Navigation – Plan du site

Storminess interannual variability and coastal hazards over the south-western Spanish coast: links to large scale atmospheric forcing

Variabilité interannuelle des tempêtes et des risques côtiers dans le sud-ouest de la côte espagnole : relations avec le forçage atmosphérique à grande échelle
Theocharis A. Plomaritis, Javier Benavente, Irene Laiz et Laura del Río
p. 275-286

Résumés

Dans le contexte de l’augmentation des risques côtiers, à travers la variabilité de la fréquence et de l’intensité des tempêtes, les dommages aux infrastructures côtières et les changements morphologiques des côtes sont mises en rapport avec la prise en compte conjointe du niveau moyen de la mer, des surcotes dues aux tempêtes, des valeurs maximales de hauteur de vague et des valeurs du jet de rive. Afin de mieux comprendre les processus physiques qui génèrent la variabilité interannuelle de ces paramètres, une réanalyse des enregistrements (HIPOCAS) de ces 44 dernières années a été effectuée. La série temporelle HIPOCAS a été validé avec des données de vents réels, de vagues et de niveau de la mer en utilisant des méthodes de corrélation linéaire et vectorielle. Dans ce travail, les changements de la durée, de la fréquence, de la chronologie et de la direction d’approche des tempêtes atlantiques sur la côte espagnole du golfe de Cadix (SO de la Péninsule Ibérique) ont été identifiés par le calcul de diverses caractéristiques des tempêtes, telles que la hauteur maximale des vagues, l’énergie totale par tempête et le regroupement des tempêtes. Les séries temporelles obtenues ont été comparées avec des indices atmosphériques à petite échelle, comme l’oscillation nord-atlantique (NAO) et le modèle de l’Atlantique Est (EA). Les résultats montrent une bonne corrélation entre les valeurs négatives de la NAO et l’augmentation des tempêtes dans tout le Golfe de Cadix. De la même façon, les valeurs négatives de la NAO ont été corrélées avec les valeurs hautes du niveau moyen de la mer et avec des épisodes de vent d’ouest et du sud-ouest, qui indiquent une plus forte probabilité de risque côtiers. Ces résultats ont été comparés avec des événements d’inondation côtière enregistrés dans la région de Cadix au cours des derniers hivers.

Haut de page

Entrées d’index

Mots-clés :

ONA, EA, tempêtes, Cadiz

Keywords :

NAO, EA, storms, Cadiz
Haut de page

Notes de la rédaction

Article soumis le 18 mai 2011, accepté le 5 septembre 2011.

Texte intégral

The authors would like to acknowledge Puertos del Estado for providing the wave buoy data for the coastal area of Cadiz and CEDEX for as well as providing the HIPOCAS data. This work has been supported by the MICORE project (EU grant FP7-ENV-2007-1-202798). It was financially supported by the Spanish Ministry of Science and Technology ("Juan de la Cierva Programme" and project "CLI-CGL2008-04736") and by the research group RNM-337 of the PAI. TAP, JB and LdR were also supported by the research group RNM-328 of the PAI. The contribution of the two reviewers in the improvement of the manuscript is greatly acknowledged.

Introduction

1Storm events are considered the main cause for shoreline erosion in many areas worldwide (Fenster et al., 2001). This way, they are attributed an essential role in coastal long-term evolution (decadal and centennial) despite the usually short time scale of their action (Morton et al., 1995). Land and sea are interacting constantly in response to natural forces. On a global basis, more than 70% of the world’s sandy beaches have experienced net erosion over the past decades (Bird, 1985). Irrespective of the causes, there is a growing socioeconomic need towards innovative coastal management and evaluation of the risks associated with the development in coastal plains. Storm activity and sea level changes can pose a risk on the coastal environment by affecting sediment transport and coastal processes over the nearshore areas. In open coastlines with large fetches, such as the Spanish Atlantic coast, large scale atmospheric phenomena are the main source of storminess variability. Hence, the interannual variability or long term trends of the above processes can affect the risk distribution over a particular stretch of coastline.

2Specific trends in wave heights have long been observed in the North Atlantic (Carter and Draper 1988; Bacon and Carter, 1991; Dupuis et al., 2006) and in the Northeast Pacific (Allan and Komar, 2000). Furthermore, Bacon and Carter (1993) observed a correlation between North Atlantic meridional atmospheric pressure gradient and wave height. Woolf et al. (2002) established a relationship between wave height anomalies and large scale atmospheric pressure patterns over the Northeast Atlantic on the basis of satellite altimetry. More precisely they attributed part of the variability to the North Atlantic Oscillation (NAO) and secondarily to the East Atlantic Pattern (EA). Finally, Dodet et al. (2010) presented a larger influence of the NAO over the south area of Europe using a 60 year long wave model forced by the 6h wind field from National Centres of Environmental Prediction (NCEP) reanalysis project (Kalnay et al., 1996).

