is
still a significant amount of water movement on this site despite the increase in water depth.
From the Pourbaix diagram (Figure 26), the intercepts of points 2 - 5 measured on DAI2, with the
exception of position 1 where there was total penetration of the metal and no corrosion parameters
could be recorded, lie in the active corrosion region, where ferrous ions are the thermodynamically
stable chemical species and corrosion will continue until all iron is consumed. Generally, with film free
corrosion mechanisms, such as occurs on concreted iron artefacts, an increase in the corrosion
potential (tending more positive) indicates an increase in the corrosion of the metal. The average
corrosion potential of the four measurement points was -0.325 ± 0.002V. This 2mV standard deviation
is comparatively small and within experimental error for the equipment and measuring procedure
suggesting that the entire vessel is in electrical connection and the same film free corrosion
mechanism applies to all areas on the DAI2. This very small standard deviation also means that it is
not possible to determine any differences in corrosion behaviour between the measurement points
based on the Ecorr data.
The pH is often a more reliable indicator of changes in localised corrosion rates. The average pH value
for the more acidic areas (2 & 3) was 6.09 ± 0.00 and for the more alkaline positions (4 & 5) was 6.68
± 0.05. There appeared to be no obvious relationship between water depth and the average pH values
as the difference in water depth from the shallowest to the deepest measurement positions was only
1.0m. However, the difference between the average pH values of 0.59 indicates that the more acidic
positions are corroding at a faster rate than the more alkaline areas. The greater corrosion rate for
position 2 (top deck of the stern section) and the total corrosion of position 1 (base of the stern section
near the rudder) could be explained by the fact that there may be some form of galvanic coupling
occurring between the lower stern area and the propeller shaft, rudder, etc, which will have different
metal compositions. This form of galvanic corrosion would cause the stern section to corrode
preferentially to the other electrically connected parts, hereby increasing the corrosion rate in that
area. The increase in the corrosion rate of position 3 (starboard gunwale) may be due to its higher
profile above the seabed, where dissolved oxygen impingement to the concreted gunwale surface will
be greater than on more protected, lower profile areas, such as on the bow ramp (5) and port gunwale
(4). However, the fact that the dtotal for position 4 (port gunwale) was 23mm, the thickest of all areas
measured during this survey tends to suggest that the more alkaline pH for position 4 may not
represent a slower corrosion rate but be due to the port gunwale having almost no residual metal
remaining and under these conditions there is no longer the driving force to maintain the lower pH and
Ecorr values inside the concretion and the solution slowly equilibrates to those values of the local
environment (higher Ecorr and more alkaline pH).
Again, it was not possible to discern the interface between the concretion and the corrosion product
layers on DAI2 so the depth of corrosion could not be measured, therefore it was not possible to
calculate the annualised corrosion rate for this wreck.
Based on the corrosion parameter measurements it is difficult to say whether the DAI2 is corroding at
a faster rate than the DAI1 as most average measurements are within their respective statistical
errors, with the exception of the average pH values of the more alkaline positions, which were 6.68 ±
0.05 on DAI2 and 7.68 ± 0.36 on DAI1. This decrease in average pH of DAI2 suggests that it may be
corroding at a slightly faster rate than DAI1. This is not unexpected as it is known that isolated iron
artefacts and steel hull structures that have been damaged either through natural phenomena (e.g.
cyclonic activity) or human intervention (e.g. salvage, explosive damage during WWII) possess higher
corrosion rates than those hull structures that are relatively intact (i.e. DAI1), where the current density
of the corrosion process can be spread over a much larger surface area effectively lowering the
corrosion rate (MacLeod and Richards 2011; Richards et al. 2011).
39
Appendix A
3.7
DAIHATSU LANDING CRAFT 3 – DAI3
Figure 27. Daihatsu Landing Craft 3 (DAI3) – starboard side view (Carpenter 2012).
Date of Inspection
23 February 2010
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). On 23 February 2012 the winds were ENE at 15 to 21 knots in the morning
tending E in the afternoon. Seas were moderately choppy with small short period wind waves (E 1.8m
at 9 seconds). The tides were semi-diurnal over the survey period and are reported in Table 1
(http://buoyweather.com). There was a slight current (<0.5 knots) running in a NNE direction.
The in-water visibility was approximately 20m. The depth to the top and base of the wreck at 1141 was
4.2 and 6.6m, respectively. The pH of seawater usually falls within the range of 7.5 to 8.3. The redox
potential range of marine environments is -0.300 to 0.000V in reducing environments and 0.000 to
+0.250V in oxidising environments. The pH and redox potential of the seawater on-site at 6.4m was
8.10 and 0.166V respectively, indicating a normal, open circulation oxidising marine environment. The
change in dissolved oxygen content, salinity and temperature of the water column with depth
measured on 23 February 2012 is shown in Table 12.
There was no significant change in salinity and temperature with increasing water depth, which is
typical of the hydrology of well mixed near coastal marine waters. The average water temperature was
27.8 ± 0.1°C and the average salinity of the water column was 35.8 ± 0.5ppK, which is within the usual
salinity range for the open ocean of 32-37ppK. The average dissolved oxygen content was 6.11 ±
0.12ppm. The change in dissolved oxygen concentration with increasing water depth is shown in
Figure 28.
40
Appendix A
41
Appendix A
The Daihatsu Landing Craft 3 (DAI3) is located on the south western side of Saipan, inside Tanapag
Harbour at a depth of about 7m. The wreck has not been positively identified but appears to be very
similar to DAI1 and DAI2 which are Daihatsu or 14m Japanese Landing Craft, constructed primarily of
welded steel and the dimensions are 14.58m in length, 3.35m in width with a 0.76m draught
(http://pwencycl.kgbudge.com/D/a/Daihatsu class.htm). The vessel is fully submerged at all times.
The wreck lies on a flat, slightly undulating sandy seabed, comprising of calcareous sediment. The
surrounding seabed is relatively devoid of marine biota with some isolated coral outcrops in close
proximity. The wreck was covered in concretion, some brown algal forms and some secondary
colonisation was evident (e.g. tunicates, soft and hard corals, etc) especially in the more protected
areas (i.e in the engine room, under the stern). Isolated hard coralline growths have formed in places
with one large coral formation on the upper stern deck adjacent to the windlass (Figure 27). It appears
that the wreck is not subjected to regular burial/exposure cycles. Sediment has partly covered the
lower profile areas and the loading area but the establishment of hard and soft corals indicates that
significant accumulation of sediment does not readily occur.
The vessel is upright on the seabed and is reasonably intact. The maximum height of the main
structure rises approximately 2m above the seabed. Some scouring has occurred around the stern
and the rudder and propeller are exposed. Many hull plates are either missing or considerably
corroded with significant areas of loss. Most of the port and starboard side hull structure in the
midships area has collapsed. The engine room of this vehicle is visible through perforated hull plates.
The engine remains in situ and is complete with the exhaust system and other associated features and
ancillary equipment. For example, a funnel is located in the starboard forward corner of the engine
room.
There are some areas of active corrosion evident on the site, especially in the shallower, more
exposed areas (i.e. upper surfaces of the stern section) indicated by the presence of the typical
red/brown “rust” spots on the surface of the vehicle. This DAI3 is not included on the WWII Maritime
Heritage Trail – Battle of Saipan therefore visitation to the site by divers and snorkelers would be less
than to those wrecks that are listed on the heritage trail. The fact that the engine, other associated
machinery and some artefacts (e.g. funnel) are still present in situ supports the fact that this site is not
visited frequently and human interference has been minimal to date.
Corrosion Survey
The corrosion parameters of eleven different areas on the DAI3 were measured over a 61 minute dive
on 23 February 2012. The results are presented in Table 13 and the on-site positions shown in Figure
29. In order to compare the corrosion data collected from the different positions measured on the DAI3
and ascertain the thermodynamically stable state of the iron, the corrosion potentials (E corr) and the pH
of the residual iron alloy surfaces were plotted on the iron Pourbaix diagram in aerobic seawater at
25°C (Figure 30). The temperature of the seawater on-site was 28°C, however this 3°C increase does
not significantly affect the nature or equilibria of the chemical species described in this diagram.
Table 13. Corrosion parameter measurements on the DAI3.
Position
Number
1
2
3
4
5
6
7
8
Description
rudder
propeller
stern, upper deck, midships
windlass
stern, upper deck, fwd windlass, midships
stbd support beam, midships
port support beam, midships
bow ramp midships, inner surface
pH
Ecorr vs
NHE
(V)
7.00
-0.339
6.28
-0.340
7.36
-0.338
7.90
-0.337
total penetration
7.10
-0.338
7.19
-0.338
7.97
-0.337
dtotal =
depth of
concretion
+
corrosion
(mm)
15
3
6
4
7
5
7
2
Water
Depth
(m)
6.0
6.6
5.2
4.2
5.1
5.6
5.8
6.4
43
Appendix A
9
10
11
stbd gunwale, bow area
port gunwale, bow area
bow ramp midships, top edge
7.95
6.54
6.20
-0.337
-0.338
-0.340
2
4
14
5.5
5.7
5.9
nd = not determined
Figure 29. Image indicating the corrosion parameter measurement positions on the DAI3
(Richards after Carpenter 2012).
44
Appendix A
(a)
Fe3+
0.8
active
0.6
FeO.OH
0.4
passive
0.2
Eh(v)
0.0
Fe2+
active
(b)
2
-0.2
11
-0.4
6 3
10 1
7
-0.6
8&9
4
Fe3O4
passive
-0.8
Fe
-1.0
immune
0
1
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 30. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the DAI3 (Richards 2012).
Generally, the areas measured on the DAI3 were covered with relatively thin aerobic concretions
(average 6 ± 4mm). There were more hard corals and more varieties of colonising biota on this wreck
compared to the shallower wrecks, such as the LVTs and the tanks, which are subjected to more
natural and human interference and the other two Daihatsu landing crafts (DAI1 and DAI2), which
were at a depth of about 11m almost twice that of DAI3. This suggests that the DAI3 is subjected to a
less aggressive environment however, the distribution of marine biota on the upper surfaces was still
sporadic suggesting there is still a significant amount of water movement on this site.
From the Pourbaix diagram (Figure 30), the intercepts of points 1 - 3, 6, 7, 10 & 11 measured on DAI3,
lie in the active corrosion region, where ferrous ions are the thermodynamically stable chemical
species and corrosion will continue until all iron is consumed. Positions 4, 8 & 9 lie on the equilibrium
line between active corrosion and the passive region, which implies that the typical aerobic corrosion
2+
mechanism where the major stable chemical species is the ferrous ion (Fe ) is in equilibrium with the
formation of an insoluble corrosion product layer of magnetite (Fe 3O4). This is a very common
corrosion state for large steel ships where a large proportion of the vessel remains are still in electrical
contact. Total penetration of the metal occurred on position 5 indicating that there was no residual
metal remaining in that particular area. Generally, with film free corrosion mechanisms, such as occurs
on concreted iron artefacts, an increase in the corrosion potential (tending more positive) indicates an
increase in the corrosion of the metal. The average corrosion potential of the nine measurement points
was -0.338 ± 0.001V. This 1mV standard deviation is comparatively small and within experimental
error for the equipment and measuring procedure suggesting that the entire vessel is in electrical
connection and the same film free corrosion mechanism applies to all areas on the DAI3. This very
small standard deviation also means that it is not possible to determine any differences in corrosion
behaviour between the measurement points based on the Ecorr data.
The pH is often a more reliable indicator of changes in localised corrosion rates. The average pH value
for the more acidic areas (2, 10 & 11) was 6.34 ± 0.18 and for the more alkaline positions (1, 3, 4, 6 9) was 7.50 ± 0.42. There appeared to be no obvious relationship between water depth and the
average pH values as the difference in water depth from the shallowest to the deepest measurement
positions was only 1.5m. However, the large difference between the average pH values (1.16)
indicates that the more acidic positions are corroding at a faster rate than the more alkaline areas. The
greater corrosion rate for position 2, which was the propeller, could be explained by the fact that the
45
Appendix A
propeller would probably possess a different metal composition and this particular Pourbaix diagram is
not applicable or there may be some form of galvanic coupling occurring between the propeller and the
engine remains. This form of galvanic corrosion would cause the propeller to corrode preferentially to
the other electrically connected parts, hereby increasing the corrosion rate in that area. More difficult to
explain is the increase in the corrosion rate of position 10, which was on the port gunwale, near the
bow and position 11 which was on the upper edge of the bow ramp as these areas appear to have
been in direct electrical contact with the rest of the vessel remains. Most of the more alkaline areas (1,
3, 4, 6 - 9) were measured on the hull remains, which are in electrical connection, dispersing the
current density over the entire hull and lowering the overall corrosion rate.
It has been observed that the pH of corroding residual metal surfaces decrease linearly with increasing
total thickness of the corrosion product layer and the encapsulating concretion (MacLeod and Richards
2011). That is, generally the thicker the concretion, the lower the surface pH but only if the concretion
layer remains essentially undisturbed (i.e. no damage occurs through human/natural interference).
Generally, this relationship applies to this wreck, where the most corroded positions (7, 9 and 5)
possessed the thickest concretion and corrosion product layers (d total), which is in agreement with the
conclusions regarding corrosion rate differences based on the pH measurements.
It was not possible to discern the interface between the concretion and the corrosion product layers on
DAI3 so the depth of corrosion could not be measured, therefore it was not possible to calculate the
annualised corrosion rate for this wreck.
Based on the corrosion parameter measurements it is difficult to say which of the Daihatsu landing
craft is corroding at a faster rate as most average measurements are within their respective statistical
errors, with the exception of the average pH values of the more alkaline positions, which were 6.68 ±
0.05 on DAI2 and 7.68 ± 0.36 and 7.50 ± 0.42 on DAI1 and DAI3, respectively. This decrease in
average pH of DAI2 suggests that it may be corroding at a slightly faster rate than both DAI1 and
DAI3. This is not unexpected as it is known that isolated iron artefacts and steel hull structures that
have been damaged either through natural phenomena (e.g. cyclonic activity) or human intervention
(e.g. salvage, explosive damage during WWII) possess higher corrosion rates than those hull
structures that are relatively intact (i.e. DAI1 and DAI3), where the current density of the corrosion
process can be spread over a much larger surface area effectively lowering the corrosion rate
(MacLeod and Richards 2011; Richards et al. 2011). In addition, it appears that DAI1 and DAI3 are
corroding at relatively similar rates, despite the fact that DAI3 is a much shallower site, where it would
be expected that the corrosion rate would be slightly higher. This would seem to suggest that human
interference (i.e. recreational diving activities) is having some impact on the deterioration rate of the
deeper DAI1 site.
3.8
JAPANESE FREIGHTER - JFR
46
Appendix A
Figure 31. Japanese Freighter (JFR) – bow view (Carpenter 2012).
Date of Inspection
21 February 2010
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). On 21 February 2012 the winds were ENE at 17 to 24 knots. Seas were
choppy with a moderate long period swell (morning NNW 2m at 10 seconds; afternoon N 2m at 10
seconds). The tides were semi-diurnal over the survey period and are reported in Table 1
(http://buoyweather.com). There was quite a strong current (>1 knot) running in a NNE direction.
The in-water visibility was approximately 20m. On the day of the survey the highest part of the wreck
was breaking the seawater surface and the depth to the base of the wreck at 1010 was 11.2m (Figure
31). The pH of seawater usually falls within the range of 7.5 to 8.3. The redox potential range of
marine environments is -0.300 to 0.000V in reducing environments and 0.000 to +0.250V in oxidising
environments. The pH and redox potential of the seawater on-site at 8.8m was 8.20 and 0.210V
respectively, indicating a normal, open circulation oxidising marine environment. The change in
dissolved oxygen content, salinity and temperature of the water column with depth measured on 21
February 2012 is shown in Table 14.
