Nat. Hazards Earth Syst. Sci.

, 12, 2425–2437, 2012

www.nat-hazards-earth-syst-sci.net/12/2425/2012/ Natural Hazards
doi:10.5194/nhess-12-2425-2012 and Earth
© Author(s) 2012. CC Attribution 3.0 License. System Sciences

A method to characterize the different extreme waves for islands

exposed to various wave regimes: a case study devoted to Reunion
Island
S. Lecacheux1 , R. Pedreros1 , G. Le Cozannet1 , J. Thiébot2 , Y. De La Torre1 , and T. Bulteau1
1 BRGM, 3 Avenue Claude Guillemin Cedex 2, BP 6009, 45060 Orléans, France
2 LUSAC, Université de Caen Basse-Normandie, site universitaire, 50130 Cherbourg, France

Correspondence to: S. Lecacheux (s.lecacheux@brgm.fr)

Received: 26 July 2011 – Revised: 29 March 2012 – Accepted: 16 April 2012 – Published: 31 July 2012

Abstract. This paper outlines a new approach devoted to the western part of the island and induce similar impacts in
the analysis of extreme waves in presence of several wave terms of runup and submersion. This method can be easily
regimes. It entails discriminating the different wave regimes transposed to other case studies and can be adapted, depend-
from offshore wave data using classification algorithms, be- ing on the data availability.
fore conducting the extreme wave analysis for each regime
separately. The concept is applied to the pilot site of Reunion
Island which is affected by three main wave regimes: south- 1 Introduction
ern waves, trade-wind waves and cyclonic waves. Several ex-
treme wave scenarios are determined for each regime, based The determination of coastal hazards maps at a regional scale
on real historical cases (for cyclonic waves) and extreme is of prime importance in implementing efficient coastal risk
value analysis (for non-cyclonic waves). For each scenario, reduction programs and in achieving proper management
the nearshore wave characteristics are modelled all around of urban development. This is especially crucial for islands
Reunion Island and the linear theory equations are used to where the living space is limited and where populations are
back calculate the equivalent deep-water wave characteris- often concentrated near the coastline.
tics for each portion of the coast. The relative exposure of Waves play a crucial role in coastal erosion and marine in-
the coastline to the extreme waves of each regime is deter- undation hazards. In shallow waters, the breaking of waves
mined by comparing the equivalent deep-water wave charac- causes a local elevation of the mean water level, called wave
teristics. setup, and the maximum vertical extent of wave uprush on
This method provides a practical framework to perform an a beach or structure above the mean water level is termed
analysis of extremes within a complex environment present- runup. A reliable assessment of extreme runup values is nec-
ing several sources of extreme waves. First, at a particular essary to map coastal hazards, especially for inundation and
coastal location, it allows for inter-comparison between var- overtopping (Ruggiero et al., 2001). At regional scales, the
ious kinds of extreme waves that are generated by different runup is usually calculated with empirical formulas, such as
processes and that may occur at different periods of the year. Stockdon et al. (2006), using deep-water wave input and as-
Then, it enables us to analyse the alongshore variability in suming a shore-normal orientation. In practice, the runup am-
wave exposition, which is a good indicator of potential runup plitude also depends on the wave incidence angle and the
extreme values. For the case of Reunion Island, cyclonic nearshore wave transformation due to the local bathymetry
waves are dominant offshore around the island, with equiv- (by refraction, reflection, diffraction and bottom friction).
alent deep-water wave heights up to 18 m for the northern Thus, even at regional scale, the determination of extreme
part. Nevertheless, due to nearshore wave refraction, south- wave characteristics must be site-specific.
ern waves may become as energetic as cyclonic waves on

Published by Copernicus Publications on behalf of the European Geosciences Union.