3Traditionally the NAO and other climatic indices of the Atlantic Ocean have been mainly linked with temperature, precipitation and large scale circulation patterns over Europe. Recent studies have also focused on analyzing the response of sea level variability to NAO, but always at a broad spatial scale (Efthymiadis et al., 2002; Woolf et al., 2003; Tsimplis and Shaw, 2008; Tsimplis et al., 2009). Main findings are that the NAO influence on sea level is dominant in winter and represents one of the causes of the high interannual variability of sea level during this season. Moreover, D.K. Woolf et al. (2003) suggest that NAO effects are probably similar in the open ocean and along coastlines in large geographical areas, although they might sometimes be masked by local phenomena. Regional correlations between mean monthly wave height and NAO values have been performed by various researchers over the Atlantic coast of Europe (Woolf et al., 2003; Dodet et al., 2010). The impact of storms on shoreline variability has been widely demonstrated by various researchers (e.g., Morris et al., 2001; Cooper et al., 2004; List et al., 2006). However, a direct relationship between interannual wave variability and coastal response has not been linked to NAO because of the general lack of long-term and detailed coastal topographic data and the influence of other local and regional aspects on coastal changes, such as, geological framework (Jackson et al., 2005). Recently O’Connor et al. (2011) demonstrated a tentative link between coastline topography and NAO-modulated external forcing, by focusing in small and well constrained tidal inlets of Northern Ireland. On the much longer data-set of Narrabeen Beach (23 years of monthly beach profiles) beach rotation was linked to the variability of wave characteristics, positively related with the Southern Oscillation Index (SOI) (Short et al., 2000).

4Although it is well known that NAO affects the storm tracks over the Atlantic (Keim et al., 2004) an in-depth investigation of the influence on the storminess and local sea level variability and the associated coastal hazards in the Gulf of Cadiz has not yet been undertaken. The present work focuses in the coastal area of the Gulf of Cadiz and goes beyond the separate analysis of wave and mean sea level as it investigates the combined occurrence of the phenomena and their contribution to the increasing severity of coastal hazards. Furthermore, other aspects of storminess are examined such as the duration of the storms, the storm direction and the total amount of storm hours in a month. A description of the study area, wave and sea level data sets is presented, followed by a section focused on the validation exercise and correction fittings that were employed in order to improve the prediction of the models during storm events. Then the interannual variability of the record is presented and correlations with large scale climatic indices are undertaken. Finally, discussion and conclusions of the results obtained and comparison with other work are detailed.

Study Area

5The Gulf of Cadiz is the sub-basin that connects the Atlantic Ocean with the Mediterranean Sea through the Strait of Gibraltar. Its northern and southern boundaries are, respectively, the southwest coast of the Iberian Peninsula and the Atlantic coast of Morocco (fig. 1).

Fig. 1 – Bathymetric map of the study area showing the HIPOCAS data grid cell and the location of the coastal buoy of Cadiz managed by Puertos del Estado.
Fig 1Carte bathymétrique de la zone d'étude indiquant la cellule de la grille des données HIPOCAS et l'emplacement de la bouée côtière de Cadix géré par Puertos del Estado.

Fig. 1 – Bathymetric map of the study area showing the HIPOCAS data grid cell and the location of the coastal buoy of Cadiz managed by Puertos del Estado. Fig 1 – Carte bathymétrique de la zone d'étude indiquant la cellule de la grille des données HIPOCAS et l'emplacement de la bouée côtière de Cadix géré par Puertos del Estado.