47
Appendix A
maximum the dissolved oxygen concentration of the water column will decrease with increasing depth.
Factors contributing to this trend are decreasing water movement, which leads to less oxygen
exchange with the atmosphere, decreasing photosynthetic activity due to less light penetration and
increasing aerobic respiration of plankton in the photosynthetic zone. The overall trend on the JFR site
was a relatively steady decrease in dissolved oxygen content with increasing water depth. This trend
coupled with the other physico-chemical measurements, are typical for a shallow, near coastal, open
circulation, oxidising marine environment, where corrosion rates are likely to be relatively high.
Wreck Site
The Japanese Freighter (JFR) is located on the south western side of Saipan, inside Tanapag Harbour
(GPS
) at a maximum depth of about 11m (Figure 21). The wreck has
not been positively identified but may be the Shoan Maru, a steamer of 5624 tons built primarily of
steel in 1937 and later requisitioned for use during WWII (McKinnon and Carrell 2011:40-44). Some of
the higher profile sections of the vessel are exposed to the atmosphere at low tides but the majority of
the wreck is fully submerged at all times.
The wreck lies on a flat, slightly undulating sandy seabed, comprising of calcareous sediment. The
surrounding seabed is interspersed with coral outcrops, especially around the bow area. The vessel is
not heavily concreted (e.g. welded overlapping hull plates are discernable) and it has a general layer
of encrustation which may be derived from calcareous and other forms of algae. Patches of low profile
hard coral formations and larger hard corals have become established particularly on the side of the
port bow which is angle towards the sea surface (Figure 31). Isolated coral out-growths exist on other
parts of the vessel but it is minimal. Coral fish predominate and a very large stonefish was noted.
Pelagic fish are prevalent around this site in contrast to the other sites investigated. As with other sites
in the lagoon freshwater run-off and associated pollution may be influencing marine growth.
It appears that the wreck is not subjected to regular burial/exposure cycles. Limited sediment coverage
has occurred on some hull structure lying on the seabed but due to the relatively high profile of the
shipwreck remains, this negates extensive burial. The establishment of hard corals indicates that
significant accumulation of sediment does not readily occur. Maximum exposure height of the main
structure is approximately 8m. Curved hull structure lying close to the seabed is undercut and free of
sand accretion. Relatively strong currents are experienced on this site and scouring under ship
structure is a likely consequence. The site lies in the comparatively protected lagoon area but due to
the extensive profile of the vessel above the seabed it is conceivable that storms would have an
impact on the ship’s structure.
The vessel lies on its starboard side and although most of the wreck is disarticulated and has
collapsed in many areas, major elements such as the engines, boilers, steering mechanism and
superstructure are generally located in close proximity to their original position. No discernible cargo
was observed. The ship was apparently torpedoed in 1943 but did not sink and was towed to Saipan
for repairs or salvage. In 1944 it was attacked again by aircraft and damaged beyond repair. During
the post-war clean up of Tanapag harbour the ship was heavily salvaged and cut-down to the
waterline because it was considered a navigational hazard. There are also reports that it was used for
explosives training during this time (McKinnon and Carrell 2011:38-40).
There are a few areas of active corrosion evident on the site, indicated by the presence of the typical
red/brown “rust” spots on the surface of the vessel but these are sporadic compared to the size of the
wreck remains. This JFR is included on the WWII Maritime Heritage Trail – Battle of Saipan and divers
are actively encouraged to visit the site provided they follow the visitation guidelines and do not
interfere with the site (i.e. disturb or attempt to remove any components).
Corrosion Survey
The corrosion parameters of twenty four different areas on the JFR were measured over two dives (80
and 61 minutes) 21 February 2012. The results are presented in Table 15 but due to the large
distribution area of the wreck the on-site positions are not shown. In order to compare the corrosion
data collected from the different positions measured on the JFR and ascertain the thermodynamically
stable state of the iron, the corrosion potentials (Ecorr) and the pH of the residual iron alloy surfaces
49
Appendix A
were plotted on the iron Pourbaix diagram in aerobic seawater at 25°C (Figure 33). The temperature of
the seawater on-site was 28°C, however this 3°C increase does not significantly affect the nature or
equilibria of the chemical species described in this diagram.
50
Appendix A
Table 15. Corrosion parameter measurements on the JFR.
Position
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
1
Description
rudder
1
stbd hull 1 , stern
1
stbd deck 1 , stern
stbd hull 2
stbd deck 2
bollard
stbd hull 3, vertical plate
stbd deck 3
stbd hull 4
stbd deck 4
stbd hull 5, under bow
stbd hull 5, vertical bow plate
port hull 1, stern
port deck 1, stern
port hull 2
port deck 2
engine block
port hull 3
port deck 3
boiler
port hull 4
port deck 4
port hull 5, bow
port deck 5, bow
pH
Ecorr vs
NHE (V)
7.62
7.28
6.76
7.76
6.95
6.87
5.48
6.75
7.94
7.67
8.11
6.81
7.93
7.96
7.86
6.38
7.89
8.14
6.47
7.89
7.45
7.01
6.71
7.21
-0.343
-0.340
-0.342
-0.338
-0.329
-0.330
-0.344
-0.334
-0.347
-0.346
-0.342
-0.345
-0.338
-0.338
-0.337
-0.330
-0.256
-0.344
-0.330
-0.330
-0.340
-0.330
-0.340
-0.356
dtotal =
depth of
concretion
+
corrosion
(mm)
8
2
9
4
8
6
13
9
2
2
4
7
1
2
1
16
1
1
5
5
2
2
3
4
Water
Depth
(m)
7.9
6.7
5.2
6.2
5.9
5.8
8.1
5.9
8.0
7.0
8.0
7.8
5.6
5.1
5.7
5.5
3.6
7.5
7.0
7.2
7.1
6.3
8.1
5.0
stbd hull and deck 1-5; port hull and deck 1-5 = traversing from stern to bow
(a)
Fe3+
0.8
active
0.6
FeO.OH
0.4
passive
0.2
Eh(v)
0.0
Fe2+
active
(b)
3,8,12,23
-0.2
7
16
-0.6
1,10
17
4
-0.4
21
5 2
19
6 22
24
11,18
Fe3O4
9,13,14,15,20
passive
-0.8
Fe
-1.0
immune
0
1
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 33. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the JFR site (Richards 2012).
51
Appendix A
Generally, the areas measured on the JFR were covered with relatively thin aerobic concretions
(average 5 ± 4mm). There were more hard corals on this wreck compared to the shallower wrecks,
such as the LVTs and the tanks, which are subjected to more natural and human interference and the
other three Daihatsu landing crafts (DAI1, DAI2 and DAI3). However, this is probably a reflection of the
larger surface area available for colonisation than an effect of changes in the local environment.
From the Pourbaix diagram (Figure 33), the intercepts of points 1-8, 10, 12, 16, 19, 21-24) measured
on the JFR, lie in the active corrosion region, where ferrous ions are the thermodynamically stable
chemical species and corrosion will continue until all iron is consumed. Positions 9, 13-15, 20 lie on
the equilibrium line between active corrosion and the passive region, which implies that the typical
2+
aerobic corrosion mechanism where the major stable chemical species is the ferrous ion (Fe ) is in
equilibrium with the formation of an insoluble corrosion product layer of magnetite (Fe 3O4). This is a
very common corrosion state for large steel ships where a large proportion of the vessel remains are
still in electrical contact. Positions 11, 17 and 18, lie in the passive magnetite region (Fe3O4) indicating
there is very little if any residual metal remaining in these areas. Generally, with film free corrosion
mechanisms, such as occurs on concreted iron artefacts, an increase in the corrosion potential
(tending more positive) indicates an increase in the corrosion of the metal. Position 17, which was the
engine block had a more positive corrosion potential of -0.256V indicating an increase in the corrosion
of the engine. The average corrosion potential of the other twenty three measurement points was 0.339 ± 0.007V. This 7mV standard deviation is comparatively small and within experimental error for
the equipment and measuring procedure suggesting that the entire vessel is in electrical connection
and the same film free corrosion mechanism applies to all areas on the JFR. This very small standard
deviation also means that it is not possible to determine any differences in corrosion behaviour
between the measurement points based on the Ecorr data.
The pH is often a more reliable indicator of changes in localised corrosion rates. From the Pourbaix
diagram the measurement positions were separated in cluster groups based on their pH values. There
appeared to be no obvious relationship between water depth and the average pH values of the
different groups as the difference in water depth from the shallowest to the deepest measurement
positions was only 3.1m. Position 7 had the most acidic pH value at 5.48 of all measurement points
indicating it has the highest corrosion rate. This position was located at the base of a 3m high vertical
hull plate on the starboard side of the wreck where water movement and oxygen impingement to the
concreted steel surface would very high and therefore the rate of the cathodic reaction (oxygen
reduction on the concretion/seawater surface) would be high concomitantly increasing the overall
corrosion rate of this high profile structural feature. Positions 16 and 19, which were measured on
collapsed deck plates on the port side of the wreck, possessed the next most acidic average pH value
of 6.42 ± 0.06. Positions 3, 5, 8, hull deck plates on the starboard side; position 22, a deck plate on the
port side; positions 12 and 23, bow hull plates on the starboard and port sides, respectively and a
bollard (6) near position 5 had the next most acidic average pH value of 6.84 ± 0.11. The next cluster
group had an average pH value of 7.50 ± 0.22 and included the rudder (1), starboard hull plates near
the stern (2, 4), a starboard deck plate (10) and hull (21) and deck plates (24) on the port side all
towards the bow section. The final cluster group had the most alkaline average pH value of 7.98 ± 0.11
and included starboard hull plates (9, 11) towards the bow, port hull and deck plates towards the stern
(13, 14, 15, 18) and the boiler (20). Position 17, which was the engine block had an alkaline pH of 7.89
and a corrosion voltage of -0.256V, almost 83mV more positive than the average Ecorr for this site (0.339V) which indicates there is very little metal remaining. This significant increase in the extent of
corrosion of the engine could be due to galvanic corrosion with the engine block preferentially
corroding with respect to the other elements of the engine, which would possess different metal
compositions.
Interpretation of the differences in the corrosion rates between so many measurement positions on this
very large site is particularly difficult but on closer inspection of the cluster groups it appears that
generally, those positions that have a very high profile, such as position 7 or are more damaged,
disarticulated and have collapsed completely, such as positions 16, 19, 3, 5, 6, 8, 12, 22 and 23 have
slightly higher corrosion rates than those areas, which are lower profile and/or are more protected from
wave action and oxygen impingement by deck supports or other structural hull features, such as
positions 1, 2, 4, 9-11, 13-15, 18, 20, 21 and 24. It also appears that generally, the bow section is
52
Appendix A
corroding at an elevated rate compared to the stern section, which would be a reflection of the bow
sections higher profile in the water column.
Generally the thicker the concretion and corrosion layer, the lower the surface pH but only if the
concretion layer remains essentially undisturbed (i.e. no damage occurs through human/natural
interference). Generally, this relationship applies to this wreck, where the most corroded positions (7,
16, 3, 5, 8, 12) possessed the thicker concretion and corrosion product layers (dtotal >8mm), which is in
general agreement with the conclusions regarding corrosion rate differences based on the pH
measurements.
Due to the thinner nature of the concretion and corrosion product layers on the JFR (5 ± 4mm), it was
possible to discern the interface between the concretion and the corrosion product layer on many of
the measurement positions so the depth of corrosion (dc) was measured at about 2 ± 1mm. It should
be noted that the dc measurements are not extremely accurate and the subsequent calculated
annualised corrosion rates should only be treated as approximations. If we assume that the JFR was
sunk in 1944 then it has been immersed for 68 years at the time of this corrosion survey. Therefore the
-1
calculated annualised corrosion rate was 0.03 ± 0.01mmy which is about a third of the average long-1
term corrosion rate for isolated iron in aerobic seawater at 0.11mmy . This lower corrosion rate is not
unexpected as the corrosion parameter measurements indicate that most of the structural remains on
this very large vessel (about 125m in length) are in electrical connection, dispersing the current density
over the entire hull thus lowering the overall corrosion rate.
3.9
AUXILIARY SUBMARINE CHASER - ASC
Figure 34. Auxiliary Submarine Chaser (ASC) – bow view (Carpenter 2012).
Date of Inspection
22 February 2010
53
Appendix A
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). On 22 February 2012 the winds were ENE at 18 to 24 knots in the morning
tending ENE at 17 to 23 knots in the afternoon. Seas were choppy with a moderate short period swell
(morning E 2.2m at 9 seconds; afternoon ENE 2.1m at 9 seconds). The tides were semi-diurnal over
the survey period and are reported in Table 1 (http://buoyweather.com). There was a slight current
(<0.5 knots) running in a NNE direction.
The in-water visibility was approximately 15m. The water depth to the top and base of the wreck at
0911 was 7.2 and 10.3m, respectively (Figure 34). The pH of seawater usually falls within the range of
7.5 to 8.3. The redox potential range of marine environments is -0.300 to 0.000V in reducing
environments and 0.000 to +0.250V in oxidising environments. The pH and redox potential of the
seawater on-site at 9.7m was 8.03 and 0.223V respectively, indicating a normal, open circulation
oxidising marine environment. The change in dissolved oxygen content, salinity and temperature of the
water column with depth measured on 23 February 2012 is shown in Table 16.
Table 16. Dissolved oxygen content, salinity and temperature of the seawater on the ASC site.
Water Depth (m)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
Average
Dissolved Oxygen
Content [ppm(S)]
6.16
6.24
6.11
6.01
6.16
5.94
5.86
5.86
5.93
5.88
5.71
5.76
6.02
5.80
5.71
5.63
5.40
5.07
5.85 ± 0.29
Salinity (ppK)
Temperature (°C)
35.9
35.8
35.7
35.6
35.7
35.7
35.6
35.7
35.7
35.7
35.7
35.7
35.7
35.7
35.7
35.7
35.7
35.7
35.7 ± 0.1
27.3
27.5
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6 ± 0.1
There was no significant change in salinity and temperature with increasing water depth, which is
typical of the hydrology of well mixed near coastal marine waters. The average water temperature was
27.6 ± 0.1°C and the average salinity of the water column was 35.7 ± 0.1ppK, which is within the usual
salinity range for the open ocean of 32-37ppK. The average dissolved oxygen content was 5.85 ±
0.29ppm. The change in dissolved oxygen concentration with increasing water depth is shown in
Figure 35.
54
Appendix A
under ship structure is a likely consequence. The site lies in the comparatively protected lagoon area
but due to the extensive profile of the vessel above the seabed it is conceivable that storms would
have an impact on the ship’s structure.
The vessel lies on its starboard side with approximately 12m of the bow section is still intact (Figure
34). The remainder of the hull structure has collapsed. The aft hull structure, stern, propeller, propeller
shaft, engines and ancillary equipment all appear to be missing. A small anchor, probably originally
stored on the deck, judging by its present location, appears to be in reasonable condition and state of
preservation. Most notably were some munitions lying inside the midships structure. The disarticulated
nature and missing features could be the result of post-war salvage and clean-up and possible
explosive training operations (McKinnon and Carrell 2011:47).
There are a few areas of active corrosion evident on the site, indicated by the presence of the typical
red/brown “rust” spots on the surface of the vessel but these are sporadic compared to the size of the
wreck remains. This ASC is included on the WWII Maritime Heritage Trail – Battle of Saipan and
divers are actively encouraged to visit the site provided they follow the visitation guidelines and do not
interfere with the site (i.e. disturb or attempt to remove any components).
