DIAMER BASHA DAM PROJECT

SOME ASPECTS OF SEISMIC HAZARDS

MARCH 2018

PREPARED BY

TALHA MUHAMMAD KAKAR

SYED KAZIM MEHDI

Consultant in

Seismology, Geophysics & Geology

Cell: 0300-5478842
 Table of Contents

Abstract

1. Introduction 1

2. Regional Tectonic and Seismicity 2

3. Seismic Data and Monitoring 4

4. Data Processing 5

5. Site Effect Analysis 7

6. Deterministic Seismic Hazard Calculation 9

7. Conclusions 11
Abstract: This report presents aspects of the seismic hazard in and around the proposed dam site
of Diamer-Basha, North Pakistan. We applied well-known approach of Nakamura for site-effect
investigation in the study area. The natural frequency of Diamer-Basha dam was calculated which
is equal to 1.35Hz. Recommended frequency of the dam (1.35 Hz) is approximately similar to the
range of strong amplification at the hard rock site on which Diamer- Basha dam is going to be
constructed. Maximum Credible Earthquake analyses was done using an appropriate empirical
equation which predict that earthquakes ranging from 6.5-7.5 Mw can occur in the investigated
area. Calculated peak ground acceleration (PGA) values range from 0.06g (minor to moderate
damage) to 0.59g (severe damage) at the dam site. It is also shown that higher PGA hazard is
adjacent to known seismogenic fault of Main Mantle Thrust (MMT).

Keywords: seismicity, site effect, Diamer-Basha dam, PGA, North Pakistan

1 Introduction

The importance of seismic aspects of new and existing dam sites is emphasized by ICOLD
bulletins (ICOLD 2016, 2010, 1989). There exist a number of efforts of seismic aspects of dams
all over the world, such as: Assad et al. (2013) for Iran, Seyrek and Tosun (2011) for the Turkey,
Rivas-Medina et al (2011) for the Spain, Augello and Yeats (2003) for the United States, Rastogi
et al. (1997) for the Indian region, and Mahdi (2005,1988) for Pakistan. The knowledge of
seismicity and site response characteristics parameters are of great importance in examining and
evaluating in seismic hazard assessment of new and existing dams’ sites. Studies of seismicity of
dams have been carried out by number of geoscientists (Assad et al. 2013; Mahdi et al. 2005;
Rastogi et al. 1997). Assad et al. (2013) described and compared the seismic earthquakes and
volume changes of the reservoir of the Latian dam, Tehran, Iran. Mahdi et al. (2005) identified
characteristics of reservoir induced seismicity at Tarbela and Manga dams, Pakistan. Rastogi et al.
(1997) analyzed seismicity around Dhamni dam, Maharashtra,

1
India and investigated the spatial-temporal correlation of seismicity with the Dhamni reservoir.
The generally used method of spectral ratio of horizontal to vertical component (H/V) suggested
by Nakamura is for site response investigations (Ahmed et al. 2009; Bard 2008; Nakamura 2000,
1996, 1989; Nogoshi and Igarashi 1971; Kanai and Tanka 1961). The H/V method is also called
the receiver transfer function technique or the non-reference site dependent technique. This
approach has been used in many applications and proved to be the low-cost technique to estimate
site effects (Ahmed et al. 2009; Bard 2008; Parolai et al. 2004).

Ground motion characterizations of dam site have been studied previously (e.g. Seyrek and Tosun
2011; Rivas-Medina et al 2011; Augello and Yeats 2003). Recently, Seyrek and Tosun (2011) and
Augello and Yeats (2003) used the deterministic seismic hazard assessment method for dam sites
in Kizilirmak basin, Turkey and northwest Oregon dam sites, USA respectively. Furthermore, they
considered and used the well-established empirical relationship of Wells and Coppersmith (1994)
for the calculation of earthquake magnitude. In contrast, Rivas-Medina et al. (2011) proposed the
probabilistic seismic hazard method using a logic tree in Northern Spain (the Itoiz Dam site).
However, no published seismic hazard work in the proposed Diamer-Basha dam site has been
reported so far. Thus, this research uses the methods that have not been used before for the analysis
of seismic hazard for the study area.