2426 S. Lecacheux et al.: A method to characterize the different extreme waves for islands

In coastal engineering, two classic approaches may be

used to determine local extreme wave characteristics. The
first approach entails performing a statistical analysis directly
at a nearshore point in front of the area of interest, as in Lee
and Jun (2006). Waves are then usually considered as fully
refracted and normal to the shore. The feasibility of this ap-
proach depends on the existence of long-term coastal wave
measurements or numerical simulations, which are rarely
available at the regional scale. In the second approach, off-
shore design wave heights are first determined for different
types of waves, corresponding to angular sectors (for exam-
ple, every 30◦ ) or seasons (winter waves, etc.). Then, they
are propagated using a nearshore wave transformation model
to calculate the local shore-normal wave characteristics re-
lated to each type (Vitousek and Fletcher, 2008; Vitousek et
al., 2008; Panigrahi and Misra, 2010). In any case, a pre-
requisite to the calculation of runup with empirical formu- Fig. 1. The pilot site and positions of the wave gauges and the
las from nearshore wave characteristics is the determination NWW3 points used in the study. The arrows indicate the sources
of “equivalent deep-water wave heights” noted H0 eq in the of the dominant wave regimes.
following. Traditionally, H0 eq is a hypothetical wave height
devised to adjust the deep-water wave height (H0 ) that may
have undergone nearshore transformations such as wave re- each regime. As the magnitude of the wave runup depends
fraction (Goda, 2000). Here, it is defined, like in Stockdon on the amount of energy carried out by the waves, the com-
et al. (2006), as the nearshore wave height Hs (refracted and parison is based on both wave height and energy.
dissipated over local bathymetry in the model) that has been This paper is organized as follows. Section 2 introduces
reverse shoaled to back calculate the corresponding deep- the pilot site as well as the data. In Sect. 3, the method is fully
water wave height. Thus, the “equivalent deep-water waves” described. Section 4 presents the results, Sect. 5 is devoted to
facilitates the inter-comparison of extreme waves between the discussion and Sect. 6 to the conclusion.
different sites and can be used in the runup parameterizations
taking into account offshore wave characteristics.
In this study, we intend to compare the relative contri- 2 Pilot site and data
bution of several types of wave regime (trade-wind waves,
cyclonic waves and southern waves) to the local wave ex- 2.1 Reunion Island and the main wave regimes
treme values around tropical islands. Thus, we propose an
adjustment of the second approach, which entails calculating Reunion Island is a French Overseas Department lying east
deep-water wave statistics before modelling the correspond- of Madagascar. It is a volcanic island with a volcanic body
ing coastal wave characteristics. The method is applied for in the northeast (the massif of the Piton des Neiges), which
Reunion Island. has been inactive for 12 000 yr, and an active portion in the
First, a new approach devoted to the analysis of extreme southeast (the Piton de la Fournaise). Due to the mountain-
waves in the presence of several wave regimes is presented. ous nature of the island’s interior, 80 % of the population is
To satisfy data homogeneity requirements, each wave regime concentrated near the shore, thereby resulting in intense an-
must be studied separately. Although a directional criterion is thropic pressures. Coastal defence against erosion and ma-
often adopted in the literature (Mathiesen et al., 1994; Mor- rine submersion is a major issue in this area. The island
ton et al., 1997; Menendez et al., 2009; Mackay et al., 2010), is exposed to three dominant wave regimes, as illustrated
using a single criterion is not always sufficient to distinguish in Fig. 1: trade-wind waves, southern waves and cyclonic
homogenous wave groups. Our approach entails (1) identi- waves. Cazes-Duvat and Paskoff (2004) describe these wave
fying homogeneous groups of waves using a classification regimes as follows.
algorithm, (2) performing an extreme wave analysis for each Trade-wind waves arrive mainly from the east and south-
group and (3) building several wave-event scenarios from the east, and their heights barely exceed 2 m, with periods rang-
wave statistics for each regime. ing between 5 and 10 s. They are generated by trade winds in
Then, for each scenario, local wave characteristics are the inter-tropical zone and are especially energetic during the
computed all around the island by means of a nearshore wave Southern Hemisphere winter, with exceptional wave heights
transformation model. The exposure of each of the island’s around 5 m. Southern waves come from the southwest, with
coastlines to the different extreme events is determined by a mean height of about 3–4 m and periods between 10 and
comparing the “equivalent deep-water waves” calculated for 20 s. They are generated at great distances from Reunion

Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012 www.nat-hazards-earth-syst-sci.net/12/2425/2012/

S. Lecacheux et al.: A method to characterize the different extreme waves for islands 2427

Island by storms in the Southern Hemisphere’s temperate

zone. Cyclonic waves are the most energetic events and oc-
cur only a few days a year between November and March.
Cyclones usually come from the northeast and then continue
their course either north of the island or, less frequently, to
the south. Cyclonic waves mainly impact the island’s north-
ern and eastern parts.

2.2 Data

Two different types of wave data were used. For non-cyclonic

wave analysis (trade-wind waves and southern waves), long-
term time series of wave parameters were extracted from the
NOAA Wavewatch3 re-analysis (hereinafter called NWW3),
covering a 13-yr period between 1997 and 2009 (Tolman,
2002). The data have a spatial resolution of 1.25◦ in longitude
and 1◦ in latitude, and a temporal resolution of three hours.
This choice was motivated by the quality of the NWW3
dataset in terms of extremes. Indeed, other re-analyses, such
as ERA-40, provide longer time series, but the extremes are
not as well reproduced (Caires et al., 2004). Two points Fig. 2. Diagram of the overall method.
near Reunion Island were used (see Fig. 1): “Point South”
(22◦ S–55◦ E) for the southern wave analysis and “Point
East” (21◦ S–56◦ 150 E) for the trade-wind wave analysis. For
height (Hs ). The scenarios were based on historical cases that
the cyclonic wave analysis, information about historical cy-
had been modified to fit the 50-yr Hs . For cyclonic waves
clones was drawn from the best track dataset from 1932 to
(Sect. 3.2), global re-analysis usually under-estimates wave
2008 provided by Méteo-France (the French Meteorological
heights because their spatial and temporal resolutions are too
Service).
coarse. Thus, long-term cyclonic wave data are not available
For the numerical modelling of waves, the bathymet-
to implement a probabilistic approach. We therefore chose to
ric data were taken from the global international database
use a deterministic approach where the scenarios correspond
SRTM30 PLUS DEM (Becker et al., 2009), gridded at about
to actual historical cases.
1 km in latitude and longitude, and the bathymetric sounds
Secondly, for each scenario, the nearshore wave fields
from the SHOM (French Naval Hydrographic and Oceano-
were simulated with the SWAN model (Booij et al.,
graphic Service) data, with a maximum resolution of 25 m
1999) until 100 m resolution grids surrounding the island
in coastal areas. Four coastal wave gauges, shown in Fig. 1,
(Sect. 3.3).
were used for the validation of wave modelling: “Port Est” at
Finally, Reunion Island was divided into coastal segments
a depth of 27 m (period of measurement: November 1996 to
of uniform orientation. For each segment and each scenario,
present), “Pointe du Gouffre” at a depth of 31 m (period of
the wave characteristics were extracted at a nearshore loca-
measurement: May 2000 to present), “Port Ouest” at a depth
tion and the “equivalent deep-water wave heights” were com-
of 25 m (period of measurement: February 1997 to today)
puted. The inter-comparison between the “equivalent deep-
and “Saint Pierre” at a depth of 33 m (period of measure-
water wave characteristics” led to the determination of the
ment: September 2000 to November 2010).
relative exposure of each segment to the extreme waves from
each regime (Sect. 3.4).
3 The overall method
3.1 Determination of extreme wave scenarios for
The method entails three main steps, summarized in the flow non-cyclonic waves
chart in Fig. 2.
Firstly, several extreme wave event scenarios were deter- 3.1.1 Identification of the wave regimes
mined for each wave regime separately. For southern and
trade-wind waves (Sect. 3.1), a probabilistic approach was As explained in Sect. 1, an assessment of extreme values
chosen based on an analysis of NWW3 data for two points must be based on a homogeneous set of independent events.
near Reunion Island. The wave regimes were first identified Classification algorithms (Butel et al., 2002) can be used to
with classification algorithms, and an extreme value anal- separate the different wave regimes from the NWW3 time
ysis led to the determination of the 50-yr significant wave series into a limited number of sea states.