Corrosion Survey
The corrosion parameters of fourteen different areas on the ASC were measured over a 55 minute
dive on 22 February 2012. The results are presented in Table 17 and the on-site positions shown in
Figure 36. In order to compare the corrosion data collected from the different positions measured on
the ASC and ascertain the thermodynamically stable state of the iron, the corrosion potentials (Ecorr)
and the pH of the residual iron alloy surfaces were plotted on the iron Pourbaix diagram in aerobic
seawater at 25°C (Figure 37). The temperature of the seawater on-site was 28°C, however this 3°C
increase does not significantly affect the nature or equilibria of the chemical species described in this
diagram.
Table 17. Corrosion parameter measurements on the ASC.
Position
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Description
hull 1, bow
deck 1, bow
hull 2, port side
hull 3, port side
hull 4, port side, end amidships
hull 5, stbd side, end amidships
box, stbd side
anchor, stbd side
hull 6, stbd side, fwd anchor
deck 2, end broken bow section
lower deck, port side
lower deck, stbd side
hull 7, end broken bow section
stanchion
pH
Ecorr vs
NHE (V)
7.54
-0.328
7.97
-0.329
7.48
-0.327
5.97
-0.327
8.01
-0.321
7.55
-0.323
6.95
-0.324
7.37
-0.330
6.08
-0.327
total penetration
8.01
-0.327
7.87
-0.328
7.99
-0.321
6.58
-0.336
dtotal = depth
of
concretion +
corrosion
(mm)
2
4
4
4
2
5
1
2
6
20
1
4
1
23
Water
Depth
(m)
9.0
8.3
8.7
8.1
7.5
7.2
9.3
10.3
9.9
9.5
8.0
9.3
8.8
9.6
56
Appendix A
Figure 36. Photomosaic indicating the corrosion parameter measurement positions on the ASC
(Richards after McKinnon and Carrell 2011:49).
Generally, the areas measured on the ASC were covered with relatively thin aerobic concretions
(average 6 ± 7mm). There were more hard corals on this wreck compared to the shallower wrecks,
such as the LVTs and the tanks, which are subjected to more natural and human interference but
possessed similar growth densities as was evident on the DAI1 and DAI2 sites (11m), suggesting
there is a significant amount of water movement on this site despite the increase in water depth (10m).
From the Pourbaix diagram (Figure 37), the intercepts of points 1, 3, 4, 6-9, 14 measured on the ASC,
lie in the active corrosion region, where ferrous ions are the thermodynamically stable chemical
species and corrosion will continue until all iron is consumed. Positions 2, 5, 11-13 lie on the
equilibrium line between active corrosion and the passive region, which implies that the typical aerobic
2+
corrosion mechanism where the major stable chemical species is the ferrous ion (Fe ) is in
equilibrium with the formation of an insoluble corrosion product layer of magnetite (Fe3O4). This is a
very common corrosion state for large steel ships where a large proportion of the vessel remains are
still in electrical contact. No corrosion parameters could be measured for position 10, where total
penetration of the metal occurred. Generally, with film free corrosion mechanisms, such as occurs on
concreted iron artefacts, an increase in the corrosion potential (tending more positive) indicates an
increase in the corrosion of the metal. The average corrosion potential of the thirteen measurement
points was -0.327 ± 0.004V. This 4mV standard deviation is comparatively small and within
experimental error for the equipment and measuring procedure suggesting that the entire vessel is in
electrical connection and the same film free corrosion mechanism applies to all areas on the ASC.
This very small standard deviation also means that it is not possible to determine any differences in
corrosion behaviour between the measurement points based on the Ecorr data.
57
Appendix A
(a)
Fe3+
0.8
active
0.6
FeO.OH
0.4
passive
0.2
Eh(v)
0.0
Fe2+
active
(b)
-0.2
4
9
-0.4
7
8
1,6
2,5,11,13
14 3 12
Fe3O4
passive
-0.6
-0.8
Fe
-1.0
immune
0
1
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 37. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the ASC site (Richards 2012).
The pH is often a more reliable indicator of changes in localised corrosion rates. Similar to the JFR the
measurement positions from the Pourbaix diagram were separated in cluster groups based on their pH
values. Again, there appeared to be no obvious relationship between water depth and the average pH
values of the different groups as the difference in water depth from the shallowest to the deepest
measurement positions was only 3.1m. Positions 4 and 9 had the most acidic average pH value at
6.02 ± 0.08 indicating the highest corrosion rate. Position 4 was on the upper surface of the port side
hull structure, where there was extensive explosive damage and position 9 was an isolated hull plate
on the collapsed starboard side of the wreck. Similarly, position 10, which had totally corroded was
located on the deck next to the broken end of the bow section where increases in micro-structural
stress of the metal would lead to concomitant increases in corrosion rate. Positions 14, a stanchion
and 9, a box both located midway along the site possessed the next most acidic pH values at 6.58 and
6.95 respectively. These were also isolated iron features and as such will corrode at a faster rate than
positions that are in electrical connection over a larger surface area (i.e. the hull structure). For
example, positions, 1, 3 and 6, which were on the hull structure and position 8, which was an anchor
lying in direct contact with the collapsed starboard hull plates lying flat on the seabed surface and was
hence, in a more protected area had a more alkaline average pH value of 7.48 ± 0.08, indicating a
lower corrosion rate than the previously mentioned positions. Similarly, positions 2, 5, 11-13 were all
part of the hull structure in relatively protected areas and possessed the most alkaline average pH
value of 7.97 ± 0.06 indicating the lowest corrosion rate. In addition, generally it appears that the
starboard side is corroding at a slightly elevated rate compared to the port side, which would be a
reflection of the increased damage, disarticulation and collapse on this side of the vessel.
Generally the thicker the concretion and corrosion layer, the lower the surface pH but only if the
concretion layer remains essentially undisturbed (i.e. no damage occurs through human/natural
interference). This relationship on the ASC was ambiguous as the average d total of the thinner
concretion and corrosion product layers was 3 ± 2mm but some of these areas had quite low surface
pHs. However, positions 9 (dtotal = 6mm) and 14 (dtotal = 23mm) had acidic pH values in conjunction
with thicker concretion and corrosion product layers, which is in agreement with the conclusions
regarding corrosion rate differences based on their pH measurements.
Unfortunately, unlike the JFR, even though the dtotals on the ASC were relatively thin (6 ± 7mm) it was
not possible to discern the interface between the concretion and corrosion product layer on the
58
Appendix A
positions so the depth of corrosion (dc) could not be measured and therefore the average long-term
annualised corrosion rate could not be estimated.
Based on the corrosion parameter measurements it is difficult to say whether the ASC is corroding at a
faster rate than the JFR as most average measurements are within their respective statistical errors,
with the exception of the average Ecorr values which were -0.327 ± 0.004V on ASC and
-0.339 ±
0.007V on the JFR. This small increase in average E corr of the ASC suggests that it may be corroding
at a slightly faster rate than JFR. This is not unexpected as it is known that isolated iron artefacts and
steel hull structures that have been damaged either through natural phenomena (e.g. cyclonic activity)
or human intervention (e.g. salvage, explosive damage during WWII) possess higher corrosion rates
than those hull structures that are relatively intact (i.e. JFR), where the current density of the corrosion
process can be spread over a much larger surface area effectively lowering the corrosion rate
(MacLeod and Richards 2011; Richards et al. 2011).
3.10
STEAMSHIP - SS
Figure 38. Steamship (SS) – boilers (Carpenter 2012).
Date of Inspection
23 February 2010
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). On 23 February 2012 the winds were ENE at 15 to 21 knots in the morning
tending E in the afternoon. Seas were moderately choppy with small short period wind waves (E 1.8m
at 9 seconds). The tides were semi-diurnal over the survey period and are reported in Table 1
(http://buoyweather.com). There was no discernible current observed during the survey.
59
Appendix A
The in-water visibility was approximately 5m, which was considerably less than the other sites
investigated during the survey period. The depth to the base of the wreck at 0955 was 10.4m. The pH
of seawater usually falls within the range of 7.5 to 8.3. The redox potential range of marine
environments is -0.300 to 0.000V in reducing environments and 0.000 to +0.250V in oxidising
environments. The pH and redox potential of the seawater on-site at 9.9m was 8.05 and 0.278V
respectively, indicating a normal, open circulation oxidising marine environment. The change in
dissolved oxygen content, salinity and temperature of the water column with depth measured on 23
February 2012 is shown in Table 18.
Table 18. Dissolved oxygen content, salinity and temperature of the seawater on the SS site.
Water Depth (m)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
10.5
11.0
Average
Dissolved Oxygen
Content [ppm(S)]
5.85
5.76
5.68
5.70
5.65
5.72
5.69
5.88
5.87
5.96
5.48
5.77
5.78
5.48
5.54
5.67
6.70
5.70
5.94
5.53
5.53
5.42
5.22
5.72 ± 0.28
Salinity (ppK)
Temperature (°C)
35.6
35.8
35.7
35.7
35.7
35.7
35.8
35.7
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.6
35.8 ± 0.1
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7
27.7 ± 0.0
There was no significant change in salinity and temperature with increasing water depth, which is
typical of the hydrology of well mixed near coastal marine waters. The average water temperature was
27.7 ± 0.0°C and the average salinity of the water column was 35.8 ± 0.1ppK, which is within the usual
salinity range for the open ocean of 32-37ppK. The average dissolved oxygen content was 5.72 ±
0.28ppm. The change in dissolved oxygen concentration with increasing water depth is shown in
Figure 39.
60
Appendix A
due to the relatively high profile of the shipwreck remains and the establishment of hard corals in some
areas it is likely that significant accumulation of sediment does not readily occur on this site.
The vessel is considerably damaged and although many parts are disconnected they lie in close
proximity to each other. Various sections are bent, twisted and distorted suggesting war damage
rather than a direct consequence of seabed impact when it sank or collapse due to corrosion. A
dislodged ships boiler appears intact as do some storage tanks. The vessel’s engine and propeller
were not observed and may have been salvaged. Overall impressions of the site were limited by the
relatively poor visibility. Generally the metal structures appear to be in a strong and robust condition.
There were no areas of active corrosion evident on the wreck remains. This steamship is not included
on the WWII Maritime Heritage Trail – Battle of Saipan therefore visitation to the site by divers and
snorkelers would be less than to those wrecks that are listed on the heritage trail.
Corrosion Survey
The corrosion parameters of twelve different areas on the SS were measured over a 65 minute dive on
23 February 2012. The results are presented in Table 19. The on-site positions are not shown as no
site plan or photomosaic has been produced for this wreck at this point in time. In order to compare the
corrosion data collected from the different positions measured on the SS and ascertain the
thermodynamically stable state of the iron, the corrosion potentials (Ecorr) and the pH of the residual
iron alloy surfaces were plotted on the iron Pourbaix diagram in aerobic seawater at 25°C (Figure 40).
The temperature of the seawater on-site was 28°C, however this 3°C increase does not significantly
affect the nature or equilibria of the chemical species described in this diagram.
62
Appendix A
Table 19. Corrosion parameter measurements on the SS.
Position
Number
1
2
3
4
5
6
7
8
9
10
11
12
Description
bow hull plate
water tank 1
water tank 2
midships hull plate
water tank 3 (smaller)
kingpost
stern hull plate
shaft
engine
boiler 1
boiler 2
anchor
pH
Ecorr vs NHE
(V)
8.00
8.04
7.43
7.99
8.00
8.04
8.00
8.03
8.04
8.03
7.97
6.38
-0.360
-0.360
-0.358
-0.353
-0.353
-0.355
-0.358
-0.355
-0.350
-0.352
-0.353
-0.357
dtotal =
depth of
concretion
+
corrosion
(mm)
2
2
1
1
3
2
9
0
2
6
6
11
dc =
depth of
corrosion
(mm)
Water
Depth
(m)
1
1
9.3
9.7
9.7
10.1
9.2
8.4
7.6
8.4
9.1
9.8
9.3
10.4
1
8
3
1
2
2
(a)
Fe3+
0.8
active
0.6
FeO.OH
0.4
passive
0.2
Eh(v)
0.0
Fe2+
active
(b)
-0.2
12
3
-0.4
1,4,5,7,10,11
2,6,8,9
Fe3O4
passive
-0.6
-0.8
Fe
-1.0
immune
0
1
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 40. Pourbiax diagram for iron (10 M) in aerobic seawater at 25°C indicating the
intercepts of the areas measured on the SS site (Richards 2012).
Generally, the areas measured on the SS were covered with relatively thin aerobic concretions
(average 4 ± 3mm). There appeared to be less hard corals on this wreck compared to the other
wrecks and this is probably a reflection of the increased turbidity on the site restricting secondary
marine growth.
From the Pourbaix diagram (Figure 40), the intercepts of points 1, 2, 4-11 measured on the SS, lie on
the equilibrium line between active corrosion and the passive region, which implies that the typical
2+
aerobic corrosion mechanism where the major stable chemical species is the ferrous ion (Fe ) is in
equilibrium with the formation of an insoluble corrosion product layer of magnetite (Fe 3O4). This is a
very common corrosion state for large steel ships where a large proportion of the vessel remains are
still in electrical contact. The intercepts of position 3 (water tank 2) and position 12 (anchor) lie in the
active corrosion zone where ferrous ions are the thermodynamically stable chemical species and
63
Appendix A
corrosion will continue until all iron is consumed. Generally, with film free corrosion mechanisms, such
as occurs on concreted iron artefacts, an increase in the corrosion potential (tending more positive)
indicates an increase in the corrosion of the metal. The average corrosion potential of the twelve
measurement points was -0.355 ± 0.003V. This 3mV standard deviation is comparatively small and
within experimental error for the equipment and measuring procedure suggesting that the entire vessel
is in electrical connection and the same film free corrosion mechanism applies to all areas on the SS.
This very small standard deviation also means that it is not possible to determine any differences in
corrosion behaviour between the measurement points based on the Ecorr data.
The pH is often a more reliable indicator of changes in localised corrosion rates. Again, there
appeared to be no obvious relationship between water depth and the pH values of the different
measurement positions as the difference in water depth from the shallowest to the deepest points was
only 2.8m. Positions 1, 2, 4-11 possessed the most alkaline average pH value of 8.01 ± 0.02 indicating
a lower corrosion rate. This is not unexpected as large steel wreck remains in electrical connection
tend to have lower corrosion rates than isolated iron alloy artefacts due to the current density being
spread over a much larger surface area. Position 3, which was the water tank 2 had a more acidic pH
value of 7.43 and position 12 which was the anchor possessed the most acidic pH value of 6.38
measured on the site. These more acidic values indicate that these features are corroding at a higher
rate than the other parts of the vessel and therefore it is probable that this particular water tank and the
anchor are electrically isolated from the rest of the vessel remains.
Generally the thicker the concretion and corrosion layer, the lower the surface pH but only if the
concretion layer remains essentially undisturbed (i.e. no damage occurs through human/natural
interference). Generally, this relationship applies to this wreck (with the exception of positions 3 and 7),
where the least corroded positions possessed the thinner concretion and corrosion product layers
(average dtotal = 2 ± 2mm) and the anchor (12), which had the thickest dtotal of 11mm possessed the
most acidic pH value. This is in general agreement with the conclusions regarding corrosion rate
differences based on the pH measurements.
Due to the thinner nature of the concretion and corrosion product layers on the SS (4 ± 3mm), it was
possible to discern the interface between the concretion and the corrosion product layer on many of
the measurement positions so the depth of corrosion (dc) was measured at about 2 ± 2mm. It should
be noted that the dc measurements are not extremely accurate and the subsequent calculated
annualised corrosion rates should only be treated as approximations. If we assume that the SS was
sunk in 1944 then it has been immersed for 68 years at the time of this corrosion survey. Therefore the
-1
calculated annualised corrosion rate was 0.02 ± 0.01mmy which is about a fifth of the average long-1
term corrosion rate for isolated iron in aerobic seawater at 0.11mmy . This lower corrosion rate is not
unexpected as the corrosion parameter measurements indicate that most of the structural remains on
this large vessel are in electrical connection, dispersing the current density over the entire hull thus
lowering the overall corrosion rate.