The main steps for investigating the seismic aspects in and around Diamer Basha dam site in the
current study are: (1) brief introduction about the regional geology and the seismotectonic features;
(2) instrumentation and utilization of available and existing seismicity data, as statistical methods
are generally based on the spatial-temporal distribution of earthquake data; (3) seismicity analysis
and data processing; (4) site effect analysis in and around Diamer Basha dam site; (6) Deterministic
seismic hazard assessment (DSHA) at the dam site in terms of magnitude and peak ground
acceleration (PGA).

2 Regional Tectonic and Seismicity

The regional geological and tectonic regime of the area around the proposed Diamer-Basha Dam
site (4500 Mega Watt) had been studied by many authors (e.g Chaudry and Ghazanfar 1990; Mona
Lisa et al. 2007). This site (73°.25’-74°.30’E and 35°25’-35°85’N) is situated along the

2
middle of Kohistan Island Arc (KIA) and sandwiched between the Eurasian and Indian Plates
which is bounded in the south by the Main Mantle Thrust (MMT) and in the north by the Main
Karakoram Thrust (MKT) as shown in Fig. 1. The Kohistan sequence represents a structurally
coherent section of a KIA terrain consisting of 30 to 40 km thick section of metamorphosed,
igneous and sedimentary rocks. The tectonic and the stratigraphic sequence of the area are
extremely complex (Jan and Howie 1981). The MMT is a 400 km long, north dipping suture zone
between the Indo-Pakistan sub-continental plate and the KIA. It extends from Khar (Bajaur
Agency) in the west to the north of Narran (Kaghan Valley) in the east. Then it takes a northeast
ward bend towards the east of Bunji and gets truncated by Raikot Fault (Kazim and Siddique
2006). The thrust inclines steeply near the surface; however, this inclination is believed to decrease
considerably with depth likewise as interpreted for other local thrust faults of the region. This
mega-thrust zone is marked by the presence of upper mantle and lower crustal rocks on Kohistan
side. On the Indian side it is marked by deep lower crustal rocks such as blue schist and eclogite.
The MKT also called the Shyok Suture Zone is a regional thrust separating the Asian mass from
KIA. The north dipping MKT is associated with significant seismic activity like 1948 earthquake
with magnitude of 6.8 Mw. It is considered that a rupture along the MKT could involve long
portions of the fault system, because it is comparatively straight over significant distances as
compared to the other structures present in the area (DBDC, 2007). This thrust is approximately at
a 94 km horizontal distance from the dam site.

From historic data it was found that majority of major events were triggered from Hindu-Kush
region which is the political boundary between Pakistan and Afghanistan as shown in Figs. 2a and
b. This zone is one of the most active zones of the world and had experienced several major
earthquakes. The earthquakes in 1983, 1985 and 1991 with moment magnitudes (Mw) 7.4, 7.4 and
6.7 have taken lives of more than 300 people in the regions of Peshawar, Chitral, Swat and
Malakand (www.pmd.gov.pk). In northern Pakistan there exists all types of geological structures
and range of earthquakes (shallow, intermediate and deep). Most of the events which have shallow
depth reported around Diamer-Basha dam site can prove to be very dangerous. Greater depth
earthquakes were reported in Hindu-Kush region which is far away from Diamer-Basha dam (Fig.
2b). The second most seismically active zone is Nanga Parbat Syntaxes and numbers of the
earthquakes are recorded on weekly basis. Numbers of Neo-tectonic features are present in

3
Nanga Parbat Syntaxes belt but Raikot-Sassi fault represents the most active feature. Majority of
events are shallow and intermediate depth (10-60) km and having Mw 3 to 6. Low to moderate size
of earthquakes are occurring continuously from long period of time which shows that it might
produce greater deformation in Nanga Parbat Syntax. Although the syntaxes are tectonically very
active due to the presence of several Neo-tectonic features, but sudden burst of accumulated strain
energy along the active Raikot-Sassi fault zone in particular can be hazardous (Kazim and
Siddique, 2006). Nanga Parbat Syntax is situated near to Diamer-Basha dam and Bunji hydro
power project as shown in Fig 2a.

A major earthquake struck the India-Pakistan border on the morning of 8 October 2005. It had a
magnitude of Mw=7.6 and was felt strongly in much of Pakistan, northern India and eastern
Afghanistan. The earthquake resulted in more than 80,000 deaths in northern Pakistan and
adjoining parts of Jammu & Kashmir (Burton et al. 2006). A total of 147 aftershocks were
registered in the first day after the main shocked. It was observed that majority of events were
shallow depthk2s and different neo-tectonic features were seen in Hazara-Kashmir area. As shown
in Figure 3.7 and 3.8 that cluster of events can be seen at in Hazara-Kashmir area and Nanga Parbat
Syntaxes. Most of the events are shallow depth which is most dangerous for man- made structures
and land sliding. Due to the presence of Neo-tectonic features, many earthquakes are recorded on
weekly basis which shows that Hazara-Kashmir Syntaxes belt is still active and it might generate
high magnitude earthquake like 2005 Kashmir earthquake.