www.nat-hazards-earth-syst-sci.net/12/2425/2012/ Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012

2428 S. Lecacheux et al.: A method to characterize the different extreme waves for islands

an expectation-maximization algorithm. After this process,

trade-wind waves were classified in a hidden state charac-
terized by eastern waves with low variance of the wave pa-
rameters. This method enabled the identification and exclu-
sion of most of the cyclones from the NWW3 time series
at “Point East”, as illustrated in Fig. 3. The few remaining
classification errors were corrected manually. Results of the
GHMM classification at “Point East” are presented in Ta-
ble 1. They have been checked visually to ensure that cy-
clonic events have been effectively separated from the trade-
wind waves in the resulting modes. This second algorithm
avoids the labour-intensive task of manually removing the
cyclonic waves from the NWW3 time series. However, in
practice, it is very sensitive to the initial conditions input to
the algorithm and is thus not straightforward to implement.
Therefore, it should be used only when the k-means cluster-
ing algorithm is unable to discriminate waves into indepen-
Fig. 3. Example of the automatic identification of cyclone Dina at dent modes.
“Point East” using the GHMM algorithm: The storm’s effects are
recognized due to the high temporal variability of the waves’ sig-
3.1.2 Statistical analysis of wave heights
nificant heights and directions.

The classification process described in Sect. 3.1.1 resulted in

At “Point South”, southern waves were iso- homogeneous groups of wave data: We chose one group for
lated from the NWW3 time series using a k-means southern waves at “Point South” and one group for trade-
clustering
 algorithm that classifies the wave’s parameters wind waves at “Point East”. For each group, an extreme
Hs (t) · cos(Dp (t)), Hs (t) · sin(Dp (t)), Tp (t) , following analysis was performed on Hs using the WAFO toolbox
the same approach as Butel et al. (2002) and Le Cozannet (Brodtkorb et al., 2000). It consists firstly in identifying in-
et al. (2011). Here, the distance between two triplets is dependent extreme wave data using the Peak Over Thresh-
calculated as follows: old (POT) method. Next, GPD (General Pareto Distribution)
*H H + laws were fitted to the data. The three parameters of each
si sj
law were fitted using the maximum likelihood, the method
d Tpi Tpj (1) of moments and the probability weighted moment method.
Dpi Dpj Lastly, the validity of the fit was evaluated with Kolmogorov-
v
u Hsi · cos(Dp ) − Hsj · cos(Dp ) 2
u
i j
!
Hsi · sin(Dpi ) − Hsj · sin(Dpj )
!2
Tpi − Tpj
!2
Smirnov and khi-2 tests. For our purposes, the best fit was
=t + +
σHs cos(Dp ) σHs sin(Dp ) σTp obtained using a GPD with coefficients calculated using the
maximum likelihood. Results are presented in Fig. 4. Con-
with Hs , the significant wave height, Tp , the peak period
sidering the length of the time series, the maximum return
and Dp , the peak direction. The identification of the wave
period was set to 50 yr.
regimes among the different classes is based on the barycen-
tre value and a seasonal criterion.
Isolating trade-wind wave regimes for “Point East” was 3.1.3 Scenario building
more difficult because these waves may have the same di-
rection, period and height as cyclonic waves in the NWW3 The statistical analysis provided some information on Hs ex-
model. This is partly due to the limitations of NWW3 in mod- treme values only, but the wave models need to be forced
elling cyclonic waves. Thus, the classification algorithm is with wave triplets (Hs , Tp , Dp ) to propagate the wave field
not able to distinguish the trade-wind waves effectively from nearshore. Different methods can be called on to assign a pe-
the cyclonic waves. Fortunately, while the trade-wind waves riod and a direction to the extreme values of Hs . A first pos-
remain relatively stationary, cyclonic waves, and particularly sibility entails performing a regression on bivariate diagrams
their direction, vary quickly. Thus, we used a semi-automatic to calculate simple laws linking Hs with Tp , on the one hand,
classification process in which the temporal variability of and Hs with Dp , on the other. Here, we chose an alternative
the wave’s parameters
 is taken into account. In practice, the method that entails using historical events and modifying the
wave signal Hs (t) · cos(Dp (t)), Hs (t) · sin(Dp (t)), Tp (t) Hs values to make them fit the 50-yr Hs without changing
is modelled as a Gaussian hidden state Markov model the period and direction. For the southern waves, three his-
chain (GHMM) using the METIS toolbox (Monbet and torical cases were selected to cover the range of possibilities
Ailliot, 2005). This statistical model is adjusted using for the direction of origin; for the trade-wind waves, only one

Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012 www.nat-hazards-earth-syst-sci.net/12/2425/2012/

S. Lecacheux et al.: A method to characterize the different extreme waves for islands 2429

Table 1. GHMM classification of the NWW3 time series at “Point East” (21◦ S 56◦ 150 E). The result obtained for the cyclonic class is a mean
of NWW3 model outputs that can be associated to cyclonic waves in the time series. Because of its coarse temporal and spatial resolutions,
NWW3 underestimates the cyclonic waves, so that the associated values are not accurate.

Hs standard Tp standard Dp standard Relative

deviation (m) deviation (s) deviation (s) occurrence of
Barycenters within the class within the class within the class the class (%)
Hs (m) Tp (s) Dp ( ◦ )
Cyclones and tropical storms 2.2 7.4 112 1 1 68 8.4
Trade-wind waves 2.4 9.3 115 0.6 1.6 32.3 43.5
Southern waves 2.3 12.2 199 0.7 1.9 25.3 48.1

Fig. 4. Return values of Hs for southern waves at “Point South” (left) and trade-wind waves at “Point East” (right) obtained with the POT
method and GPD laws. Results are presented up to a return period of 50 yr.

case was selected because the range of directions was more

limited:

– Case 1 (southern waves): based on the 1 August 2003

event with a southern origin;

– Case 2 (southern waves): based on the 2 August 2003

event coming from the southeast;

– Case 3 (southern waves): based on the 13 May 2007

event coming from the southwest;

– Case 4 (trade-wind waves): based on the 12 July 2006

event, which was the strongest trade-wind wave event
with a southeastern origin.