6As an Atlantic open coast it is influenced by large scale oceanic weather systems both in terms of precipitation, wind and large-fetch waves. The storms generated by these systems are the principal cause of storm erosion in the area (Del Río et al., 2012). On a local scale, the orientation of the coastline and the local physiographic characteristics result in sheltering effects to the north component winds and funnelling effects to south and east component winds due to the complex relief of the Strait of Gibraltar (Dorman et al., 1995). The prevailing wind and wave fields are from WSW directions, with a yearly average significant wave height of 1 m comprised of both sea and swell, generating a predominant longshore current toward the E and SE (Benavente et al., 2000). Long-term average longshore drift has been calculated for Huelva region (Western Gulf of Cadiz) by means of modelling to around 260,000 m3/yr (Medina, 1991). Changes in shoreline orientation along the Gulf coast greatly influence the approach angle of waves, which diminishes progressively toward the South, generating less significant littoral currents close to the Strait of Gibraltar and weaker longshore drift. In terms of tides the area can be described as semidiurnal and mesotidal with progressive tidal wave characteristics and a mean tidal range of 2.20 m. The surface circulation over the continental shelf is mainly wind-driven but it is also affected by local forcing mechanisms, such as the Guadalquivir River discharge (see García-Lafuente and Ruiz, 2007, for a review). It is subject to seasonal and interannual variations deeply related to the seasonal variability of the open sea circulation (Criado-Aldeanueva et al., 2009). The latter is largely affected by the large-scale atmospheric patterns over the Atlantic Ocean, roughly represented by the NAO index (Criado-Aldeanueva et al., 2009; García-Lafuente and Ruiz, 2007).

Methodology

Data Description

7In the present work hindcast model data generated by Puertos del Estado (Spain) for the nearshore area of Cadiz was employed (fig. 1). The dataset (HIPOCAS data hereinafter) consists of a 44-year reanalysis of meteorological, wave and sea level variables spanning between January 1958 and December 2001 (Guedes-Soares et al., 2002). For the wave simulation over the Gulf of Cadiz a grid of 5’ was used nested to a larger WAN model of the Atlantic Ocean. The wave model was initially forced by the NCEP reanalysis wind fields (Kalnay et al., 1996). The ocean circulation and sea level variations were simulated with the HAMSOM model over a grid of 10’ x 15’ taking into account wind and pressure forcing. The hydrodynamic model was forced with boundary conditions from the REMO atmospheric model (Sebastiao et al., 2008) that was in turn forced with NCEP data. The output time-step was 3 hours for all the parameters. Detailed description of the model setup described above can be found in Rusu et al. (2008) and Sebastiao et al. (2008).

Model Validation

8An extensive validation exercise was undertaken by Menendez et al. (2004) between the HIPOCAS data and wave buoy data collected around the Spanish coasts. However, in order to optimize the results in the Gulf of Cadiz, a new correction was applied in the present study that consisted in fitting the model wave data to observations focusing mainly in the case of storm conditions. Data from the Puertos del Estado coastal buoy of Cadiz were used for this purpose (fig. 1). In order to obtain a statistically independent data set of storm conditions a peak over threshold analysis (POT) was used for the period with simultaneous model and observation data (Kamphuis, 2000). The above analysis produced a correction based on the peak values of each storm and was then applied to the entire data set. The threshold value for the POT analysis was set as 1.5 m wave height, which is the threshold for Cadiz storms according to the local authorities, and a storm independence time (time between two consecutive storms) of 24 hours was used. Higher storm threshold values proposed for the area of Cadiz (Del Río et al., 2012; Ribera et al., 2011) were also used but the fitting coefficient obtained was not significantly different.

9Although overlapping of directional wave data between the buoy of Cadiz and the HIPOCAS data only exists over the year 2001, the period of time is long enough to cover both calm and stormy seasons. The agreement of the two directional time series was tested using the Kundu vector correlation (Kundu, 1976) between the two data sets. This correlation method produces a coefficient between the directional wave heights of the two time series and the main angle through which the first series would have to be rotated anticlockwise to match the direction of the second series.

10The sea level reanalysis carried out along the coasts of the Iberian Peninsula presented good results (Sebastiao et al., 2008). For example, in the area of the Gulf of Cadiz the results were under-predicting the observations only for the extreme peak of the storms. However, this occurs because the tidal gauges are located close to estuary mouths and their water levels are locally affected by the river discharge and they do not influence the general surge level on the continental shelf. No further validation was applied to the sea level because the calibrated time series was the only available in the area for the reanalysis period.

Data Analysis

11The corrected time series was used to calculate the monthly, seasonal and annual values of the wave heights and directions and sea level in order to identify the wave climatology in the area. Furthermore, storminess characteristics were calculated on a monthly basis in order to be compared with large scale atmospheric indices. Such characteristics were the total number of hours when the storm threshold was exceeded (Storm Index), the total storm energy per month and the number of storms per month. Correlations of the above values with the atmospheric indices were performed. For the calculation of join probability (JP) a POT analysis was used in order to construct a contingency table with storm events and the associated storm surges. The thresholds of the POT analysis were set as above.