3 Seismic Data and Monitoring

As Diamer-Basha dam is a mega structure and situated in seismically active region to know about
recent seismic activity around Diamer-Basha dam. According to the requirement of ICOLD, micro
seismic monitoring systems should be installed five years prior to the construction of dam. These
systems are very important for monitoring the seismicity around the dam. On the recommendation
of ICOLD, Water and Power Development Authority (WAPDA) has installed ten micro seismic
stations in 2007 around Diamer-Basha dam as shown in Table 1. With the help of these systems
seismicity is being recorded continuously on real time with help of Antelope software.

4
The seismicity around the Diamer-Basha dam was recorded by weak motion seismometers
(Lennart LE-3Dlit) as shown in Fig. 3a. To determine the site effect of Diamer-Basha dam Geo
SIG strong motion accelerometer was used as shown in Fig. 3b. The measuring equipment at all
stations consists of three component seismometers, 24-bit data logger quanterra Q330 with
internal GPS time receiver and PB14F packed Baler for data storage (Fig. 3c). The option data
logger preamplifier with a gain of 30 is activated which means that the default input range of Q330
of 40 Vpp is reduced to 1.33v as shown in Fig. 3c. The composite digital decimation filters with
a linear phase below 100 Hz were chosen in the data logger configuration. The sensitivity of the
LE-3Dlite Seismometer is 400 v/m/s with maximum output of ± 5V. Therefore, one count in the
recorded data corresponds to 0.2 nm/s of ground velocity sampled with 100 sps. With this
instrumentation and settings, it is expected that station will not be clipped up to local magnitude
(ML) 4.5 earthquake at the distance of 10 km and up to ML=5.2 earthquake at the distance of
about 30 km. The data logger self-noise is well below Peterson’s low-noise model (Peterson 1993)
as presented in Fig. 4. So only the self-noise of the sensor limits the recording capability. The
noise analysis will show that the detection capability of the sites is limited by self-noise of the
sensors.

The seismological data (observatories) obtained from three sources namely Water and Power
Development Authority (WAPDA), United States Geological Survey (USGS)/National
Earthquake Information Centre (NEIC) and International Seismological Center (ISC). WAPDA
data was used to find out recent seismicity and resonance frequency around Diamer-Basha dam,
while USGS and ISC data set were used for finding regional seismicity and deterministic seismic
hazard assessment (DSHA).

4 Data Processing

The data analysis was done as described in the New Manual of Standard Seismological
Observatory (Bormann, 2002). Antelope 4.11 software with Mac operating systems was used
for the data processing of the recorded events (BRTT, 2009). The processing can be

5
performed automatically or manually as shown in Fig. 5. In automatic processing, the
automatic detector was run on the whole set of data using several different streams. dB detect reads
waveform data from an input data base runs STA/LTA detecting the algorithm on the selected
channels of waveform data and write the detection time into on an output database. Only P-wave
velocities were used for automatic location. System automatically pick the travel time of P-waves
from vertical component of each known stations with help of IASPEI91 (Kennett and Engdahl,
1991) international velocity model as shown in Fig. 6. The automatic processed data has been
shown in Fig. 7a.A standard IASPEI91 (Kennett and Engdahl, 1991) velocity model was used for
the location of different events. The obtained origins were refined with the relocate antelope
module designed for locating complete catalogue of events. The individual events were relocated
many times for finding out the precise and most accurate results of the earthquake location. A
detail of different velocity models is outlined in Table 2.
Events that were detected and located with the data from at least six stations were selected for
manual processing as presented in Fig. 7b. Remaining four stations was skipped because of non-
availability of data due to power issues or heavy noise where it was difficult to mark P- and S-
wave arrival time. Individual event was reviewed and first break of P- and S-wave on vertical and
horizontal component of each channel was marked. P-wave arrival time were corrected where
necessary for each event and special emphasis is given on optimum picking of S-wave arrival time
which is very important in improving the epicenter location and depth of each event. However, to
assure the quality of the final location only distinct arrival were picked and their time uncertainty
assessed. To find the accurate location different velocity models were tested to have minimum
standard deviation during the processing of data. It has been observed that northern area velocity
model was found best velocity model for processing of earthquake data, which is outlined in Table
3. This research work is therefore restricted to the events inside and close to the WAPDA network,
or approximately inside the region between (35.25-35.85) N and (73.25-74.30) E or events which
had been recorded within one kilometer of Diamer-Basha dam’s radius. Figs. 8a and b show cross
section in north south and east west direction through the network region for depicting shallow
and major depth events. These cross sections also show that majority of the events are at shallow
depth, recorded inside the Chilas complex.