Based on the average corrosion potentials of the JFR (-0.339 ± 0.007V), the ASC (-0.327 ± 0.004V)
and the SS (-0.355 ± 0.003V) it appears that the small but statistically valid decrease in the corrosion
potential (Ecorr) of the SS suggests that this vessel is corroding at a slower rate than both the ASC and
the JFR. This would seem to suggest that human interference (i.e. recreational diving activities) is
having some impact on the deterioration rate of the JFR and ASC sites as the SS site is not on the
diving heritage trail. However, the local environment (i.e. increase in turbidity) may also be contributing
to this decrease in corrosion rate on the SS site.
64
Appendix A
4
CONSERVATION
AIRCRAFT WRECKS
4.1
ASSESSMENTS
–
ALUMINIUM
ALLOY
GRUMMAN TBM AVENGER – AVR
Figure 41. Grumman TBM Avenger (AVR) (Carpenter 2012).
Date of Inspection
20 February 2012
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). In the morning (20/6/2012) the winds were ENE at 13 to 18 knots which
tended more easterly in the afternoon, increasing to 15 to 21 knots. Seas were relatively consistent
over the entire day with breezy whitecapping conditions and moderate choppy seas with small, short
period wind waves (morning - ENE 1.5m at 10 seconds; afternoon – NE 1.7m at 10 seconds). The
tides were semi-diurnal over the survey period and are reported in Table 1. There was a consistent
and relatively strong current (~1 knot) running over the reef into the lagoon.
The in-water visibility was approximately 10-15m. The depth to the base of the wreck at 1027 was
2.7m. The pH of seawater usually falls within the range of 7.5 to 8.3. The redox potential range of
marine environments is -0.300 to 0.000V in reducing environments and 0.000 to +0.250V in oxidising
environments. The pH and redox potential of the seawater on-site at 2.7m was 8.13 and 0.208V
respectively, indicating a normal, open circulation oxidising marine environment. The change in
dissolved oxygen content, salinity and temperature of the water column with depth measured on 20
February 2012 is shown in Table 20.
65
Appendix A
66
Appendix A
other physico-chemical measurements, are typical for a shallow, near coastal, open circulation,
oxidising marine environment, where corrosion rates are likely to be relatively high.
Wreck Site
The Grumman TBM Avenger (AVR) is located on the south western side of Saipan, inside the barrier
reef near the north edge of the main channel entrance to Tanapag Harbour (GPS
) at a depth of about 3m (Figure 43). The aircraft was identified as a Grumman TBM
Avenger torpedo bomber (McKinnon and Carrell 2011), constructed most probably of duralumin
(aluminium alloy containing 3-5% Cu, 0.4-1.0% Mn, 0.3-0.6% Mg) and is 12.19m in length, 5.00m high
and had a 16.51m wingspan (http://www.flugzeuginfo.net). The aircraft remains are mostly submerged,
however the hydraulic landing gear, which is in the fully extended position (Figure 41), is exposed to
the atmosphere at extreme low tides.
Figure 43. Location of the aircraft wrecks, Saipan, CNMI (Richards 2012 after Google Earth
2012).
The aircraft remains are located on the top of the barrier reef which surrounds Tanapag Harbour and
are subjected to considerable water movement due to this high energy environment. Large quantities
of dead coral are strewn over the seabed, which is comprised of coarse grained calcareous sediment.
The surviving aircraft structure is becoming integrated into the reef as corals have developed. The
maximum exposure height of the main structure is approximately 1.5m above the seabed.
Its reef top position implies that the AVR remains are always exposed and overall sediment burial is
very unlikely. Sand is present inside and in front of the engine bay cavity which may scour in more
turbulent conditions. Localised, limited and partial exposure cycles may occur in this area. The
development of coral growth may potentially cover the aircraft remains with time however the limited
68
Appendix A
size and extent of hard coral growths on the aircraft structure after some 65 years of immersion is
likely to be a consequence of the more dynamic localised environment (including storm damage). The
smooth metal skins of aircraft generally seem to inhibit the establishment of larger forms of marine
biota unless a purchase can be made due to a break in the surface or a ferrous metal is present. The
exception can be the colonisation of the under-surfaces of areas, such as wings, where light levels
may exist that are similar to those found in the entrance to underwater caves, etc and therefore the
conditions suit the establishment of lower profile sponges, etc and small gorgonia (sea fans) that
usually colonize these features. The protruding landing struts are subjected to a relatively strong and
constant current as water streams over the reef into the lagoon and are essentially devoid of marine
growth. The much lower profile, and largely reef-shielded wing remains are less affected by excessive
water movement.
The aircraft remains lie inverted and consist principally of the central area between the prominent
wheel struts and the incomplete remains of the main wings. The engine and propeller are absent. The
aluminium skin of the wings is corroded and has a number of irregular holes and smaller perforations.
There is a disconnected gun turret ring about 20-40m north of the main site. The wheel-well openings
remain discernable and the fuselage wreckage is covered in very dense coralline growth. The body
and tail plane of the aircraft are also absent. The condition of the aluminium alloy is poor in
comparison with those aircraft wrecks located on sandy sediments in calmer areas of the lagoon.
Overall damage to the aircraft structure is not readily distinguishable between crash, storm or potential
war damage. The site is shallow (3m) and its reef top position means that it must be affected by
turbulent seas generated by storms and cyclones. McKinnon and Carrell (2011) mention that the local
surfers use the protruding landing gear as boat moorings and bright, bare aluminium surfaces are
evident suggesting some form of interference in these areas. This aircraft is included in the WWII
Maritime Heritage Trail – Battle of Saipan and lies within a Marine Conservation Area. Divers and
snorkelers are actively encouraged to visit the site provided they follow the local visitation guidelines
and do not interfere with the site (i.e. disturb or attempt to remove any cultural or natural components).
Corrosion Survey
The corrosion parameters of twelve different areas on the AVR were measured over a 54 minute dive
on 20 February 2012. The results are presented in Table 21 and the on-site positions shown in Figure
44. In order to compare the corrosion data collected from the different positions measured on the AVR
and ascertain the thermodynamically stable state of the aluminium, the corrosion potentials (E corr) and
the pH of the residual aluminium alloy surfaces were plotted on the aluminium Pourbaix diagram in
aerobic seawater at 25°C (Figure 45). The temperature of the seawater on-site was 28°C, however
this 3°C increase does not significantly affect the nature or equilibria of the chemical species described
in this diagram.
Table 21. Corrosion parameter measurements on the AVR.
Position
Number
1
2
3
4
5
6
7
8
9
10
11
12
Description
port wing
stbd wing
cylinder at bow
box aft of cylinder
aft port wing, aft wing support
port wheel support strut
port wheel support (clean metal)
aft stbd win g, aft wheel support
stbd wheel support strut
stbd wheel support
stbd side strut, aft stbd wing
gun turret remains, 20m N off site
pH
7.96
7.94
7.83
8.06
8.03
8.18
8.16
8.09
8.16
8.16
8.16
8.02
Ecorr vs NHE
(V)
-0.440
-0.439
-0.498
-0.438
-0.440
-0.439
-0.437
-0.439
-0.438
-0.437
-0.436
-0.401
Water Depth
(m)
2.3
2.6
2.9
2.6
2.4
2.0
0.5
2.5
1.9
0.5
2.6
3.8
69
Appendix A
Figure 44. Schematic plan of the TBM Avenger (AVR) indicating the corrosion parameter
measurement positions (Richards 2012 after Bell 2010 in McKinnon and Carrell 2011:89).
(a)
0.8
0.6
0.4
Al 3+
0.2
active
-
Al 2O3 .3H2 O
AlO2
passive
active
0.0
Eh(v)
-0.2
(b)
12
1,5
-0.4
-0.6
2
3
6,7,9,10,11
4,8
-0.8
-1.0
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 45. Pourbiax diagram for aluminium (3.7 x 10 M) in aerobic seawater at 25°C indicating
the intercepts of the areas measured on the AVR site (Richards 2012).
Generally, the areas measured on the AVR were covered with a very thin (<1mm) mucilaginous layer
consisting of proteinaceous and algal based material in combination with hydrated aluminium
hydroxide gels. Hence measurements of total depth of concretion and corrosion products (dtotal) are not
applicable for aluminium alloy aircraft. On some areas of the aircraft patches of blue corrosion
products, typical of hydrated oxidised copper corrosion products, were observed, indicating the use of
a copper - aluminium alloy, such as Duralumin in these areas.
70
Appendix A
From the Pourbaix diagram (Figure 45), the intercepts of all points measured on the AVR, lie in the
passive region, where Al2O3.3H2O is the dominant corrosion product (Equation 2) and forms a
continuous passivating layer, effectively slowing the corrosion rate. This is a very common corrosion
state for aluminium alloy aircraft in marine environments (MacLeod 2006).
2 Al + 3H2O → Al2O3 + 6H + 6e
+
-
(2)
However, whilst the addition of copper to aluminium (e.g. Duralumin) increases the strength it
dramatically decreases the corrosion resistance of the metal to seawater. Without a protective paint
film such alloys suffer severe pitting and total perforation can occur in a few years. In the pitting of
aluminium the copper acts as a cathodic site for the reduction of oxygen (Equation 3). More noble
impurities, such as Al3Fe, act in a similar manner. Chloride ions are known to be absorbed onto
aluminium and as little as 15ppm chloride can initiate pit growth due to the breakdown of the protective
oxide film. The corrosion of aluminium (the anodic reaction) occurs at the bottom of the pit (Equation 4)
and the aluminium ions migrate towards the interfacial region (the area between the metal and the
corrosive medium) where hydrolysis occurs (Equation 5), which makes the pit acidic. Chloride ions
migrate into the pit to form aluminium chloride (AlCl3) which dissolves in the solution. There is an
equilibrium between the formation of aluminium oxide and AlCl3 at this interfacial region (Equation 6).
When aluminium chloride forms a pit develops and when alumina (Al 2O3) forms the pit will passivate.
The chloride ions directly affect the corrosion potential of aluminium and the higher the chloride ion
concentration the more negative the corrosion potential and the faster the metal will corrode.
O2 + 2H2O + 4e → 4OH
3+
Al → Al + 3e
3+
+
Al + 3H2O → Al(OH)3 + 3H
+
Al2O3 + 6H + 6Cl 2AlCl3 (aq) + 3H2O
-
-
(3)
(4)
(5)
(6)
In addition, copper has a limited solubility in aluminium (up 2 wt%) and unless the liquid metal is
rapidly cooled copper will not be uniformly distributed throughout the grains of the aluminium phase. If
precipitation hardening (increase in hardness of the metal due to the precipitation of the CuAl 2
intermetallic phase) occurs, the areas around the grain boundaries become depleted in copper and as
such become more anodic (more reactive) than the rest of the grain. Under these conditions the metal
is subject to intergranular corrosion. In the absence of complicating factors the more reactive metal or
metal phase will have a more negative corrosion potential. For example aluminium has a corrosion
potential of -0.520V vs NHE in seawater (more reactive) but 2% copper in a solid solution of aluminium
has a corrosion potential of -0.420V vs NHE (less reactive). This difference of 100mV in the Ecorr
values is quite large and can lead to markedly different corrosion rates across the different phases of
the sheet metal (MacLeod 2006).
These types of corrosion behaviour were noted on the aircraft especially on the larger parts of the
wings and fuselage, primarily adjacent to the connecting seams of the aluminium alloy metal where
significant loss of metal occurred, either through pitting or intergranular corrosion or more likely a
combination of both mechanisms.
Oxidation of aluminium alloys is largely controlled by the passage of electrons from the metal through
defects in the passivating oxide coating to react with oxygen in the surrounding environment. Owing to
the nature of the passivating film, less negative Ecorr values generally imply a lower corrosion rate but
only if the metal composition of the alloys are very similar. Obviously the incorporation of different
alloying metals, such as copper and iron at different percentages will change the corrosion potential of
the metal. Therefore it is often difficult to determine differences in corrosion behaviour between
different measurement points on aircraft as varying alloys are used for different parts of the machines.
However, since all aluminium alloys are corroding in a common oxidising marine environment in
Tanapag Harbour, the different values of the corrosion potentials may provide a guide to the
underlying differences in alloy composition of the aircraft.
Therefore most of the Ecorr values of the positions, namely 1, 2 and 4 to 11 all fall into the average
corrosion potential range of -0.438 ± 0.001V indicating that the metal composition in these areas is
very similar. The two positions that lie outside this average Ecorr range are 3 and 12. Position 3, which
71
Appendix A
was a cylinder associated with the engine, had a more anodic Ecorr value equal to -0.498V and coupled
with a more acidic pH value of 7.83, indicates a higher corrosion rate. This increase in corrosion rate
could be due to galvanic corrosion, i.e the more reactive aluminium alloy parts of the engine in
electrical connection to the more noble metals, such as copper and iron within the engine, are
corroding at a faster rate thus providing protection to these less reactive metals. However, a box
measured aft of this cylinder (4) fell within the average Ecorr range for most of the measurement
positions hence it is more likely that the aluminium alloy composition of the cylinder is different to the
rest of the aircraft and contains less copper and more aluminium making the corrosion potential more
anodic and hence, more reactive. On the other hand, position 12, which was the gun turret remains,
located about 20m north of the main aircraft wreckage had a less negative E corr of -0.401V, indicating
that there is probably more iron associated with this section of the aircraft, which is effectively
decreasing the corrosion rate. Another factor which may also have an effect on lowering the corrosion
rate of the gun turret is the increase in the water depth. The gun turret lies in 3.8m of water, which is
1.7m deeper than the average water depth of the main wreckage (2.1 ± 0.8m). It is well known that
corrosion rates tend to fall with increases in water depth as the amount of oxygen impingement to the
metal surface decreases with the decrease in overall water movement as the water depth increases.
Marine fouling on aluminium alloys tends to be dominated by bacteria, which form thin biofilms and
unlike iron wrecks where marine organisms respond to the release of iron ions and therefore the depth
of concretion increases with increasing corrosion rates, the overall amount of marine growth found on
aircraft tends to be limited. Hence, the pH values measured on aircraft are generally very conservative,
that is the underlying acidity will be higher (i.e. pH lower) than reported, since there is no significant
reserve of acidic materials trapped under the thin protective corrosion and biofilm layer that can
effectively buffer the immediate effect of the corroding surface being directly exposed to the more
alkaline seawater (MacLeod 2006). However, owing to the inherent acidity of hydrated trivalent metal
3+
ions, such as aluminium, Al , a series of hydrolysis reactions will take place (see equation 5) in the
microenvironment of the pits or underlying the biofilm. Hence, the amount of aluminium corrosion
products will be in dynamic equilibrium with the acidity arising from the hydrolysis reactions, thus
3+
higher concentrations of Al ions will be reflected in more acidic pH values. So since the pH is a
measure of the underlying concentration of the metal ion, then more alkaline pH values will reflect
3+
lower corrosion rates, as lower concentrations of Al ions will undergo less hydrolysis and produce
less acid. This is consistent with the more acidic pH value of 7.83 measured on the cylinder at the bow
(3) compared to the average pH value of the eleven other positions of 8.08 ± 0.08 indicating the
cylinder is corroding at a faster rate than the rest of the aircraft remains due to a difference in metal
composition. This is also in general agreement with the conclusions regarding corrosion rate
differences based on the Ecorr measurements.
72
Appendix A
4.2
AICHI E13A – JAKE
Figure 46. Aichi E13A (JAKE) (Carpenter 2012).
Date of Inspection
20 February 2012
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). In the morning (20/6/2012) the winds were ENE at 13 to 18 knots which
tended more easterly in the afternoon, increasing to 15 to 21 knots. Seas were relatively consistent
over the entire day with breezy whitecapping conditions and moderate choppy seas with small, short
period wind waves (morning - ENE 1.5m at 10 seconds; afternoon – NE 1.7m at 10 seconds). The
tides were semi-diurnal over the survey period and are reported in Table 1. There was a slight current
(<0.5 knot) running in a NNE direction.