Results and Discussion

Validation

12The corrected values (Hs_c) improved the re-analysis data for the entire period of the storm and especially for storms higher than 2 m, where the HIPOCAS data had significantly and systematically overestimated the wave height. Typical results are shown in figure 2 for March 1995. It can be observed that the corrected HIPOCAS data have clearly improved (rmse 0.34) when compared with the un-corrected data (rmse 0.47).

13Differences in wave propagation direction between the measured and modelled data are presented against the significant wave height in figure 3.

Fig. 2 – Comparison between modeled (HIPOCAS), measured (buoy), and corrected data using POT analysis for significant wave height (Hs).
Fig. 2Comparaison entre des données modélisées (HIPOCAS), mesurées (bouée) et corrigées en utilisant l'analyse POT pour la hauteur significative des vagues (Hsig).

Fig. 2 – Comparison between modeled (HIPOCAS), measured (buoy), and corrected data using POT analysis for significant wave height (Hs). Fig. 2 – Comparaison entre des données modélisées (HIPOCAS), mesurées (bouée) et corrigées en utilisant l'analyse POT pour la hauteur significative des vagues (Hsig).

Fig. 3 – Comparison of the difference between mean wave direction for the corrected (HIPOCAS) and measured (Buoy) data for 2001.
Fig. 3Comparaison de la différence de la direction moyenne des vagues pour les données corrigées (HIPOCAS) et mesurées (bouée) pour 2001.

Fig. 3 – Comparison of the difference between mean wave direction for the corrected (HIPOCAS) and measured (Buoy) data for 2001. Fig. 3 – Comparaison de la différence de la direction moyenne des vagues pour les données corrigées (HIPOCAS) et mesurées (bouée) pour 2001.

Grayscale presents the data density.
L’échelle des niveaux de gris montre la densité des données.

14There is a large scatter for wave heights smaller than 1m. However, as it can be seen by the density of the plot, large differences are only present over a few events, while the majority of the data show small deviations. Such deviations are reduced to a variance of 4.5 degrees for waves between 1 m and 1.5 m also with larger densities concentrated in small differences. For the larger wave heights the scatter is minimized and the HIPOCAS data strongly agree with the observed data. The correlation coefficient obtained is 0.71 (perfect correlation is 1) with an average angle difference of 4.67 degrees. In figure 4 the zonal and meridional components of the waves are plotted for the HIPOCAS, observed and corrected data.

Fig. 4 – Comparison between modeled (HIPOCAS), measured (buoy), and corrected data using POT analysis for (A) the zonal (East-West) and (B) the meridional components of the significant wave height (Hsig).
Fig. 4Comparaison entre des données modélisées (HIPOCAS), mesurées (bouée) et corrigées en utilisant l'analyse POT pour (A) la composante de la zone (Est-Ouest) et (B) la composante méridienne de la hauteur significative des vagues (Hsig).

Fig. 4 – Comparison between modeled (HIPOCAS), measured (buoy), and corrected data using POT analysis for (A) the zonal (East-West) and (B) the meridional components of the significant wave height (Hsig). Fig. 4 – Comparaison entre des données modélisées (HIPOCAS), mesurées (bouée) et corrigées en utilisant l'analyse POT pour (A) la composante de la zone (Est-Ouest) et (B) la composante méridienne de la hauteur significative des vagues (Hsig).

15A good agreement is observed between both components, particularly for significant storm events of southwest directions, which are the main oceanic storm direction for the Gulf of Cadiz. By rotating the HIPOCAS data by 4.67 degrees an improvement of the east-west component of the wave is observed mostly during storm conditions (fig. 4, around 10/22/2001). The locally generated small storms with predominant southeast directions are not well represented probably due to spatial constraints of the atmospheric forcing than can not resolve the local east wind acceleration over the Strait of Gibraltar (fig. 4, around 08/011/2001). The above locally generated storms are not affecting the eastern coastline of the Gulf of Cadiz, where this study is focusing, because of the small fetches and its general orientation. However, the above events can generate coastal erosion events further west, over the Portuguese coast of Algarve region (Garcia et al., 2005).

Analysis

Wave Climate

16The mean annual cycle for the corrected significant wave height (Hs_c) and the associated wave directions for the coastal area of the city of Cadiz are presented in figure 5.