6
5 Site-effect Analysis

From a civil engineering point of view, site effects are one of the most important parameters and
it could be well defined by means of the site amplification and the predominant frequency of the
medium response (CGS, 2007).

5.1 Site effect of Diamer Basha- sites

Two digital, three components strong ground motion station (Kinemetric) have been in operation
since 04 February 2005 in the area near the Diamer-Basha dam site. The accelerometers are located
on hard rock as well as on alluvium (Table 4). The 8 October 2005 earthquake was recorded along
with several of its aftershocks. The waveforms recorded at hard rock and alluvium sites are
presented in Table 5. It also shows the recorded aftershocks for October 2005earthquake. We
selected October 2005 earthquake based on two reasons: (1) this is the only source of data which
was recorded at Diamer-Basha dam site and (2) no major event like October 2005 earthquake with
7.6 Mw has ever been recorded at this site.

The KMI strong motion analyst and Geo Dk2AS software’s were used for the processing of strong
motion data. The waveforms of the 8 October 2005 earthquake and its aftershocks are well
recorded by the hard rock site. However, the alluvium site had its start time wrongly set 01.01.1980
and hence it is difficult to identify the main event and its aftershocks. The main earthquake
recorded by the hard rock site has duration of 80 sec. There was no record similar to that duration
at the alluvium site. Therefore, it has to be assumed that the alluvium site was not triggered by the
earthquake of 8 October 2005. The method using the reference site could not be applied to the data
because it was impossible to identify an event recorded by the hard rock and the alluvium site
simultaneously. Figs. 9a−c show the acceleration of 8th October 2005 earthquake recorded at the
hard rock site. All three-component acceleration has been recorded with vertical component
acceleration (0.304 m/s2), less than it was suggested for Diamer-Basha dam (0.5m/s2). Remaining
aftershocks have negligible vertical component acceleration. Main objective of SMA installation
around the dam site is to continuously monitor the PGA value of any earthquake. For example, the
suggested and estimated PGA for Tarbela dam was 0.1m/s2, but after 1973 earthquake event, more
than that was recorded at SMA of Tarbela dam

7
during the construction. After this, Tarbela dam was redesigned for 0.6 m/s2 PGA and around 9
SMA has been installed around the dam.

5.2 Site-effect analysis from horizontal to vertical component ratios (H/V)

The Horizontal to vertical component ratios (H/V) method was used to calculate site response at
hard rock. The Fourier spectra were calculated in the time window from 10 to 22 sec. The observed
energy of the signal disperses over frequencies between 0.5 and 20 Hz (see Figs. 10a−c). Fig. 10a
shows high amplification from 0.5 to 20 Hz as compared to the remaining aftershocks, thus making
it possible to differentiate main shock from aftershocks. Figs. 11a−c display the spectral ratio of
the X/Z components (solid line) and the Y/Z (dashed line). Both ratios show strong amplification
of the signal around 2.6 Hz with the value of amplification of the signal varying from 5 to 6 Hz.
Amplification of the signal is also observed around 6 and 9–12 Hz. Aftershocks were processed in
the same fashion. From a theoretical point of view, a hard rock site with 2.6 Hz resonant frequency
is a feature of large earthquakes. The time window used to analyze data from the main earthquake
is 12 sec, which is much longer that the time window of the aftershocks (4–8 sec.). Spectral ratio
of the X/Z (solid line) and the Y/Z (dashed line) components show a strong amplification around
1.7 to 2.5. The value of the amplification is not well constrained by the data. It changes from one
earthquake to another and is strongly influenced by the selection of the time window used to
calculate the spectra.

5.3 Natural frequency of Diamer-Basha Dam

The natural frequency of the Diamer Basha dam site was obtained using the following equation
(Chopra, 1988).