For each case, the wave parameters (Hs , Tp , Dp ) were ex-

tracted on the eight NWW3 boundary points surrounding the
Fig. 5. Tracks of the four cyclones considered in this study: Gamede island (see Fig. 1). Then, the Hs parameter was multiplied by
(purple crosses), Dina (red diamonds), Colina (green squares) and the same coefficient at each of the eight points. For the south-
Hollanda (blue dots). ern wave cases (Cases 1 to 3), this coefficient was calculated
in such a way as to ensure that the new Hs was equal to the
50-yr value at “Point South” (where the statistical analysis
was performed). For the trade-wind wave case (Case 4), it

www.nat-hazards-earth-syst-sci.net/12/2425/2012/ Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012

2430 S. Lecacheux et al.: A method to characterize the different extreme waves for islands

Fig. 6. Nested grids used for the modelling of waves with SWAN.

was calculated so that the 50-yr value was reached at “Point difficulty when creating cyclonic wind fields is to reproduce
East”. The Tp and Dp parameters were not modified. the asymmetric structure of the cyclone, which is controlled
by many factors, including hurricane motion, environmental
3.2 Determination of the extreme wave scenarios conditions, vertical shear and surface friction (Phadke et al.,
for cyclonic waves 2003). For the three selected cyclones, the basic parameters
(track, maximum wind speed, radius of maximum wind, min-
3.2.1 Scenario building imum central pressure) were defined, and for Dina and Hol-
landa we also had the 30 kt wind radius, which yields infor-
The cyclonic wave scenarios correspond to historical cy- mation about the cyclone’s structure. Two parametric wind
clones selected on two criteria: their intensity (greater than field models were used. The first, based on Holland (1980),
category 4 on the Dvorak scale) and their track (very close reproduces symmetric cyclonic wind fields (idealized by
to the island). We chose three intense cyclones having tracks concentric circles) with the basic parameters. The second,
representative of the different possible configurations, as il- based on Xie et al. (2006), provides asymmetric cyclonic
lustrated in Fig. 5. Dina (January 2002) came from the north- wind fields. This latter model is a modification of Holland’s
east and passed north of Reunion Island (category 6). Colina approach in which the storm’s asymmetric structure is de-
(January 1993) came from the north and passed west of Re- picted by the radial extent of 30 kt winds in the four quad-
union Island (category 4). Hollanda (February 1994) came rants (north, east, south, and west). Xie et al. (2006) showed
from the northeast, and it is one of the rare cyclones to have that this approach significantly improves the accuracy of the
passed south of Reunion Island (category 4). Additionally, wind model, but its use is limited by the availability of infor-
cyclone Gamede (February 2007) was used for validation mation about the radial extent of the wind. Thus, for Colina
and to complete sensitivity tests because neither offshore nor we used the symmetric wind model and for Dina and Hol-
coastal wave gauges were operational during the three se- landa, we used the asymmetric model. The wind fields were
lected cyclones. gridded at 0.1◦ with a time step of 30 min.
3.2.2 Reconstitution of the wind fields
3.3 Wave modelling
The application of wave models to tropical cyclones re-
quires the reconstitution of a 2-D surface wind input over Seven wave scenarios were identified: four modified his-
the entire course of the storm. For this study, we used para- torical events for non-cyclonic waves (corresponding to a
metric wind models to create the wind fields. The main 50-yr Hs return period) and three historical cyclones. For

Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012 www.nat-hazards-earth-syst-sci.net/12/2425/2012/

S. Lecacheux et al.: A method to characterize the different extreme waves for islands 2431

Fig. 7. Positions of the 74 segments into which the Reunion Island

coastline has been divided. Fig. 8. Description of the process used to determine the equivalent
deep-water wave heights (H0 eq ) from the deep-water wave height
of the seven scenarios (Hs 0 ) using numerical modelling and linear
each scenario, the nearshore wave fields were simulated with theory equations (Eq. 2 to Eq. 4 in Sect. 3.4.2).
SWAN. This is a third-generation model devoted to the sim-
ulation of wave characteristics nearshore, as described in
Booij et al. (1999). It solves the spectral action balance around the island. The Reunion Island coastline was broken
equation and takes into account wave generation, propaga- down into 74 segments, as depicted in Fig. 7. A nearshore
tion, dissipation, and non-linear wave interactions. Figure 6 point on the 40-m isobath was associated with each segment
presents the three nested grid levels defining the model do- to extract the wave parameters (Hs , Tp , Dp ) from the R2 grids
main. The first grid (R0), covering a large portion of the for each scenario and at each computational time step. The
Indian Ocean, enables the wave fields to be generated with choice of the 40-m isobath was motivated by the accuracy
wind forcing alone. The spatial resolution is about 10 km, of the model nearshore. Indeed, although the resolution of
and the time step is one hour. The second grid (R1), with a the R2 grids is good (about 100 m), the very steep slopes of
1 km resolution, is centred on Reunion Island with a com- the bathymetry around the island reduce the sensitivity of the
putation time step of 30 min. Lastly, four grids (R2N, R2E, model close to the shore. The model’s accuracy for depths of
R2S, R2W) around Reunion Island describe the coastline and less than 40 m is expected to be insufficient at a few locations
the nearshore waters with a resolution of about 100 m and a along the coast (especially in the southeast).
computation time step of 30 min.
3.4.2 Determination of the equivalent deep-water waves
For the cyclonic scenarios, the simulations were per-
formed in a transitional mode to reproduce the entire course The “equivalent deep-water waves” were calculated from the
of the cyclones. The wave fields were first generated at grid local wave conditions extracted at a 40-m depth with the pro-
R0 with the parametric cyclonic wind only. Next, the results cedure summarized in Fig. 8. Here, we assume that the wave
were applied as boundary conditions in the nested grids to transformation due to refraction and shoaling follows the lin-
propagate the wave fields up to the four R2 (second rank) ear theory equations and that the depth contours are evenly
grids. For the non-cyclonic scenarios, the simulations were spaced and parallel to the shoreline. The complete mathemat-
performed in a stationary mode. First, the wave triplets (Hs , ical description is beyond the scope of this study but can be
Tp , Dp ) were propagated in grid R1 by forcing the bound- found in Dean and Dalrymple (1991). The basic equations
aries with the conditions established in Sect. 3.1.3. Then, the are:
four R2 grids were nested with the results of grid R1. s s
Cg1 cos(θ1 )
3.4 Local comparisons and extreme wave Hs2 = Hs1 Ks Kr with Ks = and Kr = (2)
Cg2 cos(θ2 )
characteristics assessment
gT22 2π d
3.4.1 Segmentation of the coast L2 = tanh( ) (3)
2π L2
The assessment of local extreme wave characteristics was where Hs1 and T1 are the wave height and period at a 40-m
performed for coastal segments of uniform orientation depth, and Hs2 , T2 and L2 are, respectively, the wave height,