Fig. 5 – Average seasonal cycle for the entire reanalysis period of (A) Significant wave height and (B) Mean wave direction.
Fig. 5Cycle saisonnier moyenne pour la période entière de réanalyse de (A) la hauteur significative des vagues et (B) la direction moyenne des vagues.

Fig. 5 – Average seasonal cycle for the entire reanalysis period of (A) Significant wave height and (B) Mean wave direction. Fig. 5 – Cycle saisonnier moyenne pour la période entière de réanalyse de (A) la hauteur significative des vagues et (B) la direction moyenne des vagues.

Note: correction values for the Gulf of Cadiz described in section "Model validation" were applied in both cases.
Note : les valeurs de correction pour le Golfe de Cadix décrites dans la section "Validation du modèle" ont été appliquées dans les deux cas.

17The average wave heights over the area are higher during the winter months and part of the autumn. From the comparison of Hs_c and their associated directions it can be seen that there is a high energy period of the year that starts in November and extends up to March when the mean wave heights are higher with average wave height values close to 1m and they have more southerly directions. These values classify the coastline as a low energy one according to Tanner (1960) and Hegge et al. (1996). Over the rest of the period mean wave height is significantly lower with more westerly directions. This annual variability represents the typical wave climatology of the region. A similar pattern is also observed for the mean monthly wave periods with higher values during the stormy season and lower values during the rest of the year. Based on the above results the wave climate in the Gulf of Cadiz can be separated in a storm (November - March) and a calm (April - October) season.

18The monthly means can deviate substantially from the average seasonal cycle especially during the storm season (mainly in December and January), when mean monthly values of Hs up to 2 m can be observed. Both the storm mean monthly values (mean value of significant wave height that exceeds the storm threshold) and their anomalies for the whole year and for the storm seasons were correlated with the NAO index for significant wave height, wave direction and residual sea level.

Correlations with Climate Indices

19The results (tab. 1) show a weak correlation between the annual Hs_c and NAO that increases to -0.58 for the storm season months (fig. 6A).

Tab. 1 – Comparison of the correlation coefficients between mean monthly values and anomalies and NAO index for wave height (Hs_c), wave direction (Dir) and mean sea level (Niv).
Tab. 1Comparaison des coefficients de corrélation entre les valeurs moyennes mensuelles, les anomalies et l'indice NAO pour la hauteur des vagues (Hs_c), la direction des vagues (Dir) et le niveau moyen de la mer (Niv).

 

Hannual

Hstorm

Dirannual

Dirstorm

Nivannual

Nivstorm

Mean Values

-0,32

-0,58

-0,11

-0.136*

-0,37

-0,67

Anomalies

-0,43

-0,59

-0,12

-0.133*

-0,42

-0,7

Significance levels are < 99%.
Les niveaux de confiance sont < 99 %.

Fig. 6 – Correlations between NAO index and (A) mean monthly significant wave height, (B) Storm Index (number of hours that the wave height exceeds the storm threshold) and (C) residual monthly mean sea level, for the months of the storm season (November - March). (D): Correlation between the Storm Index and the mean monthly significant wave height.
Fig. 6Corrélations entre l'indice NAO et (A) la hauteur moyenne mensuelle de vague significative, (B) le Storm Index (nombre d'heures que la hauteur des vagues dépasse le seuil de tempête) et (C) le résidu du niveau moyen de la mer mensuel, pour les mois de la saison des tempêtes (Novembre - Mars). (D) Corrélation entre le Storm Index et la hauteur significative des vagues moyenne mensuelle.

Fig. 6 – Correlations between NAO index and (A) mean monthly significant wave height, (B) Storm Index (number of hours that the wave height exceeds the storm threshold) and (C) residual monthly mean sea level, for the months of the storm season (November - March). (D): Correlation between the Storm Index and the mean monthly significant wave height. Fig. 6 – Corrélations entre l'indice NAO et (A) la hauteur moyenne mensuelle de vague significative, (B) le Storm Index (nombre d'heures que la hauteur des vagues dépasse le seuil de tempête) et (C) le résidu du niveau moyen de la mer mensuel, pour les mois de la saison des tempêtes (Novembre - Mars). (D) Corrélation entre le Storm Index et la hauteur significative des vagues moyenne mensuelle.

Grayscale represents the data density.
L’échelle de nuances de gris représente la densité des données.