T = 1.4 Hs / /¯(ES)¯¯ (1)

where T is time period of vibration of the dam in seconds, is height of the dam in feet, is Young's
modulus of elasticity of concrete in pounds per square inch. At the Diamer-Basha dam site Hs =
274 m (898.95 ft.), Es = 20 GPA = 2900653 lb/in2 (data provided by the Diamer-Basha

8
dam consultant BDC). These give T =0.739 sec or natural frequency = 1.35Hz approximately
similar to the range of strong amplification at the hard rock site on which Diamer-Basha dam is
going to be constructed. This suggests a dynamic response analysis for changing the frequency of
Diamer-Basha dam using length-height ratio (LH) and width-height ratio (WH), which are
inversely proportional to the natural frequency of dam (Tsai et al., 2009).

6 Deterministic Seismic Hazard Calculation

6.1 Maximum Credible Earthquake (MCE)/Maximum Earthquake Potential

In order to calculate deterministic seismic hazard, it is important to have an understanding of the

maximum credible earthquake (MCE) that will affect a site. According to ICOLD guidelines, the
MCE is the largest reasonable conceivable earthquake that is possible along a recognized fault or
within a geographically defined tectonic province under the presently known or presumed tectonic
framework. The MCE for each potential earthquake source was calculated using the equations of
Wells and Coppersmith (1994); and Bonilla et al. (1984) and Nowroozi (1985). These equations
are as follows:

Mw = 5.02 + 1.22 Log L (2)

Ms = 6.04 + 0.708 Log L (3)

Ms = 1.259 + 1.244 Log L (4)

L is rupture length in Km in equations (2) and (3), which are Wells and Coppersmith (1994) and
Bonilla et al. (1984) empirical equations respectively and in meters in equation (4), which is
Nowroozi (1985) equation. Application of Ms dependent relations (3 and 4) needs conversion of
the earthquakes to Mw and Ms to Mw are performed using Scurrilous (2006). Table 6 summarizes
the results obtained from above mentioned equations.

9
6.2 Diamer-Basha dam deterministic seismic hazard assessment

Due to insufficient strong motion records for Pakistan identified by Ambraseys and Bommer
(1991), Boore et al. (1997) attenuation equation is adopted to estimate PGA at Diamer-Basha dam
site. The attenuation equation of Boore et al. (1997) has been used previously to assess seismic
hazard of the North-West Himalayan Fold and Thrust Belt, Pakistan (Mona Lisa et al. 2007). Boore
et al., (1997) attenuation equation was defined for shallow earthquakes in distance range of rupture
distance Rrup ≤ 80 km and magnitude range of Mw (4.0-8.5) in western North America.

ln (Y) = b1 + b2 (M w − 6) + b3 (M w − 6)2 + bs ln r + bv lnVS,30 IVA (5)

1
Y = Peak ground acceleration in g, r = (r 2jb + h 2 ) 2
rjb= shortest distance in km from the station

to the surface projection of the fault rupture, h = depth and VS, 30=shear wave velocity (in the upper
30 m), b1 = b1RS for reverse slip earthquakes and b1ALL = if mechanism is not specified.

Boore et al. (1997) gives best result for reverse faulting, which can be major source of earthquake
in study area. This equation also encompasses earthquakes having depth greater than 30 km
covering both shallow and intermediate earthquakes. Three respective faults/zones have been
considered as potential sources for future earthquakes (Fig. 1): (1) MMT (Zone-1), (2) MKT
(Zone-2) and (3) Chilas complex fault (Zone-3). The maximum possible earthquakes established
for each zone were assumed to occur on each feature in the respective zones. A depth of 20 km
was assumed for these earthquakes. Table 5.2 shows the results of PGA calculations for the
different faults which have been calculated with help of Boore et al., (1997) attenuation
relationships. The maximum value of PGA (0.59) was found for Diamer-Basha dam which might
be produced due to MMT while considering an epicentral distance of 1 km and 8 Mw. Due to the
seismological importance of the MMT, the uncertainty of its dip and depth extent, different
scenarios were modeled to form an understanding of its effect on the expected PGA values. The
PGA values were calculated for closest as well as incremental epicentral distances from1 to 20 km
(Table 7). The MMT yielded the highest average respective PGA values of 0.2-0.59g obtained as
shown in Fig. 12. In Zone-2, only one fault, MKT was considered in the computation
10
as shown in Table 7. Based on the magnitude of 7.8 and depth at 20km, PGA of 0.059 g is
calculated as shown in Fig. 12. The highest PGA value of 0.11g can be obtained from Chilas
Complex (Zone-3) as shown in Figure 5.1 located 40km from the Diamer-Basha dam.