www.nat-hazards-earth-syst-sci.net/12/2425/2012/ Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012

2432 S. Lecacheux et al.: A method to characterize the different extreme waves for islands

Fig. 9. Comparison between wave heights and periods recorded by the wave gauges at “Pointe du Gouffre” (left) and “Port Est” (right) and
the wave parameters calculated with SWAN using the parametric wind fields.

period and wavelength at depth d. Ks and Kr are, respec- and the term H02 eq L0 , which is a good indicator of wave en-
tively, the shoaling coefficient and the refraction coefficient, ergy.
where Cg is the group velocity and θ is the angle between the
wave crest and the depth contours. Deep-water wavelength
(L0 ) and period (T0 ) are assumed to be interdependent and 4 Results
linked by Eq. (4):
4.1 Cyclonic waves
gT 2
L0 = 0 . (4)
2π 4.1.1 Validation with cyclone Gamede
The curved shape of island coastlines, with differently ori-
ented sides, implies that waves may not be fully refracted The track of cyclone Gamede is plotted in Fig. 5. The
at a 40-m depth, within the model’s accuracy limitations. wind field was created with both symmetric and asymmet-
For this reason, the shore-normal approach cannot be di- ric models described in Sect. 3.2.2. Until 25 February 2007,
rectly assumed to back calculate the “equivalent deep-water Gamede’s structure remained quite symmetric. Then, the cy-
wave height”. Thus, given the initial wave conditions at 40-m clonic motion abruptly changed, thereby inducing asymme-
depth, the wave parameters were first determined for shal- try in the cyclone structure. We compared the Hs computed
lower depths with Eqs. (2) and (3) up to a depth limit cal- with both wind fields with the two coastal wave gauges mea-
culated as D lim = Hs /0.7 (to avoid exceeding the breaking surements at “Port Est” and “Pointe du Gouffre” when the
depth). The objective of this first step was to calculate the computed wave heights exceeded 3 m (see Fig. 9). If asym-
wave characteristics at very shallow depths, beyond the limit metric wind fields are used, the computed Hs is quite compa-
of resolution of the model, in order to approximate the fully rable to the buoy measurements: the rms difference is 0.63 m
refracted wave characteristics and minimize the wave inci- at “Pointe du Gouffre” and 0.9 m at “Port Est”. At the peak
dence angle. Then, the “equivalent deep-water wave heights” of the storm, both wave gauges indicated Hs values of 7 m,
H0 eq were calculated by dividing the wave heights calcu- while the simulations estimated 6.2 m at “Pointe du Gouf-
lated previously with the shoaling coefficient Ks . fre” and 6.4 m at “Port Est”. Concerning the wave periods,
since measurements of Tp were not available, we compared
3.4.3 Comparison between the scenarios the computed Tp with the recorded Ts (the average period of
the 1/3 highest waves). The ratio of Ts /Tp is approximately
The comparison between the sites and scenarios was based equal to 0.93, according to Goda (2000). For both cases,
on both the “equivalent deep-water wave heights” (H0 eq ) the computed wave periods compare well with the records.

Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012 www.nat-hazards-earth-syst-sci.net/12/2425/2012/