The in-water visibility was approximately 20m. The depth to the base of the wreck at 1156 was 6.8m.
The pH of seawater usually falls within the range of 7.5 to 8.3. The redox potential range of marine
environments is -0.300 to 0.000V in reducing environments and 0.000 to +0.250V in oxidising
environments. The pH and redox potential of the seawater on-site at 2.7m was 8.19 and 0.211V
respectively, indicating a normal, open circulation oxidising marine environment. The change in
dissolved oxygen content, salinity and temperature of the water column with depth measured on 20
February 2012 is shown in Table 22.
73
Appendix A
Table 22. Dissolved oxygen content, salinity and temperature of the seawater on the JAKE site.
Water Depth (m)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
Average
Dissolved Oxygen
Content [ppm(S)]
6.65
6.56
6.54
6.56
6.52
6.52
6.53
6.52
6.45
6.45
6.49
6.54
6.48
6.43
6.44
6.51 ± 0.06
Salinity (ppK)
Temperature (°C)
35.6
35.7
35.8
35.9
35.9
35.9
35.9
36.1
36.1
36.1
36.1
36.2
36.2
36.2
36.2
36.0 ± 0.2
27.4
27.4
27.3
27.3
27.2
27.2
27.2
27.1
27.1
27.1
27.1
27.0
27.0
27.1
27.0
27.2 ± 0.1
There was no significant change in salinity and temperature with increasing water depth, which is
typical of the hydrology of well mixed near coastal marine waters. The average water temperature was
27.2 ± 0.1°C and the average salinity of the water column was 36.0 ± 0.2ppK, which is within the usual
salinity range for the open ocean of 32-37ppK. The average dissolved oxygen content was 6.51 ±
0.06ppm. The change in dissolved oxygen concentration with increasing water depth is shown in
Figure 47.
For open circulation ocean environments, there is usually a surface maximum in the dissolved oxygen
concentration. This maximum is a direct result of absorption from the atmosphere interface, increased
water movement and photosynthetic activity by plants and cyanobacteria. Typically, after this surface
maximum the dissolved oxygen concentration of the water column will decrease with increasing depth.
Factors contributing to this trend are decreasing water movement, which leads to less oxygen
exchange with the atmosphere, decreasing photosynthetic activity due to less light penetration and
increasing aerobic respiration of plankton in the photosynthetic zone. However, the relatively small
standard deviation between the measurements and a decrease of only 0.22ppm over the 7.0m depth
range indicates that there is very little variation in the dissolved oxygen content with increasing water
depth over such a shallow depth range, which is not unexpected. Hence, this trend coupled with the
other physico-chemical measurements, are typical for a shallow, near coastal, open circulation,
oxidising marine environment, where corrosion rates are likely to be relatively high.
74
Appendix A
snorkelers are actively encouraged to visit the site provided they follow the local visitation guidelines
and do not interfere with the site (i.e. disturb or attempt to remove any cultural or natural components).
Corrosion Survey
The corrosion parameters of eleven different areas on the JAKE were measured over a 44 minute dive
on 20 February 2012. The results are presented in Table 23 and the on-site positions shown in Figure
48. In order to compare the corrosion data collected from the different positions measured on the
JAKE and ascertain the thermodynamically stable state of the aluminium, the corrosion potentials
(Ecorr) and the pH of the residual aluminium alloy surfaces were plotted on the aluminium Pourbaix
diagram in aerobic seawater at 25°C (Figure 49). The temperature of the seawater on-site was 28°C,
however this 3°C increase does not significantly affect the nature or equilibria of the chemical species
described in this diagram.
Table 23. Corrosion parameter measurements on the JAKE.
Position
Number
1
2
3
4
5
6
7
8
9
10
11
Description
stbd wing tip
stbd wing near fuselage
stbd fwd fuselage
boss, port side
port propeller
port fwd fuselage
port wing near fuselage
port wing near tip
float
aft fuselage midway
vertical propeller
pH
8.01
7.86
8.05
7.94
8.03
7.93
8.02
8.02
7.83
8.13
8.04
Ecorr vs NHE
(V)
-0.448
-0.448
-0.421
-0.442
-0.442
-0.426
-0.448
-0.448
-0.463
-0.447
-0.442
Water Depth
(m)
5.9
5.9
6.2
6.3
6.6
6.3
6.0
6.4
6.3
6.0
5.4
76
Appendix A
Figure 48. Schematic plan of the Aichi E13A (JAKE) indicating the corrosion parameter
measurement positions (Richards 2012 after Bell 2010 in McKinnon and Carrell 2011:56).
(a)
0.8
0.6
0.4
Al 3+
0.2
active
-
Al 2O3 .3H2 O
AlO2
passive
active
0.0
Eh(v)
-0.2
(b)
6
2
-0.4
3
10
9 41,5,7,8 11
-0.6
-0.8
-1.0
2
3
4
5
6
7
8
9
10 11 12
pH
77
Appendix A
-6
Figure 49. Pourbiax diagram for aluminium (3.7 x 10 M) in aerobic seawater at 25°C indicating
the intercepts of the areas measured on the JAKE site (Richards 2012).
Generally, the areas measured on the JAKE were covered with a very thin (<1mm) mucilaginous layer
consisting of proteinaceous and algal based material in combination with hydrated aluminium
hydroxide gels. Hence measurements of total depth of concretion and corrosion products (dtotal) are not
applicable for aluminium alloy aircraft. Unlike the AVR, there was no observed evidence of the typical
blue copper corrosion products on the JAKE.
From the Pourbaix diagram (Figure 49), the intercepts of all points measured on the JAKE, lie in the
passive region, where Al2O3.3H2O is the dominant corrosion product (Equation 2) and forms a
continuous passivating layer, effectively slowing the corrosion rate. This is a very common corrosion
state for aluminium alloy aircraft in marine environments.
Some sections of the JAKE, such as the wings and the float showed discreet areas of perforation,
however the extent was significantly less than that observed on the AVR, suggesting pitting corrosion
was the preferred corrosion mechanism in this environment with considerably less intergranular
corrosion occurring. This would indicate that the environment on the AVR site is considerably more
aggressive than that experienced by the JAKE, which is not unexpected as the AVR site is much
shallower and sits on a reef platform which experiences much greater overall water movement.
The Ecorr values of eight of the eleven measurement positions, namely 1, 2, 4, 5, 7, 8, 10 and 11 all fall
into the average corrosion potential range of -0.446 ± 0.003V indicating that the metal composition of
the wings, aft fuselage, propellers and boss head are very similar. Positions 3 and 6, measured on
forward part of the fuselage, had an average Ecorr of -0.424 ± 0.002V, which is 22mV more positive
than the other areas, indicating that this section is corroding at a slightly slower rate. This would not be
unexpected as this area of the fuselage is in direct electrical contact with what appears to be
machinery associated with the engine, where there would be more contact with iron, copper and other
less reactive metals concomitantly lowering the overall corrosion rate in this area. Position 9, which
was the disarticulated float, possessed the most negative Ecorr equal to -0.463V and the most acidic pH
value of 7.83, indicating the aluminium alloy composition of the float is different to the rest of the
aircraft and contains less copper and more aluminium making the corrosion potential more anodic and
hence, more reactive.
78
Appendix A
4.3
MARTIN PBM MARINER – MNR
Figure 50. Martin PBM Mariner (MNR) (Carpenter 2012).
Date of Inspection
20 February 2012
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). In the morning (20/6/2012) the winds were ENE at 13 to 18 knots which
tended more easterly in the afternoon, increasing to 15 to 21 knots. Seas were relatively consistent
over the entire day with breezy whitecapping conditions and moderate choppy seas with small, short
period wind waves (morning - ENE 1.5m at 10 seconds; afternoon – NE 1.7m at 10 seconds). The
tides were semi-diurnal over the survey period and are reported in Table 1. There was a slight current
(<0.5 knot) running in a NNE direction.
The in-water visibility was approximately 25m. The depth to the base of the wreck at 1349 was 7.0m.
The pH of seawater usually falls within the range of 7.5 to 8.3. The redox potential range of marine
environments is -0.300 to 0.000V in reducing environments and 0.000 to +0.250V in oxidising
environments. The pH and redox potential of the seawater on-site at 6.8m was 8.19 and 0.222V
respectively, indicating a normal, open circulation oxidising marine environment. The change in
dissolved oxygen content, salinity and temperature of the water column with depth measured on 21
February 2012 is shown in Table 24.
79
Appendix A
range indicates that there is very little variation in the dissolved oxygen content with increasing water
depth over such a shallow depth range, which is not unexpected. Hence, this trend coupled with the
other physico-chemical measurements, are typical for a shallow, near coastal, open circulation,
oxidising marine environment, where corrosion rates are likely to be relatively high.
Wreck Site
The Martin PBM Mariner (MNR) (Figure 50) is located in Tanapag Harbour approximately 600m SSE
of Mañagaha Island (GPS
) at a depth of about 7.0m (Figure 43). After
extensive historical and archaeological investigation the aircraft was positively identified in 2010 as a
Martin PBM Mariner U.S. twin-engined maritime patrol flying boat (McKinnon and Carrell 2011:74),
constructed most probably of duralumin (aluminium alloy containing 3-5% Cu, 0.4-1.0% Mn, 0.3-0.6%
Mg) and is 24.33m in length, 8.38m high and had a 35.97m wingspan (McKinnon and Carrell 2011:75).
The aircraft remains are totally submerged at all times.
The aircraft remains are located on a relatively flat, undulating seabed comprising primarily of fine
calcareous sediment. There are sporadic large coral outcrops in close vicinity to the wreck. Some
lower profile sections, such as the wings were covered in a thin layer of fine sediment. There appears
to be evidence of localised, seasonal exposure/reburial cycles on the site, however, most of the
remains are exposed and total burial seems unlikely to occur. The maximum exposure height of the
main structure is approximately 1.5m above the seabed. A thin mucilaginous layer consisting of
proteinaceous and algal forms cover the aluminium surfaces with coral growth evident on various parts
of the aircraft, especially near areas where the presence of ferrous components would encourage
more secondary colonisation. A steady and generally light current affecting the site did not visibly
move sediment.
The main wreckage is lying inverted on the seabed and consists principally of the wings with twin
engine compartments minus engines and propellers with other wreck remains, including gun turrets,
tail sections and a portion of cockpit, etc distributed over a relatively large area. Most of the aircraft
structures and components, although damaged and disconnected, remain in relatively good condition
and still retain strength and resilience. However, extensive corrosion is evident on parts of the wings
and nacelles. There was also a large anchor, with chain and cable covered in extensive coralline
growth, west of the major wing structure.
Overall damage to the aircraft structure is quite extensive and the site highly disarticulated and
scattered over a relatively large area which may indicate a catastrophic wrecking event (McKinnon and
Carrel 2011:85). However, there is evidence of recent anchor damage and over the past two years the
site has been frequented by more divers. Many smaller artefacts and components have been moved
from their original positions and piled up in one area on the site. It is also possible that some form of
salvage occurred as the engines and propellers are missing. The site is relatively shallow (7m) and
may be affected by turbulent seas generated by storms and cyclones. This aircraft is included in the
WWII Maritime Heritage Trail – Battle of Saipan and lies within a Marine Conservation Area. Divers
and snorkelers are actively encouraged to visit the site provided they follow the local visitation
guidelines and do not interfere with the site (i.e. disturb or attempt to remove any cultural or natural
components).
Corrosion Survey
The corrosion parameters of fifteen different areas on the MNR were measured over a 68 minute dive
on 20 February 2012. The results are presented in Table 25 and the on-site positions shown in Figure
52. In order to compare the corrosion data collected from the different positions measured on the MNR
and ascertain the thermodynamically stable state of the aluminium, the corrosion potentials (Ecorr) and
the pH of the residual aluminium alloy surfaces were plotted on the aluminium Pourbaix diagram in
aerobic seawater at 25°C (Figure 53). The temperature of the seawater on-site was 28°C, however
this 3°C increase does not significantly affect the nature or equilibria of the chemical species described
in this diagram.
81
Appendix A
Table 25. Corrosion parameter measurements on the MNR.
Position
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Description
port wing near tip
port wing near tip under sediment
float lying on port wing
float lying fwd of port wing
top surface fwd port engine
stbd side port engine aft 5
top surface dihedral wing stbd side
nacelle stbd wing
stbd wing near tip
float lying aft stbd wing
upper deck gun turret
tail fin
tail gun turret
plate aft fuselage
gun turret fwd port wing
pH
7.95
8.04
8.13
7.91
7.84
8.03
7.80
8.27
7.97
8.11
8.35
8.01
8.78
7.88
8.41
Ecorr vs NHE
(V)
-0.464
-0.464
-0.465
-0.450
-0.463
-0.463
-0.464
-0.414
-0.464
-0.448
-0.323
-0.439
-0.448
-0.458
-0.460
Water Depth
(m)
6.8
6.8
6.3
6.9
5.4
5.5
5.2
6.2
6.7
6.7
6.5
6.8
6.8
6.4
7.0
82
Appendix A
Figure 52. Schematic plan of the Martin PBM Mariner (MNR) indicating the corrosion parameter
measurement positions (Richards 2012 after Bell 2010 in McKinnon and Carrell 2011:78).
83
Appendix A
(a)
0.8
0.6
0.4
Al 3+
0.2
0.0
Eh(v)
-0.2
active
-
Al 2O3 .3H2 O
AlO2
passive
active
(b)
12
4
-0.4
11
5,7,14
1,9
-0.6
10
3
8
13
15
2,6
-0.8
-1.0
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 53. Pourbiax diagram for aluminium (3.7 x 10 M) in aerobic seawater at 25°C indicating
the intercepts of the areas measured on the MNR site (Richards 2012).
Generally, the areas measured on the MNR were covered with a very thin (<1mm) mucilaginous layer
consisting of proteinaceous and algal based material in combination with hydrated aluminium
hydroxide gels. Hence measurements of total depth of concretion and corrosion products (d total) are not
applicable for aluminium alloy aircraft. On some areas, especially position 7, which was the very
corroded area on the dihedral wing, patches of blue corrosion products, typical of hydrated oxidised
copper corrosion products, were observed, indicating the use of a copper - aluminium alloy, such as
Duralumin in these areas.
From the Pourbaix diagram (Figure 53), the intercepts of all points measured on the MNR, lie in the
passive region, where Al2O3.3H2O is the dominant corrosion product (Equation 2) and forms a
continuous passivating layer, effectively slowing the corrosion rate. This is a very common corrosion
state for aluminium alloy aircraft in marine environments.
The surfaces of the aluminium alloy sections of the MNR, such as the wings, floats, engine cowlings,
nacelles, etc were quite corroded and pitting and intergranular corrosion mechanisms had caused
significant pitting and perforation of the residual metal. The remains of the MNR are obviously more
deteriorated than the JAKE even though they lie in a similar environment at similar depths. However,
this is not unexpected as the remains of the JAKE are almost intact, whereas the MNR remains are
more damaged, disarticulated and spread over a wider area, therefore there would be more
intergranular corrosion occurring on the MNR aircraft remains due to increased stress and metal
fatigue.
The Ecorr values of the measurement positions, namely 1 to 7, 9, 10, 12, and 14 all fall into the average
corrosion potential range of -0.458 ± 0.008V indicating that the metal composition of these areas are
very similar. The average pH values for all these positions was 7.97 ± 0.10, however two positions, 5
and 7, had more acidic pH values of 7.84 and 7.80, respectively. This indicates that these areas, which
were on the top surfaces of the engine remains and the dihedral wing were corroding at a faster rate
than the other measurement positions. This was supported by the extensive pitting observed in these
two areas as compared to the other measurement positions. However, it is likely that the reason why
the local Ecorr of positions 5 and 7, that had these lower pH values, were similar to the potentials of the
other measurement positions is that the pH reflects the local microenvironment of the position while
the Ecorr reflects the average voltage of the corrosion cell that consists of the areas that are electrically
84
Appendix A
connected to the point of measurement. Similar behaviour was observed on the large iron shipwrecks
in Saipan.