20The correlations between the mean monthly anomalies and NAO show a significant increase (p < 0.02) only for the annual values, suggesting that NAO has an effect on the wave variability that prevails during calm conditions. For the case of storm wave height the increase in the correlation is not statistically significant. Weak but significant correlations were observed between NAO and annual wave directions. However, over the storm seasons the significance levels decreased due to the lower number of observation (220 instead of 528) that influences the degrees of freedom of the correlations. The residual water levels also presented significant and strong correlations for the storm season months. This is due to the higher abundance of low pass fronts and the increase in wind speed due to the southern shift of the Atlantic storm tracks. All correlations have a negative sign because wave height and residual sea level increase with negative NAO values. For the case of EA the above parameters show weak but significant correlations (tab. 2) among which the most pronounced is that of the significant wave height during the storm months.

Tab. 2 – Comparison of the correlation coefficients between mean monthly values and anomalies and EA for wave height, wave direction and mean sea level. Confidence levels are < 99%.
Tab. 2Comparaison des coefficients de corrélation entre les valeurs moyennes mensuelles, les anomalies et l'indice EA pour la hauteur des vagues, la direction des vagues et le niveau moyen de la mer. Les niveaux de confiance sont < 99 %.

 

Hannual

Hstorm

Dirannual

Dirstorm

Nivannual

Nivstorm

Mean Values

0,2

0,34

0,15

0,21

-0,17

-

Anomalies

0,24

0,35

0,16

0,21

-0,16

-

21In all cases the correlation coefficients were smaller than with NAO, as expected, since the EA is the second prominent mode of low-frequency variability over the North Atlantic (Barnston and Livezey, 1987). In this case the correlations are positive for wave height and direction, since positive EA values are responsible for zonally extended storm tracks that affect the southern coasts of Europe (Wettstein and Wallace, 2010).

22Slightly higher correlations (-0.67, p < 0.01) were obtained between the Storm Index and the NAO for the storm season (fig. 6B), and very high between the Storm Index and Hs_c (0.90, p < 0.01, fig. 6D). However, the monthly maximum wave height obtained from individual storms produced a weak correlation (-0.41, p < 0.01) with NAO over the storm season. Similar results were obtained for the total energy of the storm waves for each month where the correlation coefficient with NAO was -0.50 (p < 0.01). The above results suggest that although negative NAO values increase the storminess over the study area, they do not control the magnitude of the wave height which is probably affected by mesoscale atmospheric patterns. Finally, the correlation between mean monthly residual water levels and NAO is high (-0.67, p < 0.01). These results are in agreement with previous studies in the area using the same reanalysis data for southern Europe (Marcos et al., 2009). The main mechanism that drives the residual sea level response to the NAO is both hydrostatic and non-hydrostatic (Woolf et al., 2003). The HIPOCAS data used include both effects since they are produced using a barotropic version of the HAMSOM model (Ratsimandresy et al., 2008).

Joint Probability

23For the estimation of the JP the residual sea level (RSL) and Hs_c were used in POT analysis. For each event the peak Hs_c and peak RSL were used in order to construct a contingency table. Tidal variations were not taken into account since in the Iberian Peninsula tidal–surge energy transfer is low (Ratsimandresy et al., 2008). The results of the JP are presented in figure 7.

Fig. 7 – Observed joint probability distribution of storm wave heights and residual mean sea level obtained from POT analysis of the corrected HIPOCAS data.
Fig. 7Distribution observée de la probabilité conjointe des hauteurs des vagues de tempête et du résidu du niveau moyen de la mer résultant de l'analyse POT des données HIPOCAS corrigées.

Fig. 7 – Observed joint probability distribution of storm wave heights and residual mean sea level obtained from POT analysis of the corrected HIPOCAS data. Fig. 7 – Distribution observée de la probabilité conjointe des hauteurs des vagues de tempête et du résidu du niveau moyen de la mer résultant de l'analyse POT des données HIPOCAS corrigées.

Greyscale indicates the number of events.
L’échelle de nuances de gris indique le nombre d'événements.

24A total of 369 events were identified using POT with a large proportion of them (50%) corresponding to low energy events (< 2 m Hs_c). For these events the RSL showed a large spread that is mainly concentrated in positive values between 0 and 15 cm. For the rest of the events a clear trend is obtained where larger wave heights are observed together with positive RSL, with values up to 30 cm for the extreme wave height events of 4-6 m that have a return period in the area of Cadiz between 3 and 4 years respectively. The above results show dependence between the RSL and the peak storm Hs_c, hence the JP cannot be calculated as a simple convolution between the Hs_c and RSL histograms.