S. Lecacheux et al.: A method to characterize the different extreme waves for islands 2433

These tests show that the accuracy of the model with asym- southern side of the island. At a 40 m depth (on R2 grids), Hs
metric parametric wind fields is satisfactory. Using symmet- ranges between 5 and 6 m from the “Pointe de la Table” to the
ric wind fields produces larger errors when the structure of “Pointe des Aigrettes” and Tp reaches 17 s. On the northern
the cyclone becomes very asymmetric. We used the symmet- side of the island, Hs is very low.
ric model for cyclone Colina only. Considering that it was
a slow-moving cyclone with a translation speed of less than 4.3 Impact of the different wave regimes on the coast
7.5 m s−1 and a rather straight track, we assume that the sym-
metric model still yielded a good representation of the wind The objective of the comparison is to determine which zones
field. of Reunion Island will be most severely at risk of setup and
runup in cases of extreme events, and with respect to which
4.1.2 Results of the simulations wave regime.
Figure 11 presents the characteristics of the “equivalent
The results of the simulations at grid R1 are shown in Fig. 10 deep-water waves”, with the highest H02 eq × L0 obtained for
but the following results are extracted from the R2 grids. each regime with the method described in Sect. 3.4.2. We
Dina is the cyclone that generates the most energetic and distinguish three areas of influence:
widespread waves. The waves mainly affect the eastern and
the northern sides of the island. They are particularly a threat – On the large half of the island, extending from the north-
for the eastern side because they have a shore-normal di- west to the southeast (clockwise from segment 72 to
rection and therefore are very little refracted. At a depth of segment 34), Dina’s waves are dominant. H0 eq reaches
40 m, the most energetic waves occur at the “Pointe Champ a maximum of about 18 m near the “Pointe Champ
Borne”, with Hs up to 18 m and Tp about 17 s. Further south, Borne” (segment 15). In general, H0 eq is higher in the
Hs varies between 16 m and 12 m, depending on the orienta- eastern part of the area, with most of the values exceed-
tion of the coast. On the northern side, the wave angle of ing 13 m. In the northern part, H0 eq decreases rapidly
incidence is higher, but the waves are still very energetic down to 6 m towards the extreme west. T0 remains ho-
nearshore. Going from east to west, Hs decreases until 8 m. mogeneous in the area, with values of approximately
Cyclone Hollanda is an interesting event because it generates 17 s.
waves coming from the southeast that strongly impact the
southern side. From the “Pointe de la Table” to the “Pointe de – From the southeast to the southwest (clockwise from
l’Etang-Salé”, Hs decreases from 13 m to 8 m with Tp from segment 34 to segment 53), Hollanda becomes domi-
14 s to 12 s at a 40 m depth. The waves of cyclone Colina nant nearshore. Its original track, passing to the south-
affect only the northern side of the island, with shore-normal east of the island, induces the generation of waves from
waves having Hs values mainly between 10 m and 12 m and the southeast that are less dissipated than the ones gen-
Tp values around 14 s at 40 m depth. None of the three se- erated by Dina. Even if southern waves have longer pe-
lected cyclones generated important waves on the island’s riods in this area (T0 up to 17 s), the waves of Hollanda
western side. remain slightly more energetic, with a maximum H0 eq
of about 13.3 m at the “Pointe Langevin” (segment 40)
4.2 Non-cyclonic waves and an associated T0 of about 14 s. In the remainder of
the area, most of the H0 eq exceed 10 m.
4.2.1 Validation with the 13 May 2007 southern
wave event – From the southwest to the northwest (clockwise from
segment 54 to segment 71), cyclonic waves and south-
The 13 May 2007 southern wave event particularly affected ern waves (Case 3) have very similar energies. Indeed,
the western portion of the coast. The R2 simulations dis- this area is not directly exposed to cyclonic waves,
played good agreement with the measurements of the “Port which are very dissipated when reaching the coast, with
Ouest” and “Saint Pierre” wave gauges. At “Saint Pierre”, H0 eq up to 8 m but T0 not exceeding 12 s. On the other
the wave gauge measured a maximum Hs of 5.9 m, and the hand, southern waves have a shore-normal incidence,
simulation estimated 5.5 m. At “Port Ouest”, the maximum with H0 eq less than 6 m but high T0 of about 17 s. This
measured Hs was 3.9 m, while the simulation gave 3.8 m. area is the one with the lowest energy.

4.2.2 Results of the simulations The angles to the shore (θ) of the initial waves, extracted
at a 40-m depth, ranged from 1◦ to 70◦ . When applying the
Figure 10 also shows the results of the simulations at grid method described in Sect. 3.4.2., the angles at depth D lim
R1 for each non-cyclonic scenario. The trade-wind wave ranged for the most part from 1◦ to 30◦ , depending on the
scenario (Case 4) is clearly less energetic than the three position and the orientation of the segments. A few studies
other southern wave scenarios. Case 3 is the dominant show that for this range of angles, the impact of the wave an-
non-cyclonic event, with shore-normal waves impacting the gle on the amplitude of the setup or runup is either negligible

www.nat-hazards-earth-syst-sci.net/12/2425/2012/ Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012

2434 S. Lecacheux et al.: A method to characterize the different extreme waves for islands

Fig. 10. Wave fields modelled with SWAN for each scenario at grid R1. For the cyclonic scenarios (performed with a transitional mode), the
figure corresponds to the peak of the waves near Reunion Island. The colour bar indicates the amplitude of Hs and the vectors indicate wave
direction.

or at most minor. As to wave runup at dikes, the report of Van angles, the comparison between the “equivalent deep-water
der Meer and Janssen (1995) shows that when 0◦ < θ <30◦ , waves” for the different sites and scenarios is relevant in
long-crested waves cause almost the same runup as shore- terms of the impact of setup and runup.
normal waves, with a reduction factor exceeding 0.9. For the
wave setup and runup at beaches, little research has been per-
formed. Among the studies that do exist, Hsu et al. (2006) 5 Discussion
found that the percentage decrease is less than 20 % for in-
cident angles between 0◦ and 30◦ . Thus, with our range of This method calls for many steps and concepts which in-
troduce additional levels of uncertainty. The first level is

Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012 www.nat-hazards-earth-syst-sci.net/12/2425/2012/

S. Lecacheux et al.: A method to characterize the different extreme waves for islands 2435

Fig. 11. Characteristics of the most energetic “equivalent deep-water waves” (characterized by the highest H02 eq × L0 ) obtained for each
regime. In the lowest panel, the scenario producing the most energetic waves within each regime is distinguished.