7 CONCLUSION

Results reveal that most seismically active zone is MMT, where numbers of events are recorded
on weekly basis. Majority of events are shallow and intermediate depth (10-40) km and having
magnitude 3-5ML (Local magnitude). The frequency dependent response of Diamer-Basha dam
based on spectral ratios for hard rock site exhibits frequency dependent amplifications with peaks
at 2.6, 5-6 and 9-12 Hz. The most relevant amplification for the Diamer Basha dam site was
observed using the waveform data of the 8th October 2005 earthquake and its aftershocks at
2.6 Hz closer to the natural (resonance) frequency of the dam (1.35 Hz) which is almost equal to
the natural of hard rocks.

Maximum Credible Earthquake (MCE) analyses based on three different equations predict that
earthquakes ranging from 6.5-7.5 Mw can occur in the investigated area. The deterministic seismic
hazard assessment approach around Diamer-Basha dam predicts three major potential zones i.e.
MKT, MMT and Chilas complex. Out of these three major potential zones, MMT yielded the
highest Peak Ground Acceleration (PGA) value of 0.59 obtained by Boore et al., (1997)
attenuation relationship with the assumption of rupture taking place immediately underneath the
dam site (hypocentral distance of 20 km).

11
REFERENCES

Ahmed S, Shah MD, Qaisar M, Khan K (2009) Estimation of site response characteristics at four
sites of seismic monitoring network in Pakistan. Nucleus, 46(1-2): 9–14

Ambraseys NN, Bommer JJ (1991) Database of European strong ground-motion records. Eur
Earthq Eng. 5: 18–37

Asadollahfardi G, Aghda MF, Rahbar M (2013) Case Study of Reservoir Triggered Seismicity,
Latian dam, Tehran, Iran. 82:553–560

Bonilla MG, Mark RK, Lienkaemper JJ (1984) Statistical relations among earthquake magnitude,
surface rupture length, and surface fault displacement. Bull Seismol Soc Am. 74(6): 2379–2411

Boore DM, Joyner WB, Fumal TE (1997) Equations for estimating horizontal response spectra
and peak acceleration from western North American earthquakes: a summary of recent work.
Seismol Res Lett. 68(1): 128–153

Bormann P (2002) New manual of seismological observatory practice (NMSOP-2). IASPEI, GFZ
German Research Centre for Geosciences, Potsdam
BRTT (2009) Antelope 4.11, Boulder Real Time Technologies, Inc

Burton PW, Hainzl S, Lasocki S (2006) Special Issue. Tectonophysics 403: 1–2

CGS (2007) Seismic hazard evolution for Diamer-Basha Dam project. Council for Geosciences

Chaudhry MN, Ghazanfar M (1990) Position of the Main Central Thrust in the tectonic framework
of Western Himalaya. Tectonophysics 174(3): 321–329

Chopra AK (1988) Earthquake response analysis of concrete dams. In Advanced dam engineering
for design, construction, and rehabilitation. Springer US

DBDC (2007) Dynamic response of Diamer-Basha Dam. Annual report of Diamer-Basha Dam
Consultant

Hull AG, Augello A, Yeats RS (2003) Deterministic seismic hazard analysis in northwest
Oregon, U.S.A. Proc. Pacific Conference on Earthquake Engineering. Paper No. 157.
12
ICOLD (1989) Selecting seismic parameters for large dams. Bulletin 72. Committee on Seismic
Aspects of Dams Design, ICOLD, Paris

ICOLD (2010) Selecting seismic parameters for large dams. Revision of Bulletin 72, Committee
on Seismic Aspects of Dams Design, ICOLD, Paris

Jan MQ, Howie RA (1981) The mineralogy and geochemistry of the metamorphosed basic and
ultrabasic rocks of the Jijal complex, Kohistan, NW Pakistan. J Petrol. 22(1): 85–126

Kennett BN, Engdahl ER (1991) Travel times for global earthquake location and phase
identification. Geophys J Int. 105(2): 429–465

Mahdi SK (1988) Tarbela Reservoir A Question of Induced Seismicity. Proc. 2nd International
Conference on Case histories in Geotechnical Engineering, St, Louis, Mo., Paper No. 3.49