25In more detail and in accordance with the results presented above between the storm and RSL correlations with the NAO, the JP analysis was undertaken separately for positive and negative NAO events. In general 166 storm events occurred during a positive NAO phase while 201 events occurred during a negative NAO phase (tab. 3). For NAO phases with an index larger/smaller than +/- 1 and +/- 1.5 it can be seen that the negative NAO phases show almost twice the events than the positive ones. This difference is not present in extreme NAO phases ( -2 > NAO > +2 ) probably due to the small number of events (tab. 3).

Tab. 3 – Number of storm events identified for different positive and negative NAO thresholds.
Tab. 3Nombre de tempêtes identifié pour différents seuils de la NAO positifs et négatifs.

 

0

0,5

1

1,5

2

NAO +

166

105

52

20

6

NAO -

201

149

89

44

6

Ratio (+/-)

0,83

0,7

0,58

0,45

1

Total

367

254

141

64

12

26The JP analysis for positive and negative NAO events with index larger/smaller than ± 1.5 is presented in figure 8, where it can be seen that during strong positive and negative phases of NAO the JP of the wave–surge follows a different pattern.

Fig. 8 – Observed joint probability distribution of storm wave heights and residual mean sea level obtained from POT analysis of the corrected HIPOCAS data for (A) storm events during NAO > +1.5 and (B) storm events during NAO < -1.5.
Fig 8Distribution observée de la probabilité conjointe des hauteurs des vagues de tempête et du résidu du niveau moyen de la mer résultant de l'analyse POT des données HIPOCAS corrigées pour (A) les tempêtes au cours d'une NAO > +1.5 et (B) les tempêtes au cours d'une NAO < -1.5.

Fig. 8 – Observed joint probability distribution of storm wave heights and residual mean sea level obtained from POT analysis of the corrected HIPOCAS data for (A) storm events during NAO > +1.5 and (B) storm events during NAO < -1.5. Fig 8 – Distribution observée de la probabilité conjointe des hauteurs des vagues de tempête et du résidu du niveau moyen de la mer résultant de l'analyse POT des données HIPOCAS corrigées pour (A) les tempêtes au cours d'une NAO > +1.5 et (B) les tempêtes au cours d'une NAO < -1.5.

Grayscale indicates the probability of occurrence.
L’échelle de nuances de gris indique la probabilité d’occurrence.

27Positive NAO events (fig. 8A) are concentrated in weak storm events (Hs_c < 2.5 m) with mainly small RSL. These events most probably correspond to locally generated storms. On the other hand, for the negative NAO events (fig. 8B) the same pattern that was observed in the full analysis with a positive trend between storm wave heights and RSL is repeated. This case is composed of both weak and strong storm events with large return periods.

Conclusions

28Re-analysis wave and sea level data for a period of 44 years (HIPOCAS data) were used to investigate the connection between large scale atmospheric circulation (NAO, EA) and the wave climate and sea level in the area of Cadiz. Significant improvement of the wave data was obtained after applying correction functions derived from the coastal wave buoy in the area of Cadiz. In general, the HIPOCAS data correctly represented the directional storm climate in the Gulf of Cadiz and mainly the one coming from the Atlantic. The locally accelerated easterly winds in the area of the Strait of Gibraltar are not well represented due to the relative large scale of the re-analysis data. However, these events are not affecting storm-related hazards along the coastline of Cadiz, as they are mostly related to high atmospheric pressure situations; furthermore, shoreline orientation and short fetch determine a negligible impact of easterly waves along the coast of Cadiz.

29In terms of wave activity two seasons can be distinguished: the storm and the calm season. The former extends from November to March and shows higher mean monthly significant wave height and distinct period and wave direction than the calm season. Based on these results further analysis was undertaken following the above seasonal pattern and not the atmospheric season convention. NAO presented negative correlations with the monthly parameters of the storm season. When the mean wave climatology was subtracted from the data this correlation was extended to the entire year (anomalies). Positive correlations were obtained with the EA pattern that probably represents the zonal extension of the storm tracks over the study area during positive EA phases. Better correlations were identified for the total storm hours (Storm Index) and the residual mean sea level but not with the maximum wave height. The above results suggest that although negative NAO values increase the storminess over the study area they do not control the magnitude of the wave height, which is probably affected by mesoscale atmospheric patterns.