associated with the data (cyclone best tracks and global wave a return period. Statistical approaches exist for calcu-
re-analysis for non-cyclonic waves) and their conversion into lating the return periods of cyclonic waves. Classically,
wave scenarios. The second level relates to the numerical a synthetic database is first generated with a statistical
modelling of nearshore waves (with limitations linked to the model, calibrated against a historical dataset in order to
simplified representation of physical processes and bathy- enlarge the sample of recorded storms. Then, the wave
metric data). The last level is due to the use of linear theory fields are modelled for each storm in the database, and
equations to calculate the “equivalent deep-water waves”, a statistical analysis is performed on the results to de-
considering idealized waves and bathymetry. From the au- termine the extreme values (James and Mason, 2005;
thors’ points of view, the most important source of uncer- Hardy et al., 2003). This approach is resource intensive
tainty lies in the first level: because it requires (1) the development of a statistical
model representative of a specific cyclonic basin and (2)
– For non-cyclonic waves, the lack of long-term re- large computer capacities. This is the reason it was not
analysis, with a valid representation of extreme waves, applied in this study, which aims at presenting a general
remains an obstacle to achieving a reliable analysis method. However, the application of the present method
of extremes. In this study, the statistical analysis was with a probabilistic approach for both cyclonic and non-
performed on a 13-yr interval of data and was set to cyclonic cases would enable the analysis of the results
a 50-yr return period. The use of a longer time se- to be pursued further as it would improve the accuracy
ries would enable the statistical analysis to be extended of the comparison between the extreme waves from the
to the 100-yr value, which is the one conventionally different regimes.
used in coastal engineering for the design of protec-
tive structures. A new re-analysis, with a time coverage
expanded to 1988–2011, is currently produced through 6 Conclusions
the IOWAGA project (http://www.ifremer.fr/iowaga). It
uses the CFSR dataset (NCEP Climate Forecast System A general approach was proposed for determining the ex-
Re-analysis), which is efficient for reproducing extreme treme wave events that may cause most damage along the
events (Saha et al., 2010). The use of such a re-analysis various coastlines of tropical islands. It provides local in-
in future work would certainly improve the statistical formation about extreme waves from each wave regime all
analysis of non-cyclonic waves. around the island and enables (1) inter-comparison of these
extremes for a particular location and (2) assessment of the
– For the cyclonic scenarios, we chose to base the analy- alongshore variability in wave exposition. We think that this
sis on historical cases alone, which were not related to method is particularly appropriate for complex environments

www.nat-hazards-earth-syst-sci.net/12/2425/2012/ Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012

2436 S. Lecacheux et al.: A method to characterize the different extreme waves for islands

presenting various origins of extreme waves. First, the clas- References

sification of waves into different regimes enables improve-
ment of the data homogeneity required for the analysis of Becker, J. J., Sandwell, D. T., Smith, W. H. F., Braud, J., Binder, B.,
extremes compared to classical methods using a single cri- Depner, J., Fabre, D., Factor, J., Ingalls, S., Kim, S.-H., Ladner,
R., Marks, K., Nelson, S., Pharaoh, A., Trimmer, R., Von Rosen-
terion such as angular sectors or seasons. Then, this process
berg, J., Wallace, G., and Weatherall, P.: Global Bathymetry and
provides a complete overview of the different kinds of ex- Elevation Data at 30 Arc Seconds Resolution: SRTM30 PLUS,
treme waves that may threaten the coast, which is useful for Mar. Geodesy, 32, 4, 355–371, 2009.
planners and decision-makers. While cyclonic waves usually Booij, N., Ris, R.C., and Holthuijsen, L.H.: A third-generation wave
cause most reported wave-induced damages, southern and model for coastal regions, Part I, Model description and valida-
trade-wind waves do disrupt coastal infrastructure or road- tion, J. Geophys. Res., 104, C4, 7649–7666, 1999.
maintenance activities during the Southern Hemisphere win- Brodtkorb, P. A., Johannesson, P., Lindgren, G., Rychlik, I., Rydén,
ter. J., and Sjö, E.: WAFO – a Matlab toolbox for analysis of random
The exposure of each part of the coastline to different waves and loads, in: Proceeding of the 10th Int. Offshore and
categories of extreme events was determined for the pilot Polar Eng. Conf., Seattle, USA, III, 343–350, 2000.
site of Reunion Island. Offshore, cyclonic waves appear to Butel, R., Dupuis, H., and Bonneton, P.: Spatial variability of wave
conditions on the French Atlantic coast using in-situ data, J.
be the most energetic waves around the island. However,
Coast. Res., 36, 96–108, 2002.
for the western part of the coast, cyclonic waves are sub- Caires, S., Sterl, A., Bidlot, J. R., Graham, N., and Swail, V.: In-
ject to a strong refraction, inducing an important loss of en- tercomparison of different wind-wave reanalyses, J. Climate, 17,
ergy nearshore. Thus, non-cyclonic waves with longer peri- 1893–1912, 2004.
ods may finally become as energetic as cyclonic waves when Cazes-Duvat, V. and Paskoff, R.: Les littoraux des Mascareignes
reaching the coast and generate a similar runup. entre nature et aménagement, L’Harmattan, 186 pp., 2004.
Our method can be easily implemented for other case stud- Dean, R. G. and Dalrymple, R. A.: Water wave mechanics for en-
ies because it can be adapted depending on the data availabil- gineers and scientists, Singapore: World Scientific, 2nd Edn.,
ity. In this study, we applied this method at a regional scale, 353 pp., 1991.
using data with a global coverage (global re-analysis, cyclone Goda, Y.: Random Seas and Design of Maritime Structures, World
best tracks, global bathymetric databases, etc.) to provide an Scientific, 443 pp., 2000.
Hardy, T. A., McConochie, J. D., and Mason, L. B.: Modeling tropi-
overview of extreme waves all around the island. The results
cal cyclone wave population of the Great Barrier Reef, J. Waterw.
obtained with this approach may be used for the mapping Port. C-Asce, 129, 2, 104–113, 2003.
of marine submersion hazards by calculating the associated Holland, G.: An analytic model of the wind and pressure profiles in
wave setup and runup with empirical formulas. However, the hurricanes, Mon. Wea. Rev., 108, 1212–1218, 1980.
main limitation of this application for tropical islands is the Hsu, T.-W., Hsu, J.-R.-C., Weng, W.-K., Wang, S.-K., and Ou, S.-
presence of fringing reefs and small bays along some parts of H.: Wave setup and setdown generated by obliquely incident
the coast, for which empirical formulas are not adapted. The waves, Coast. Eng., 53, 865–877, 2006.
use of wave-by-wave models able to simulate the full pro- James, M. and Mason, L.: Generation of a synthetic tropical cyclone
cess of runup and submersion may be more suited to this case database, in: Coasts $ Ports ‘99. Challenges and directions for
but requires fine topo-bathymetric data and higher computa- the new century, proceedings of the 14th Australasian Coastal
tional capacities. Thus, the assessment of marine submersion and Ocean Engineering conference and the 7th Australasian Port
and Harbour conference, Barton, A.C.T., National committee on
maps in this context requires working at higher resolution
Coastal Ocean Engineering, Insitution of Engineers, Australia,
and adapting the method case by case. 381–386, 1999.
Le Cozannet, G., Lecacheux, S., Delvallee, E., Desramaut, N., Oliv-
eros, C., and Pedreros, R.: Teleconnection Pattern Influence on
Acknowledgements. We wish to thank Valérie Monbet and Pierre
Sea-Wave Climate in the Bay of Biscay, J. Climate, 24, 641–652,
Ailliot for kindly providing the METIS toolbox and for useful
2011.
recommendations on sea-wave classifications. We are also grateful
Lee, D.-Y. and Jun, K.-C.: Estimation of design wave height for the
to Hubert Quetelard from Météo-France (Réunion) for making
water around the Korean peninsula, Ocean Sci. J., 41, 4, 245–
available the cyclone best-track database and the DDE for providing
254, 2006.
the buoy measurements. This study was supported by the DREAL
Mackay, E. B. L., Challenor, P. G., and Bahaj, A. S.: On the use
La Réunion, and the BRGM research project RISCOTE.
of discrete seasonal and directional models for the estimation of
extreme wave conditions, Ocean Eng., 37, 5–6, 425–442, 2010.
Edited by: S. Tinti
Mathiesen, M., Goda, Y., Hawkes, P. J., Mansard, E., Martin, M. J.,
Reviewed by: two anonymous referees
Peltier, E., Thompson, E. F., and Van Vledder, G.: Recommended
practice for extreme wave analysis, J. Hydraul. Res., 32, 6, 803–
814, 1994.
Menendez, M., Mendez, F.J., Izaguirre, C., Luceno, A., and Losada,
I. J.: The influence of seasonality on estimating return values of
significant wave height, Coast. Eng., 56, 3, 211–219, 2009.

Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012 www.nat-hazards-earth-syst-sci.net/12/2425/2012/

S. Lecacheux et al.: A method to characterize the different extreme waves for islands 2437

Monbet, V. and Aillot, P.: Metocean Time Series (METIS) Stockdon, H. F., Holman, R. A., Howd, P. A., and Sallenger Jr., A.
Toolbox Documentation, Ifremer, 52 pp., available at: http: H.: Empirical parameterization of setup, swash and runup, Coast.
//www.univ-ubs.fr/adminsite/objetspartages/liste fichiergw.jsp? Eng., 53, 573–588, 2006.
OBJET=DOCUMENT&CODE=1222201292142&LANGUE= Tolman, H. L.: User manual and system documentation of
1, 2005. Wavewatch-III version 2.22., NOAA/NWS/NCEP/MMAB Tech-
Morton, I. D., Bowers, J., and Mould, G.: Estimating return period nical Note, 222, 133 pp., 2002.
wave heights and wind speeds using a seasonal point process Van der Meer, J. W. and Janssen, J. P. F. M: Wave run-up and wave
model, Coast. Eng., 31, 1–4, 305–326, 1997. overtopping at dikes, in: American Society of Civil Engineers,
Panigrahi, J. K. and Misra, S. K.: Numerical hindcast of extreme Wave forces on inclined and vertical wall structures, 1–27, 1995.
waves, Nat. Hazards, 53, 361–374, 2010. Vitousek, S. and Fletcher, C. H.: Maximum annually recurring wave
Phadke, A. C., Martino, C. D., Cheung, K. F., and Houston, S. H.: heights in Hawaii, Pacific Sci., 62, 4, 541–553, 2008.
Modeling of tropical cyclone winds and waves for emergency Vitousek, S., Fletcher, C. H., and Barbee, M.: A practical approach
management, Ocean Eng., 30, 4, 553–578, 2003. to mapping extreme wave inundation: consequences of sea-level
Ruggiero, P., Komar, P. D., McDougal, W. G., Marra, J. J., and rise and coastal erosion, in: Proceedings of Solutions to Coastal
Beach R. A.: Wave runup, extreme water levels and the erosion Disasters, Oahu, Hawaii, 13–16 April, 85–96, 2008.
of the properties backing beaches, J. Coast. Res., 17, 407–419, Xie, L., Bao, S., Pietrafesa, L. J., Foley, K., and Fuentes, M.: A real-
2001. time hurricane surface wind forecasting model: formulation and
Saha, S., Moorthi, S., Pan, H.-L., Wu, X., Wang, J., Nadiga, S., verification, Mon. Wea. Rev., 134, 1355–1370, 2006.
Tripp, P., Kistler, R., Woollen, J., Behringer, D., Liu, H., Stokes,
D., Grumbine, R., Gayno, G., Wang, J., Hou, Y.-T., Chuang H.-
Y., Juang, H.-M. H., Sela, J., Iredell, M., Treadon, R., Kleist,
D., Van Delst, P., Keyser, D., Derber, J., Ek, M., Meng, J., Wei,
H., Yang R., Lord, S., Van Den Dool, H., Kumar, A., Wang,
W., Long, C., Chelliah, M., Xue, Y., Huang, B., Schemm, J.-K.,
Ebisuzaki, W., Lin, R., Xie, P., Chen, M., Zhou, S., Higgins, W.,
Zou, C.-Z., Liu, Q., Chen, Y., Han, Y., Cucurull, L., Reynolds, R.
W., Rutledge, G., and Goldberg, M.: The NCEP Climate Fore-
cast System Reanalysis, B. Am. Meteor. Soc., 91, 1015–1057,
doi:10.1175/2010BAMS3001.1, 2010.

www.nat-hazards-earth-syst-sci.net/12/2425/2012/ Nat. Hazards Earth Syst. Sci., 12, 2425–2437, 2012