Interestingly, the gun turrets, 11, 13 and 15 and the damaged nacelle 8, had an average pH value of
8.45 ± 0.20, which was more alkaline than the surrounding seawater at pH 8.19 and was much more
alkaline than the mean pH of 7.97 ± 0.10 for the rest of the aircraft remains. This significant increase in
pH indicates there are higher concentrations of iron associated with these positions and they are being
cathodically protected by the corroding aluminium alloy sections, which are more reactive. This is
supported by the fact that position 11, upper deck gun turret and position 8, the damaged nacelle,
which is close association with extensive engine remains had significantly less negative corrosion
potentials (-0.323V and -0.414V, respectively) than the rest of the aircraft remains (-0.458 ± 0.008V)
indicating that most of the associated aluminium alloy had corroded away and the average voltage of
the corrosion cell was moving towards the more positive corrosion potential of iron in seawater.
4.4
KAWANISHI H8K – EMILY
Figure 54. Kawanishi H8K (EMILY) (Carpenter 2012).
Date of Inspection
21 February 2012
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). On 21 February 2012 the winds were ENE at 17 to 24 knots. Seas were
choppy with a moderate long period swell (morning NNW 2m at 10 seconds; afternoon N 2m at 10
seconds). The tides were semi-diurnal over the survey period and are reported in Table 1
(http://buoyweather.com). There was a slight current (<0.5 knot) running in a NNE direction.
85
Appendix A
The in-water visibility was approximately 15m. The depth to the base of the wreck at 0834 was 8.0m.
The pH of seawater usually falls within the range of 7.5 to 8.3. The redox potential range of marine
environments is -0.300 to 0.000V in reducing environments and 0.000 to +0.250V in oxidising
environments. The pH and redox potential of the seawater on-site at 7.4m was 8.18 and 0.237V
respectively, indicating a normal, open circulation oxidising marine environment. The change in
dissolved oxygen content, salinity and temperature of the water column with depth measured on 21
February 2012 is shown in Table 26.
Table 26. Dissolved oxygen content, salinity and temperature of the seawater on the EMILY
site.
Water Depth (m)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
Average
Dissolved Oxygen
Content [ppm(S)]
7.20
7.17
7.08
7.07
7.05
7.05
7.04
7.04
7.01
7.00
6.99
7.01
6.98
6.98
6.99
6.97
6.90
6.85
6.82
7.01 ± 0.09
Salinity (ppK)
Temperature (°C)
35.9
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.8
35.7
35.8
35.8 ± 0.0
27.1
27.1
27.1
27.1
27.1
27.1
27.0
27.0
27.0
27.0
27.0
27.0
27.0
27.0
27.0
27.0
27.0
27.0
27.0
27.0 ± 0.1
There was no significant change in salinity and temperature with increasing water depth, which is
typical of the hydrology of well mixed near coastal marine waters. The average water temperature was
27.0 ± 0.1°C and the average salinity of the water column was 35.8 ± 0.0ppK, which is within the usual
salinity range for the open ocean of 32-37ppK. The average dissolved oxygen content was 7.01 ±
0.09ppm. The change in dissolved oxygen concentration with increasing water depth is shown in
Figure 55.
86
Appendix A
area. Most of the aircraft structures and components, although damaged and disconnected, remain in
relatively good condition and still retain strength and resilience. However, extensive corrosion is
evident on the nacelles.
Overall damage to the aircraft structure is quite extensive and the site highly disarticulated and
scattered over a relatively large area which may indicate a catastrophic wrecking event (McKinnon and
Carrel 2011:70). This aircraft wreck is a popular dive site, however there is little evidence of anchor
damage. It appears that the cockpit has been repositioned so divers can sit in the pilots seat and many
smaller artefacts and components have been moved from their original positions and are piled up near
the Korean and Japanese monuments present on the site. The site is relatively shallow (9m) and may
be affected by turbulent seas generated by storms and cyclones. This aircraft is included in the WWII
Maritime Heritage Trail – Battle of Saipan and divers and snorkelers are actively encouraged to visit
the site provided they follow the local visitation guidelines and do not interfere with the site (i.e. disturb
or attempt to remove any cultural or natural components).
Corrosion Survey
The corrosion parameters of sixteen different areas on the EMILY were measured over a 46 minute
dive on 21 February 2012. The results are presented in Table 27 and the on-site positions shown in
Figure 56. In order to compare the corrosion data collected from the different positions measured on
the EMILY and ascertain the thermodynamically stable state of the aluminium, the corrosion potentials
(Ecorr) and the pH of the residual aluminium alloy surfaces were plotted on the aluminium Pourbaix
diagram in aerobic seawater at 25°C (Figure 57). The temperature of the seawater on-site was 27°C,
however this 2°C increase does not significantly affect the nature or equilibria of the chemical species
described in this diagram.
88
Appendix A
Table 27. Corrosion parameter measurements on the EMILY.
Position
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Description
stbd wing tip
stbd wing midway
nacelle 1 stbd wing
nacelle 2 stbd wing
broken port wing
nacelle 3 port wing
port wing midway
port wing tip
cockpit
bow gun turret
engine 1 propeller
engine 1 boss
engine 2 bent propeller
plane part aft nacelle 2
engine 3 propeller
engine 3 boss
pH
8.15
8.13
7.92
7.94
7.93
7.52
7.94
8.04
8.14
8.02
8.19
7.97
8.03
8.27
8.11
8.06
Ecorr vs NHE
(V)
-0.474
-0.475
-0.474
-0.475
-0.475
-0.474
-0.475
-0.475
-0.465
-0.472
-0.450
-0.450
-0.452
-0.461
-0.465
-0.465
Water Depth
(m)
8.7
8.0
7.6
7.5
6.8
7.6
8.5
9.0
8.4
8.9
7.5
7.1
8.3
8.2
7.8
7.4
89
Appendix A
Figure 56. Schematic plan of the Kawanishi H8K (EMILY) indicating the corrosion parameter
measurement positions (Richards 2012 after Bell 2010 in McKinnon and Carrell 2011:63).
(a)
0.8
0.6
0.4
Al 3+
0.2
active
-
Al 2O3 .3H2 O
AlO2
passive
active
0.0
Eh(v)
-0.2
(b)
12,13
3,4,5,7
-0.4
6
-0.6
11
8,10
14
1,2,9,15,16
-0.8
-1.0
2
3
4
5
6
7
8
9
10 11 12
pH
-6
Figure 57. Pourbiax diagram for aluminium (3.7 x 10 M) in aerobic seawater at 25°C indicating
the intercepts of the areas measured on the EMILY site (Richards 2012).
Generally, the areas measured on the EMILY were covered with a very thin (<1mm) mucilaginous
layer consisting of proteinaceous and algal based material in combination with hydrated aluminium
hydroxide gels. Hence measurements of total depth of concretion and corrosion products (d total) are not
applicable for aluminium alloy aircraft. Unlike the AVR and MNR, there was no observed evidence of
the typical blue copper corrosion products on the EMILY.
From the Pourbaix diagram (Figure 57), the intercepts of all points measured on the EMILY, lie in the
passive region, where Al2O3.3H2O is the dominant corrosion product (Equation 2) and forms a
continuous passivating layer, effectively slowing the corrosion rate. This is a very common corrosion
state for aluminium alloy aircraft in marine environments.
The surface of the aluminium alloy sections of the EMILY wing were generally in good condition with
the exception of the nacelles where discrete areas of pitting and perforation of the residual metal were
obvious, however the extent was significantly less than that observed on the AVR and MNR,
suggesting pitting corrosion was the preferred corrosion mechanism in this environment with
considerably less intergranular corrosion occurring.
The Ecorr values of the measurement positions 1 to 8 and 10 all fall into the average corrosion potential
range of -0.474 ± 0.001V indicating that the metal composition of these areas are very similar. This
would not be unexpected as positions 1 to 8 were all part of the wing structure and position 10 was the
bow gun turret, which would have been connected to this area in the past. The cockpit (9), the plane
section aft of nacelle 2 (14) and the propeller (15) and boss (16) of engine 3 had an average corrosion
potential of -0.464 ± 0.002V, whilst the boss of engine 1 (12) and the propellers of engine 1 (11) and 2
(13) had an average corrosion potential of -0.451 ± 0.001. These more positive average Ecorr values
indicate that the metal composition of these positions are different to the wing section and probably
contain more noble minor alloying constituents, such as copper and iron (e.g. the cast propellers and
bosses) and/or are electrically connected to different metal components (e.g. the cockpit). The 13mV
increase (more positive) in the Ecorr of positions 11 to 13 is readily explained by the fact that engines 1
and 2 were considerably more intact and the propellers and bosses were electrically connected to
90
Appendix A
engine components of differing metal composition (e.g. iron and copper). On the other hand, only
small amounts of iron were associated with the cockpit (9) and the plane section aft of nacelle 2 (14)
and the propeller (15) and boss (16) of engine 3 were separated from the main engine components.
The average pH value for positions 1 to 8 and 10 was 7.95 ± 0.17 but positions 9, 11 to 16, which
possessed more positive Ecorr values, had a more alkaline average pH value of 8.11 ± 0.09. This
increase in pH indicates that the corrosion rate in these areas is less than on the wing and turret and it
is likely these areas are being cathodically protected to some extent and/or the composition of the
metal contains less reactive elements, such as iron and copper, which is in agreement with the
conclusions based on the corrosion potential data.
4.5
CONSOLIDATED PB2Y CORONADO - CRDO
Figure 58. Coronado (CRDO) (Carpenter 2012).
Date of Inspection
22 February 2012
Environmental Conditions
Generally fine weather conditions with an average daily temperature of 29ºC over the survey period
(20-24 February 2012). On 22 February 2012 the winds were ENE at 18 to 24 knots in the morning
tending ENE at 17 to 23 knots in the afternoon. Seas were choppy with a moderate short period swell
(morning E 2.2m at 9 seconds; afternoon ENE 2.1m at 9 seconds). The tides were semi-diurnal over
the survey period and are reported in Table 1 (http://buoyweather.com). There was a slight current
(<0.5 knots) running in a NNE direction.
The in-water visibility was approximately 20m. The depth to the base of the wreck at 1234 was 7.7m.
The pH of seawater usually falls within the range of 7.5 to 8.3. The redox potential range of marine
91
Appendix A
environments is -0.300 to 0.000V in reducing environments and 0.000 to +0.250V in oxidising
environments. The pH and redox potential of the seawater on-site at 7.7m was 8.06 and 0.263V
respectively, indicating a normal, open circulation oxidising marine environment. The change in
dissolved oxygen content, salinity and temperature of the water column with depth measured on 23
February 2012 is shown in Table 28.
92
Appendix A
increasing aerobic respiration of plankton in the photosynthetic zone. However, the relatively small
standard deviation between the measurements and a decrease of only 0.38ppm over the 9.0m depth
range indicates that there is very little variation in the dissolved oxygen content with increasing water
depth over such a shallow depth range, which is not unexpected. Hence, this trend coupled with the
other physico-chemical measurements, are typical for a shallow, near coastal, open circulation,
oxidising marine environment, where corrosion rates are likely to be relatively high.
Wreck Site
The Coronado (CRDO) is located on the south western side of Saipan, inside Tanapag Lagoon at a
depth of about 7m. The aircraft remains were positively identified as a Consolidated PB2Y Coronado
in 2012, a U.S. four engine maritime patrol flying boat, constructed most probably of duralumin
(aluminium alloy containing 3-5% Cu, 0.4-1.0% Mn, 0.3-0.6% Mg) and is 24.16m in length, 8.38m high
and had a 35.05m wingspan (http://www.flugzeuginfo.net). The aircraft remains are totally submerged
at all times.
The aircraft remains are located on a relatively flat, undulating seabed comprising primarily of fine
calcareous sediment. There are sporadic large coral outcrops in close vicinity to the wreck and a very
large reef formation lies to the south east of the major site. Some lower profile sections, such as the
wings and tail planes were covered in a thin layer of fine sediment. Most of the remains are exposed
and total burial seems unlikely to occur. The maximum exposure height of most remains is
approximately 50cm with the engine being the exception rising about 1.5m above the seabed. A thin
mucilaginous layer consisting of proteinaceous and algal forms cover the aluminium surfaces with
coral growth evident on various parts of the aircraft, especially near areas where the presence of
ferrous components would encourage more secondary colonisation. A steady and generally light
current affecting the site did not visibly move sediment.
The aircraft remains are disconnected and scattered over a very large area. Among the components
identified were a single detached engine, a rectangular box structure with dials, etc, aerial mast,
cockpit canopy with windscreen wipers, hatch covers, float support, a chair, concreted forks and a
number of unidentifiable hull sections. Overall damage to the aircraft structure is quite extensive and
the site highly disarticulated and scattered over a relatively large area which may indicate a
catastrophic wrecking event. The site is relatively shallow (7m) and may be affected by turbulent seas
generated by storms and cyclones. This aircraft is not included on the WWII Maritime Heritage Trail –
Battle of Saipan therefore visitation to the site by divers and snorkelers would be less than to those
wrecks that are listed on the heritage trail. The fact that smaller artefacts (e.g. chair, forks, etc) are still
present in situ supports the fact that this site is not visited frequently and human interference has been
minimal to date.
Corrosion Survey
The corrosion parameters of fourteen different areas on the CRDO were measured over a 69 minute
dive on 22 February 2012. The results are presented in Table 29 and the on-site positions shown in
Figure 60. In order to compare the corrosion data collected from the different positions measured on
the EMILY and ascertain the thermodynamically stable state of the aluminium, the corrosion potentials
(Ecorr) and the pH of the residual aluminium alloy surfaces were plotted on the aluminium Pourbaix
diagram in aerobic seawater at 25°C (Figure 61). The temperature of the seawater on-site was 28°C,
however this 3°C increase does not significantly affect the nature or equilibria of the chemical species
described in this diagram.
Table 29. Corrosion parameter measurements on the CRDO.
Position
Number
1
2
3
4
5
6
Description
wing tip
stbd wing
fuselage
fuselage
cockpit
fuselage
pH
7.87
8.05
7.96
7.83
7.97
7.93
Ecorr vs NHE
(V)
-0.459
-0.464
-0.469
-0.454
-0.461
-0.460
Water Depth
(m)
7.6
7.2
7.3
7.5
7.4
7.6
94
Appendix A
7
8
9
10
11
12
13
14
float strut
unidentified piece of wreckage
unidentified piece of wreckage
hatch cover
unidentified piece of wreckage
engine
control panel
float ?
8.08
8.01
7.95
8.07
7.93
7.89
7.93
7.99
-0.499
-0.451
-0.451
-0.466
-0.480
-0.460
-0.483
-0.451
7.6
7.3
7.5
7.7
7.5
6.3
7.3
7.8
95
Appendix A
Figure 60. Schematic plan of the Consolidated PB2Y Coronado (CRDO) indicating the corrosion
parameter measurement positions (Richards 2012 after Harvey and Raupp 2012).
96
Appendix A
(a)
0.8
0.6
0.4
Al 3+
0.2
0.0
Eh(v)
-0.2
active
-
Al 2O3 .3H2 O
AlO2
passive
active
(b)
1,4,12
-0.4
11,13
-0.6
8,9,14
2,10
8
9
7
3,5,6
-0.8
-1.0
2
3
4
5
6
7
10 11 12
pH
-6
Figure 61. Pourbiax diagram for aluminium (3.7 x 10 M) in aerobic seawater at 25°C indicating
the intercepts of the areas measured on the CRDO site (Richards 2012).