Mahdi SK, Chaudhry MM, Siddique M (2005) Characteristics of reservoir induced seismicity at
Tarbela and Mangla dams. Proc. 73rd Annual Meeting of ICOLD, Tehran, Iran. Paper No.
018-W1

Mahdi SK, Siddique M (2006) The Kashmir Hazara earthquake and its implications on Jhelum
thrust zone, International Conference on Earthquake Engineering, UET Lahore, Pakistan
MonaLisa, Khwaja AA, Jan Q (2007) Seismic Hazard Assessment of the NW Himalayan Fold-
and-Thrust Belt, Pakistan, Using Probabilistic Approach. J Earthq Eng. 11: 257–301

Nogoshi M, Igarashi T (1971) On the amplitude characteristics of microtremors (part 2). J

Seismol Soc Japan. 24: 26–40 (in Japanese with English abstract)

Nakamura Y (1989) A Method for Dynamic Characteristics Estimation of Subsurface using

Microtremors on the Ground Surface. Quarterly Report Railway Tech Res Inst. 30(1): 25–33

Nakamura Y (1996) Real-time information systems for hazards mitigation. Proc. 11th WCEE.,
Acapulco, Mexico

Nakamura Y (2000) Clear identification of fundamental idea of Nakamura’s technique and its
applications. Proc. 12th WCEE., Auckland, New Zealand

Nowroozi AA (1985) Empirical relations between magnitudes and fault parameters for
earthquakes in Iran. Bull Seismol Soc Am. 75(5): 1327–1338

13
Parolai S, Bindi D, Baumbach M, Grosser H, Milkereit C, Karakisa S, Zünbül S (2004)
Comparison of different site response estimation techniques using aftershocks of the 1999 Izmit
earthquake. Bull Seismol Soc Am. 94(3): 1096–1108

Rastogi BK, Mandal P, Kumar N (1997) Seismicity around Dhamni Dam, Maharashtra , India.
Pure Appl Geophys. 150:493–509.

Rivas-Medina A, Santoyo MA, Luzón F, Benito B, Gaspar-Escribano JM, Garcia-Jerez A (2011)

Seismic Hazard and Ground Motion Characterization at the Itoiz Dam (Northern Spain). Pure
Appl Geophys. doi: 10.1007/s00024-011-0405-0

Scordilis EM (2006) Empirical global relations converting Ms and Mb to moment magnitude. J

Seismol. 10:225–236.

Seyrek E, Tosun H (2011) Deterministic approach to the seismic hazard of dam sites in Kızılırmak
basin, Turkey. Nat Hazards. 59: 787–800. doi: 10.1007/s11069-011-9795-7

Tsai PH, Hsu SC, Lai J (2009) Effects of Core on Dynamic Responses of Earth Dam. In Slope
Stability, Retaining Walls, and Foundations. Selected Papers from the 2009 GeoHunan
International Conference (pp. 8-13). ASCE

Wells DL, Coppersmith KJ (1994) New empirical relationships among magnitude, rupture length,
rupture width, rupture area, and surface displacement. Bull Seismol Soc Am. 84, 974– 1002

Figure Captions:

Figure 1 Location of proposed Diamer-Basha Dam with respect to the major tectonic features of
the region (Kazim and Rana, 1982; Mehdi, 2012).
Figure 2a: Location of seismic events based on different depths within 320 km of the Dam site
(modified from Quittmeyer and Jacob, 1979) (CGS, 2007).

Figure 2b: Historical seismic events from25AD to 1902(QuittmeyerandJacob,1979) within 320

km of the Dam site (CGS, 2007).
Figure 3: (a) Seismometer Lennartz LE-3Dlite and (b) Episensor (www.geosig.com).
(c) Details seismicstation map of WAPDA.
Figure 4: Self noise of the seismometer, Lennartz LE-3Dlite (www.geosig.com).
Figure 5: Automatic and manual processingof an event recorded on a three-component station
(WAPDA, 2004)
Figure 6: IASP91 velocity model which was used for automatic processing
(www.iris.edu/dms/products/emc).

14
Figure 7a: Epicenters determined from (a) automatic picks on at least six seismic stations
network (small squares). The IASPEI velocity model was used for automatic analysis and
(b) manual revised picks for the earthquake within the seismic stations network area
Figure 8a: Cross-section in (a) North-South direction through the seismic stations network region
and (b) East-West direction through the seismic stations network region.