Generally, the areas measured on the CRDO were covered with a very thin (<1mm) mucilaginous
layer consisting of proteinaceous and algal based material in combination with hydrated aluminium
hydroxide gels. Hence measurements of total depth of concretion and corrosion products (d total) are not
applicable for aluminium alloy aircraft. Unlike the AVR and MNR, there was no observed evidence of
the typical blue copper corrosion products on the CRDO.
From the Pourbaix diagram (Figure 61), the intercepts of all points measured on the CRDO, lie in the
passive region, where Al2O3.3H2O is the dominant corrosion product (Equation 2) and forms a
continuous passivating layer, effectively slowing the corrosion rate. This is a very common corrosion
state for aluminium alloy aircraft in marine environments.
The surfaces of the aluminium alloy sections of the CRDO were quite corroded and pitting and
intergranular corrosion mechanisms had caused significant pitting and perforation of the residual
metal. The remains of the CRDO are obviously more deteriorated than the JAKE and EMILY even
though they lie in a similar environment at similar depths. However, this is not unexpected as the
CRDO remains are more damaged, disarticulated and spread over a wider area, therefore there would
be more intergranular corrosion occurring on the CRDO aircraft remains due to increased stress and
metal fatigue.
The Ecorr values of the measurement positions 1 to 6, 8 to 10, 12 and 14 all fall into the average
corrosion potential range of -0.459 ± 0.006V indicating that these areas have similar metal
compositions and are corroding at a similar rate. The average corrosion potential of positions 7 (float
strut), 11 (unidentified wreckage) and 13 (control panel) was -0.487 ± 0.008V, which is 28mV more
negative than the other positions indicating that these areas have different metal compositions (i.e.
higher aluminium contents) and are therefore corroding at a higher rate. Unfortunately the average pH
values of all groups of measurements all fell within the statistical standard sample deviation so
differences in pH cannot be used as an indication of differences in corrosion rates.
97
Appendix A
5
CONCLUSIONS
In general, the physico-chemical measurements (pH, Eredox, dissolved oxygen, salinity, temperature,
etc) of the local environment surrounding the wreck sites in Saipan are typical for a shallow, near
coastal, open circulation, oxidising marine environment, where corrosion rates are likely to be relatively
high for both ferrous alloy wrecks and aluminium alloy aircraft.
All of the wrecks and the aircraft were mostly exposed with only very thin layers of sediment covering
some lower profile areas lying on the seabed, which would be particularly mobile during periods of
excessive water movement (i.e. storm and cyclonic activity). Hence, natural protection via seasonal
sediment burial would be very unlikely for any of the wrecks surveyed in 2012.
The corrosion parameters of a number of different areas on each of the ten iron alloy wrecks in Saipan
were measured during the survey period from 20-24 February 2012. In order to compare the corrosion
data collected from the different positions measured on the iron wrecks the corrosion potentials (Ecorr)
and the pH of the residual iron alloy surfaces were plotted on the iron Pourbaix diagram in aerobic
seawater at 25°C. Generally, the intercepts of all points measured on the iron alloy wrecks either lay in
the active corrosion region, where ferrous ions are the thermodynamically stable chemical species and
corrosion will continue until all iron is consumed, lay on the equilibrium line between active corrosion
and the passive region, which implies that the typical aerobic corrosion mechanism where the major
2+
stable chemical species is the ferrous ion (Fe ) is in equilibrium with the formation of an insoluble
corrosion product layer of magnetite (Fe3O4) or lay in the passive magnetite region (Fe 3O4) indicating
there was very little if any residual metal remaining in those areas. Generally, with film free corrosion
mechanisms, such as occurs on concreted iron artefacts, an increase in the corrosion potential
(tending more positive) indicates an increase in the corrosion of the metal.
However, since Ecorr data describes the electrochemical environment of the iron alloy that is electrically
connected to the measurement point and as such, it is not as sensitive to changes in localised
corrosion processes as the value of the pH recorded at the same point, provided no damage has
occurred to the protective concretion layer. It has been shown that pH data is a useful guide to
changes in corrosion rate, since as the corrosion rate increases the pH decreases (becomes more
acidic). It has also been observed that the pH of corroding residual metal surfaces decrease linearly
with increasing total thickness of the corrosion product layer and the encapsulating concretion (d total).
That is, generally the thicker the dtotal, the lower the surface pH but only if the concretion layer remains
essentially undisturbed (i.e. no damage occurs through human/natural interference).
The average corrosion parameters (Ecorr, pH values and dtotal) of all measurement points on each iron
wreck are shown in Table 30. However, for many of the wrecks there are no statistically valid
differences between the average corrosion parameter measurements as they fall within the
maxima/minima range calculated from the standard deviations for each set of data points, making it
difficult to determine any differences in corrosion rates between the wrecks based on the corrosion
parameter data. However some conclusions can be drawn if only based on some of the corrosion
parameter data in conjunction with the environmental and historical information.
Table 30. Average corrosion parameter measurements for all iron alloy wrecks measured in
Saipan, CNMI from 20-24 February 2012.
Wreck
Sherman Tank 1
Sherman Tank 3
LVT1
LVT2
DAI1
DAI2
Average
Corrosion
Potential vs NHE
(V) (all points)
-0.305 ± 0.003
-0.320 ± 0.003
-0.322 ± 0.002
-0.295 ± 0.028
-0.334 ± 0.009
-0.325 ± 0.002
Average pH
(acidic)
6.17 ± 0.32
6.09 ±0.00
Average pH
(alkaline)
Average pH
(all points)
Average
dtotal
(mm)
7.68 ± 0.36
6.68 ± 0.05
6.86 ± 0.09
6.34 ± 0.33
6.40 ± 1.18
7.13 ± 0.79
7.22 ± 0.80
6.39 ± 0.34
19 ± 11
6±4
18 ± 7
7±4
9±7
11 ± 12
98
Appendix A
DAI3
JFR
ASC
SS
-0.338 ± 0.001
-0.339 ± 0.007
-0.327 ± 0.004
-0.355 ± 0.003
6.34 ± 0.18
7.50 ± 0.42
7.15 ± 0.66
7.29 ± 0.68
7.34 ± 0.72
7.83 ± 0.49
6±4
5±4
6±7
4±3
The average corrosion potential for Tank 1 was -0.305 ± 0.003V, which is only 15mV more positive
than Tank 3, hence, it is not possible to determine any differences in corrosion behaviour between the
measurement points on Tank 3 and between Tank 1 and Tank 3 based on the Ecorr data. The decrease
of 0.52 pH units for Tank 3 indicates that there has been a statistically significant increase in the
corrosion rate of Tank 3 as compared to Tank 1 and in conjunction with the thinner d totals measured on
Tank 3, suggests that the natural and cultural impacts of the local environment on Tank 3 are more
aggressive than those experienced by Tank 1. More importantly, as there appears to be more tourist
activity associated with Tank 3, it may be this increase in human interference that is causing the
accelerated deterioration of Tank 3.
It is difficult to say whether the LVT2 is corroding at a faster rate than the LVT1 as all average
measurements are within their respective statistical errors. However, considering the extent of
deterioration of the LVT2 as compared to the LVT1 it would appear that the natural and cultural
impacts on the LVT2 would be greater than those experienced by the LVT1.
Based on the average pH values of the more alkaline positions on the Daihatsu wrecks, which were
6.68 ± 0.05 on DAI2 and 7.68 ± 0.36 and 7.50 ± 0.42 on DAI1 and DAI3, respectively some differences
in corrosion rate can be ascertained. The decrease in average pH of DAI2 suggests that it may be
corroding at a slightly faster rate than both DAI1 and DAI3. This is not unexpected as it is known that
isolated iron artefacts and steel hull structures that have been damaged either through natural
phenomena (e.g. cyclonic activity) or human intervention (e.g. salvage, explosive damage during
WWII) possess higher corrosion rates than those hull structures that are relatively intact (i.e. DAI1 and
DAI3), where the current density of the corrosion process can be spread over a much larger surface
area effectively lowering the corrosion rate. In addition, it appears that DAI1 and DAI3 are corroding at
relatively similar rates, despite the fact that DAI3 is a much shallower site, where it would be expected
that the corrosion rate would be slightly higher. This would seem to suggest that human interference
(i.e. recreational diving activities) is having some impact on the deterioration rate of the deeper DAI1
site.
Based on the corrosion parameter measurements it is difficult to determine any changes in corrosion
behaviour of the larger shipwrecks, JFR, the ASC and the SS as most average measurements are
within their respective statistical errors. However, based on the average corrosion potentials of the
JFR (-0.339 ± 0.007V), the ASC (-0.327 ± 0.004V) and the SS (-0.355 ± 0.003V) it appears that the
small but statistically valid decrease in the corrosion potential (E corr) of the SS suggests that this vessel
is corroding at a slower rate than both the ASC and the JFR and the small increase in the average E corr
of the ASC suggests that it may be corroding at a slightly faster rate than JFR. This is not unexpected
as steel hull structures that have been extensively damaged (i.e. ASC) possess higher corrosion rates
than those hull structures that are relatively intact (i.e. JFR and the SS). This would seem to suggest
that human interference (i.e. recreational diving activities) is having some impact on the deterioration
rate of the JFR and ASC sites as the SS site is not on the diving heritage trail. However, the local
environment (i.e. increase in turbidity) may also be contributing to this decrease in the corrosion rate
on the SS site.
The corrosion parameters of a number of different areas on each of the five aluminium alloy aircraft
wrecks in Saipan were measured during the survey period from 20-24 February 2012. In order to
compare the corrosion data collected from the different positions measured on the aircraft the
corrosion potentials (Ecorr) and the pH of the residual aluminium alloy surfaces were plotted on the
aluminium Pourbaix diagram in aerobic seawater at 25°C. From these Pourbaix diagrams, the
intercepts of all points measured on all aircraft, lie in the passive region, where Al 2O3.3H2O is the
dominant corrosion product and forms a continuous passivating layer, effectively slowing corrosion
rates. This is a very common corrosion state for aluminium alloy aircraft in marine environments.
However this particular Pourbaix diagram is only applicable to pure aluminium in seawater and most of
the aircraft manufactured during WWII used a variety of aluminium alloys consisting mainly of
aluminium but including varying concentrations of minor alloying constituents (e.g. iron, copper,
99
Appendix A
magnesium, manganese, zinc and silicon) in order to change the functionality of the aluminium. One of
the most common alloying metals used was copper (e.g. Duralumin) which was added to aluminium to
increase its strength, however the presence of the copper dramatically decreased the corrosion
resistance of the metal to seawater. The other issue that will increase the deterioration rates of the
aircraft is galvanic corrosion, where the more reactive aluminium alloys will corrode faster effectively
protecting the more noble metals, such as iron and copper. All these issues combined makes it
extremely difficult to determine any differences in corrosion rates based on the corrosion parameter
data. For example, the average corrosion potential and pH values for all measurement points for each
wreck are shown in Table 31 (columns 2 and 3) and from this data it is obvious that there are no
statistically valid differences between any of the average corrosion parameter measurements as all fall
within the maxima/minima range calculated from the standard deviations for each set of data points.
Table 31. Average corrosion parameter measurements for all aluminium alloy aircraft wrecks
measured in Saipan, CNMI from 20-24 February 2012.
Wreck
Avenger (AVR)
Aichi E13A (JAKE)
Mariner (MNR)
Kawanishi H8K (EMILY)
Coronado (CRDO)
Average
Corrosion
Potential vs
NHE (V)
(all points)
-0.440 ± 0.021
-0.443 ± 0.011
-0.446 ± 0.037
-0.467 ± 0.009
-0.465 ± 0.014
Average
pH
(all points)
8.06 ± 0.11
7.99 ± 0.09
8.10 ± 0.26
8.02 ± 0.17
7.96 ± 0.07
Average
Corrosion
Potential vs
NHE (V)
(major group)
-0.438 ± 0.001
-0.446 ± 0.003
-0.458 ± 0.008
-0.474 ± 0.001
-0.459 ± 0.008
Average
pH
(major
group)
Average
Water
Depth
(m)
8.08 ± 0.08
8.01 ± 0.07
7.97 ± 0.10
7.95 ± 0.17
7.95 ± 0.07
2.2 ± 0.9
6.1 ± 0.3
6.4 ± 0.6
8.0 ± 0.6
7.4 ± 0.4
However, since all aluminium alloys are corroding in a common oxidising marine environment in
Tanapag Lagoon, the different values of the corrosion potentials may provide a guide to the underlying
differences in alloy composition of the aircraft. Since the corrosion potentials of aluminium alloys
containing higher concentrations of less reactive metals, such as copper, become more positive (more
anodic) it is possible to determine differences in the metal compositions of the major structural
components of the aircraft. Hence, if the average corrosion potentials of the largest group of
measurement points with similar Ecorr values on each of the wrecks are compared (Table 31 - column
4) some differences between the aircraft metal compositions become more apparent. Unfortunately,
again the average pH values for the same group of points are within statistical error and thus, cannot
be used in the interpretation. So based on this average E corr data, the metal composition of the
aluminium alloys for each aircraft, in order of decreasing concentrations of incorporated copper (or
other less reactive metals) is Avenger > Jake > Mariner ~ Coronado > Emily. That is the Avenger may
have the highest concentration of copper in this group of aluminium alloys measured whilst the Emily
will have the lowest based on this data set. This may have consequences for the corrosion rates of
these aircraft as higher concentrations of copper will increase the rate of pitting and intergranular
corrosion if the aircraft are subjected to similar environmental conditions and other complicating
factors, such as increases in corrosion through stress and metal fatigue, are absent. Since this is not
the case with these aircraft (i.e. the Avenger lies in a very aggressive, shallower marine environment
and the Coronado is extensively damaged with separate sections strewn over a very large area) this
highlights the problem with interpreting corrosion data based on only one set of corrosion parameter
measurements.
In conclusion, it is obvious that there are problems with determining differences in corrosion behaviour
of wrecks based only on one set of corrosion parameter measurements. A holistic approach must be
taken using all the data obtained including the environmental and historical information in order to
understand the corrosion processes occurring on a wreck site. Hence, continued observation of the
sites and further corrosion measurements in the future may assist in corroborating or refuting the
aforementioned inferences.
100
Appendix A
6
RECOMMENDATIONS
The submerged shipwrecks and aircraft wrecks located in Saipan are a significant part of World War II
history and are one of the main tourist attractions in Saipan. It is therefore important that appropriate
management plans are implemented to ensure the future preservation of these sites. The scope of the
work includes ongoing monitoring of the status of the natural and cultural attributes of the wrecks and
the integrity of these underwater archaeological sites. Specific guidelines detailing the range of
corrosion aspects to be documented on a regular basis will empower Saipan HPO field staff to
implement regular and effective monitoring surveys integral for their future preservation. The
guidelines will provide consistent and comparative data, which will assist in the implementation of any
future conservation management strategies.
Regular site inspections are an integral part of the overall management strategy for a submerged site.
The primary focus of an on-site corrosion survey is to collect as much pertinent information as possible
to assist in ascertaining the extent of deterioration and structural integrity of a site. Further inspections
are then required at regular intervals and especially after any severe storm or cyclonic activity so any
changes in the integrity of the site are noted by direct comparison with earlier surveys. The more
surveys carried out the better as it will provide more information regarding the rate of deterioration and
the inherent stability of a site, which will assist in recognising which sites are a priority for future
implementation of appropriate in situ conservation management strategies.
The first step in implementing any corrosion survey is to gather the information outlined in the On-Site
Corrosion Survey Data Sheet (Appendix B). Much of this information is self explanatory but some
basic explanations and examples of some of the criteria included in this form are described below.
Weather and Sea Conditions; Swell and Tidal Information; Current – The amount of oxygen
impingement to the surface of a metal will directly affect the corrosion rate. Without direct access to
probes to measure the dissolved oxygen concentration in the water column it is imperative that the
amount of water movement on a site is documented. For example, any increase in water movement
(increased swell, tidal movement, current, etc) will increase the amount of oxygen available to a metal
surface and in turn, increase the corrosion rate.
Water Temperature – The effects of water temperature on corrosion rates is complicated by its effect
on biological growth, however in the absence of biological considerations the rate of corrosion would
be expected to double with every 10°C rise in water temperature. On the other hand, increases in
water temperature will increase the growth rate of encrusting organisms and the depth of the
concretion layer on the metal surface, which may reduce the corrosion rate. In addition, the
concentration of dissolved oxygen decreases with increasing temperature, therefore it is important to
measure the water temperature on-site and when possible, the annual ranges in an area should also
be noted.
Water Depth to Wreck (minimum, maximum) – The depth range of the submerged site from the
shallowest to the deepest section, include the depths of any large structural features (e.g. the
shallowest section of the Freighter is the top of the bow at 2m, the major structure at 5-7m and the
seabed is 11m). The depth of a site may have an influence on the corrosion rate because in general,
as water depth increases the amount of water movement decreases, decreasing the amount of oxygen
availability to a metal surface and hence, the corrosion rate. In addition, changes in the maximum and
minimum water depth are a simple way to monitor the overall collapse of the vessel on the seabed.
Visibility – This should be an approximation. Visibility on submerged sites is quite variable and
influenced by many factors, some of which can affect the deterioration rate of sites. For example,
increased water movement can lift sediment into the water column, which can then essentially
sandblast the metal surface and rapidly erode any protective corrosion/concretion layers, thereby
increasing the corrosion rate. Alternatively, sites where the visibility is more often than not, poor may
discourage diving activity and therefore decrease damage by limiting human disturbance.
101
Appendix A
Distance from Land/Reef – The distance and direction of a submerged site from land, reef or manmade construction can have an influence on the amount of water movement a site experiences and
hence, the corrosion rate. For example, a site located in the lee of an island may be protected from
seasonal increases in water movement (e.g. during monsoons, typhoons, etc) effectively lowering the
average corrosion rate. Alternatively a site located adjacent to a reef may experience increased water
movement and hence, an increased corrosion rate.
Freshwater Influence – In general, metal corrosion rates decrease with decreasing salinity, hence
metals in freshwater are generally better preserved than those located in marine environments.
Therefore if there is a large increase in the volume of freshwater from rain water run off, rivers, etc on
a site, the salinity will decrease thus reducing the corrosion rate.
Site Dimensions (area) – This is essentially a measure of the scatter of debris on a submerged site
which may have an influence on localised turbulence and increase the surface area exposed to
dissolved oxygen and hence, change the average corrosion rates of the different sections scattered
over a site.
Site Orientation – The orientation of a submerged site can affect the corrosion rate by changing the
amount of water movement around a site. For example, a wreck that has a list to port may show signs
of increased corrosion on the more exposed starboard side compared to the more protected port side
of the vessel (e.g. Jake aircraft). Another point that has to be considered, especially on wrecks that are
not upright (e.g. Auxiliary Submarine Chaser), is the increase in stress on the vessel’s hull structure
causing increased corrosion rates in the long term.
Composition of Dominant Wreck Material – It is important to identify the dominant material/s a
submerged site primarily consists of as it will have a significant effect on the type and amount of
biological growth on the metal surface, the primary corrosion mechanisms and hence, the corrosion
rates. For example, iron promotes biological growth and is characterised by relatively thick concretion
layers and significant amounts of secondary marine growth, such as corals, etc. This semi-permeable
protective layer essentially changes the nature of the local micro-environment from that of normal
seawater and effectively slows down the rate of corrosion. On the other hand, aluminium is biologically
inert and generally characterised by little marine growth, often only being covered by a thin gelatinous
layer of corrosion products and marine algae.
Another factor to consider is galvanic corrosion. When two dissimilar metals are in direct electrical
contact with each other the more active metal (e.g. aluminium) will corrode faster than normal and the
other more noble metal (e.g. iron) will be protected. For example, aluminium in direct physical contact
with iron will corrode at a faster rate than just aluminium on its own and therefore the structural
integrity of the aluminium of the galvanic couple will deteriorate at a faster rate.
Dominant Encrusting Organisms on Surface (type, abundance, photograph) – This need only be a
very general survey, photographically documenting the main encrusting organisms present on the
dominant material types (e.g. iron, aluminium, etc) comprising the submerged site. If the site is large or
scattered over a large area then fully document a few areas (e.g. bow, midships and stern) that can be
monitored at regular intervals in the future. The type and abundance of colonising organisms can have
a significant effect on the rate of corrosion of metals and the degradation rate of organic materials. For
example, a relatively thick concretion layer may decrease the corrosion rate of iron by effectively
separating the metal surface from the seawater and protecting the underlying metal from physical
damage but conversely, areas that are covered in large, very prominent encrusting organisms may
increase localised water turbulence and this in turn, may cause an increase in the corrosion rate. In
addition, documenting any areas where changes have occurred (i.e. through storm damage or human
interference) can assist in monitoring the rate of recolonisation. If the damage is extensive then fully
document a few areas that can be monitored at regular intervals in the future.
Evidence of Active Corrosion – Evidence of active corrosion on iron is characterised by the typical
red/brown coloured corrosion products (rust). It is more difficult to identify active corrosion on
aluminium due to the protective oxide layer that forms on the metal surface and often the first sign of
active corrosion is total perforation of thinner structural plates, however sometimes it can be
102
Appendix A
characterised by localised areas of white/grey pustules. In addition, copper-aluminium alloys, such as
Duralum suffer from extensive corrosion in seawater. In this case, active corrosion is characterised by
a combination of the white/grey aluminium oxide pustules and the typical blue/green copper corrosion
products. If the active corrosion is relatively uniform over a site then fully document a few areas (e.g.
bow, midships and stern) that can be monitored at regular intervals in the future. It is imperative that
these areas of active corrosion are accurately documented in the initial survey (water depth, general
description of position and photographic documentation) so the information gathered on any
subsequent surveys can be directly compared to this baseline survey so any changes in the number
and/or extent of the active areas can be noted. Obviously, a submerged site exhibiting increased
active corrosion indicates that there is an increase in the corrosion rate.
Evidence of Damage – Damage caused to submerged sites by human interference and/or periods of
excessive water movement (storms, typhoons, etc) is easily identified by large areas of exposed metal
generally devoid of secondary marine growth. Often the metal will show signs of active corrosion. It is
imperative that these damaged areas are accurately documented in the initial survey (water depth,
general description of position and photographic documentation) so the information gathered on any
subsequent surveys can be directly compared to this baseline survey so any changes in the corrosion
activity, extent of colonisation, etc can be noted in the future. A submerged site with large expanses of
damage will exhibit increased localised corrosion rates.
Evidence of Structural Collapse – It is imperative that a submerged site is accurately documented over
its entire length during the initial survey, concentrating on areas that would be more prone to structural
collapse. In this way, any changes in the structural integrity of a site can be accurately monitored.
Evidence of Human Disturbance – It is imperative that a submerged site is accurately photographically
documented over its entire length and breadth during the initial survey so any evidence of human
disturbance (e.g. broken corals caused by diver damage, damage due to inappropriate anchoring
procedures, removal of artefacts, etc) can be monitored in the future. If feasible, photographically
document any conglomeration of artefacts during the initial survey so any changes in the condition of
the artefacts or more importantly, removal of the artefacts from the site can be monitored during
subsequent surveys.
In addition, some general points to consider when performing on-site corrosion surveys are outlined
below.
1.
2.
3.
It may be advisable to conduct the first dive on any submerged site as a reconnaissance
survey on order to plan what information and documentation is actually required for the
initial survey and how the survey will be carried out.
It is imperative that an initial survey is carried out on every submerged site so the
information gathered during subsequent surveys can be directly compared to this baseline
survey. This is necessary in order to ascertain if any changes have occurred to the site
with respect to its corrosion state (increases in the extent and number of areas exhibiting
active corrosion; changes in structural integrity; changes in the extent of damage due to
human disturbance, such as anchor damage, salvage, pollution, etc) in the future. The
baseline survey must be as comprehensive as possible then it be used as the basis for
subsequent surveys where any changes that occur are documented rather than
duplicating the initial survey.
One of the most important aspects of the corrosion survey is the photographic
documentation of any changes that occur on a site. This will then allow meaningful
comparisons to be made in the future to ascertain if any significant changes have occurred
to the corrosion rate and structural integrity of a particular site.
Without access to the underwater corrosion equipment, the most important aspect of the regular site
inspections is the photographic documentation of any changes that occur on a site. This will then allow
meaningful comparisons to be made in the future to ascertain if any significant changes to a particular
site have occurred. In addition, it is recommended that another full corrosion and environmental survey
using the underwater corrosion survey equipment is performed in another few years. In this way, from
comparisons of the regular site inspection results and the additional corrosion parameter data for each
103
Appendix A
wreck site, it will be possible to ascertain if there is indeed any effect from diving tourism on the sites
and if it is at all comparable to the detrimental effects afforded by natural occurrences, such as
seasonal storm and cyclonic activity. Finally, using a combination of information gathered from these
surveys it will be possible to prioritise these submerged sites with respect to their overall in situ
management requirements and the most appropriate management plans determined and applied to
each site.
104
Appendix A
7
REFERENCES
Denton, G.R.W., Concepcion, L.P., Galt Siegrist, H., Vann, D.T., Wood, H.R., and Beardon, B.G.
2001. Contaminant Assessment of Surface Sediments from Tanapag Lagoon, Saipan, Unpublished
Technical Report No. 93, University of Guam, Water and Environmental Research Institute of the
Western Pacific, Guam.
Grove, E. 1976. World War II Tanks. Second Edition. Excalibur Books, New York, NY.
La Que, F.L. 1975. Marine Corrosion, John Wiley, New York, NY.
MacLeod, I.D., 1998. In situ corrosion studies of iron and composite wrecks in South Australian
waters: implications for site managers and cultural tourism, Bulletin of the Australasian Institute for
Maritime Archaeology, 22:81-92.
MacLeod, I.D., 2006. In situ corrosion studies on wrecked aircraft of the Imperial Japanese Navy in
Chuuk Lagoon, Federated States of Micronesia, The International Journal of Nautical Archaeology,
35:1-9.
MacLeod, I.D., Berger, M., Richards, V.L., Jeffery, W. and Hengeveld, M. 2007. Dynamic interaction of
marine ecosystems with wrecks in Chuuk Lagoon, Federated States of Micronesia in Metal 07:
Proceedings of the Interim Meeting of the ICOM-CC Metal Working Group, Amsterdam, 17-21
September 2007, eds C. Degrigny, R. van Langh, I. Joosten and B. Ankersmit, Rijksmuseum,
Amsterdam, pp. 51-54.
MacLeod, I.D. and Richards, V. 2011. In situ conservation surveys of iron shipwrecks in Chuuk Lagoon
and the impact of human intervention, Australian Institute for the Conservation of Cultural Material
Bulletin, 32:106-122.
McKinnon, J. and Carrell, T. 2011. Saipan WWII Invasion Beaches Underwater Heritage Trail.
Unpublished final report for Grant Agreement No. GA-2255-09-028, Department of the Interior,
National Park Service, Washington, DC.
Richards, V., Carpenter, J. and Kasi, K. 2011. Shipwrecks of the Ningaloo Reef - Conservation
Surveys 1992-2009 in Shipwrecks of the Ningaloo Reef: Maritime Archaeological Projects from 19782009, ed. J. Green, Special Publication No. 15, Australian National Centre of Excellence for Maritime
Archaeology, Department of Maritime Archaeology, Western Australian Museum, Fremantle, pp. 1264.
105
Appendix A
APPENDIX A
ON-SITE CONSERVATION SURVEY DATA SHEET
Date of Survey
Time of Survey
Aim of Survey
Personnel
Site (name, date and type)
Location
Distance from Land/Reef
Site Classification
Site Dimensions (length, width, area)
Site Orientation
Seabed Topography
Marine Macrofauna and Flora (type and abundance) (photograph)
Wreck Specific Types of Marine Life (photograph)
Composition of Dominant Wreck Material (in situ observation, cargo influence)
Exposed Artefacts (type, material, apparent condition, degree of completeness, distribution)
Degree of Site Exposure (area, height above seabed)
Evidence of Seasonal Exposure
Evidence or Potential for Storm, Cyclone Influence
Evidence of Human Disturbance (salvage, pollution, modern contamination, water activities)
Weather Conditions
Sea Conditions
Swell
Current (rate, direction, speed)
Tidal Information
Freshwater/Saltwater Influence (rivers, springs, sea water)
Water Temperature (surface, at depth)
Salinity/Conductivity Water (surface, at depth)
Dissolved Oxygen Content Water (surface, at depth)
pH Water (surface, at depth)
Redox Potential Water (surface, at depth)
Water Depth (minimum, maximum)
Visibility (material type in suspension)
106
Appendix A
General Sediment Composition (in situ observation)
Mobility of Sediment Surface (rippling, direction and frequency)
Sediment Slope
Probe Depth to Wreck Material (extent of burial)
Depth to Stable Seabed (evident by black/anaerobic layer)
Sediment Gradation (changes in colour)
Sediment Photography (surface, gradation, at depth)
Sediment Sampling (sample all significant layers)
Sediment Analysis (particle size distribution, inorganic elements, organic content, nutrients,
micro-organisms)
pH Sediment (measure all significant layers)
Redox Potential Sediment (measure all significant layers)
Timber Infestation by Marine Borers (active, depth to non activity)
Probe Depths of Timbers (exposed, buried)
pH Profiles of Timbers (exposed, buried)
13
Timber Samples (wood identification, maximum water content, FT-IR, C-NMR, py-gc-ms)
Corrosion Potential Metals (concretion/metal interface)
Surface pH Metals (concretion/metal interface)
Depth of Concretion and Graphitisation
Depth of Concretion
Depth of Graphitisation
Sample Concretion (optional)
Sample Metals (optional)
107
Appendix A
APPENDIX B
ON-SITE CORROSION SURVEY DATA SHEET
Date of Survey: ______________________________________________________
Time of Survey: ______________________________________________________
Personnel: __________________________________________________________
Site (name, date and type): ______________________________________________
Location & GPS Co-ordinates: __________________________________________
Weather and Sea Conditions: ___________________________________________
Swell and Tidal Information: ____________________________________________
Current (rate, direction, speed): __________________________________________
Water Temperature: ___________________________________________________
Water Depth to Wreck (minimum, maximum): ______________________________
Visibility (metres): ____________________________________________________
Distance and Direction from Land/Reef: __________________________________
Freshwater Influence (e.g. rivers, springs, rain water run off): __________________
____________________________________________________________________
Site Dimensions (length, width, area): _____________________________________
Site Orientation (e.g. upright, list to port or starboard, upside down): _____________
Composition of Dominant Wreck Material (e.g. iron, aluminium): ______________
____________________________________________________________________
Dominant Encrusting Organisms on Surface (type, abundance & photograph):
____________________________________________________________________
____________________________________________________________________
Evidence of Active Corrosion (depth, position & photograph): Y/N
____________________________________________________________________
____________________________________________________________________
Evidence of Dynamite and/or Storm Damage (depth, position & photograph): Y/N
____________________________________________________________________
____________________________________________________________________
Evidence of Structural Collapse (depth, position & photograph): Y/N
____________________________________________________________________
____________________________________________________________________
Evidence of Human Disturbance (e.g. salvage, pollution) (depth, position &
photograph): Y/N
108
Appendix A
____________________________________________________________________
____________________________________________________________________
109