Figure 9: Three component accelerograms of the 8 October 2005 earthquake (a) Location 2a;
(b) Location 2b aftershock; (c) Location 2c aftershock recorded at the hard rock site. X and Y
are horizontal components and Z is the vertical component.

Figure 10: The three component average Fourier spectra of (a) 8 October 2005 earthquake. The
seismic signal is strong at the lower end of the spectrum 0.4 Hz up to 10-15 Hz. Energy of the
seismic signal starts to rapidly decay above 15 Hz; (b) the Location 2b aftershock. The seismic
signal is strong in the frequency range starting from 1Hz to 10-15 Hz. Its energy gradually
decays below 1 Hz and starts to rapidly decay above 15Hz and (c) the Location 2c aftershock.
The seismic signal is strong in the frequency range starting from 4Hz to 15 Hz. Energy of the
seismic signal gradually decays below 4 Hz and starts to rapidly decay above 15Hz. solid-X,
dash-Y and dotted represent Z component

Figure 11a: The spectral ratio of the X/Z components (solid line) and the Y/Z (dashed line). Both
ratios show strong amplification of the signal around (a) 2.6 Hz and the amplification of the signal
is observed as well at around 6 Hz and 9-12 Hz; (b) 2.5-3 Hz and the amplification of the signal
is observed as well at around 5 Hz and 9-12 Hz and (c) 2.5-3 Hz and the amplification of the signal
is observed as well at around 6 Hz and 9-12 Hz.

Figure 512: Shows the PGA of three zones (MMT1-MMT4) with different epicentral distance,
MKT-5 and Chilas-6 complex.

15
Fig 1

16
Fig 2

17
Fig 3

18
Fig 4

19
Fig 5

20
Fig 6

21
Fig 7

Fig 8

22
Fig 9

23
Fig 10 a

Fig 10b

24
Fig 10c

Fig 11a

25
Fig 11b

Fig 11c

26
Fig 12

27
Table 1

Latitude Longitude
Station # (deg) (deg) Elevation(M)
DBD1 35.7255 73.6462 2127.9
DBD2 35.6634 73.9455 1828.8
DBD3 35.5743 73.8529 1232.4
DBD4 35.5142 73.7590 1209

Basha Dam
DBD5 35.5859 73.7097 1857.6
DBD6 35.5001 73.3665 1000.9
DBD7 35.4273 73.6205 2048.9
DBD8 35.4800 73.6930 2004.1
DBD9 35.3820 73.8564 2092.2
DBD10 35.4369 74.1514 1250.5

Table 2

Model RMS Max Epi Median Epi Max Depth Median Depth
(Km) (Km)
Beloussov (1980) 0.179 19.3 0.9 10.0 1.3
IASP91(1991) 0.245 20.4 2.3 43.4 2.3
Khurshid (1984) 0.211 15.9 1.3 31.6 1.8
Tarbela 0.443 16.8 2.6 25.6 4.2
CRUST2.0 (1998) 0.193 12.3 1.2 31.6 1.2
Diamer-Basha (2007) 0.149 0.0 0.0 0.0 0.0

28
Table 3
Depth[km] Vp[km/s] Vs[km/s]
-3.0 6.2 3.6
0.3 6.4 3.7
3.0 6.5 3.75
67.0 8.05 4.62

Table 4

Name Description of the site Installed on

Location 1 Alluvium at right abutment 04-02-2005
Location 2 Hard rocks at left abutment 04-02-2005

Table 5
Fault Epicentral Magnitude PGA
distance
MMT1 1 8 0.592
MMT2 5 8 0.474
MMT3 10 8 0.339
MMT4 20 8 0.213
MKT 94 7.8 0.059
Chilas 40 7.4 0.114

Table 6
Maximum Credible Earthquake
Expected
Seismic Feature Type Rupture Length Wells & Nowroozi Bonilla
Coppersmith 1985 1984
Km 1994

MMT Reverse 113.33 7.2 Mw 6.9 Mw 7 Mw

29
Table 7

Channel [m/s2 ]
Signal
X Y Z
Start Time Duration
[s]
Alluvium Site
Location 1 1/1/1980 0:39:01 0.11 0.1 0.04 17
Hard rock Site
Location 2a 10/8/2005 8:51:02 0.66 0.8 0.3 80
Location 2b 10/8/2005 9:02:57 0.07 0.11 0.052 18
Location 2c 10/8/2005 9:26:40 0.06 0.08 0.045 15

30