30Joint probability analyses showed dependence between storm conditions and positive residual mean sea level on the basis of 367 events. This dependence is more pronounced over storm events with large return periods (Hs_c > 3 m). Study of storm events over distinct NAO phases showed a progressive abundance of storm events during negative NAO phases. At extreme negative NAO phases the coexistence of large RSL and large storm events are present. This is not the case in positive NAO phases, where small storm events are present with disperse RSL response. In terms of coastal hazards and risk the coexistence of storm events and high RSL can potentially increase the vulnerability of the coastal areas to erosion and/or flooding episodes.

Haut de page

Bibliographie

Allan J.C., Komar P.D. (2000) – Are ocean wave heights increasing in the eastern North Pacific? Eos, Transactions, American Geophysical Union, 47, 561-567.

Bacon S., Carter D.J.T. (1991) – Wave Climate Changes in the North-Atlantic and North-Sea. International Journal of Climatology 11, 5, 545-558.

Barnston A.G., Livezey R.E. (1987) – Classification, Seasonality and Persistence of Low-Frequency Atmospheric Circulation Patterns. Monthly Weather Review 115, 1083-1126.

Benavente J., Gracia F.J., López-Aguayo F. (2000) – Empirical model of morphodynamic beachface behaviour for low-energy mesotidal environments. Marine Geology 167, 375–390.

Bird E.C.F. (1985)Coastal changes. Wiley & Sons, New York, 219 p.

Carter D.J.T., Draper L. (1988) – Has the north-east Atlantic become rougher? Nature 332, 6164, 494-494.

Cooper J.A.G., Jackson D.W.T., Navas F., McKenna J., Malvarez G. (2004) – Identifying storm impacts on an embayed, high-energy coastline: examples from western Ireland. Marine Geology 210, 1-4, 261-280.

Criado-Aldeanueva F., García Lafuente J., Navarro G., Ruiz J. (2009) – Seasonal and interannual variability of the surface circulation in the eastern Gulf of Cadiz (SW Iberia). Journal of Geophysical Research 114, C01011 doi:10.1029/2008JC005069.

Del Río L., Plomaritis T.A., Benavente J., Valladares M., Ribera P. (2012) – Establishing storm thresholds for the Spanish Gulf of Cádiz coast, Geomorphology. 143-144(0), 13-23, doi:10.1016/j.geomorph.2011.04.048

Dodet G., Bertin X., Taborda R. (2010) – Wave climate variability in the North-East Atlantic Ocean over the last six decades. Ocean Modelling 31, 3-4, 120-131.

Dorman C.E., Beardsley R.C., Limeburner R. (1995) – Winds in the strait of gibraltar. Quarterly Journal of the Royal Meteorological Society 121, 1903-1921.

Dupuis H., Michel D., Sottolichio A. (2006) – Wave climate evolution in the Bay of Biscay over two decades. Journal of Marine Systems 63, 3-4, 105-114.

Efthymiadis D., Hernandez F., Le Traon P.-Y. (2002) – Large-scale sea-level variations and associated atmospheric forcing in the subtropical north-east Atlantic Ocean. In Deep Sea Research Part II: Topical Studies.Oceanography 49, 3957-3981.

Fenster M.S., Dolan R., Morton, R.A. (2001) – Coastal storms and shoreline change: signal or noise? Journal of Coastal Research 17, 3, 714-720.

Garcia T., Ferreira Ó., Matias A., Dias J.A. (2005) – Coastal Hazards Representation for Praia do Barril (Algarve, Portugal). In Proceedings of the 2nd Meeting in Marine Sciences, Valencia – Spain, ISSN 0749-0208). Journal of Coastal Research SI 49, 28 – 33.

García-Lafuente J., Ruiz J. (2007) – The Gulf of Cadiz pelagic ecosystem: A review. Progress in Oceanography 74, 228-251.

Guedes Soares C., Carretero Albiach J.C., Weisse R., Alvarez-Fanjul E. (2002) – A 40 years hindcast of wind, sea level and waves in European waters. In 21st International Conference on Offshore Mechanics and Arctic Engineering Olso, Am. Soc. of Mech. Eng., 669–675.

Hegge B., Eliot I., Hsu J. (1996) – Sheltered sandy beaches of southwestern Australia. Journal of Coastal Research 12, 748-760.

Jackson D.W.T., Cooper J.A.G., Del Río L. (2005) – Geological control of beach morphodynamic state. Marine Geology 216, 297-314.

Kalnay E., Kanamitsu M., Kistler R., Collins W., Deaven D., Gandin L., Iredell M., Saha S., White G., Woollen J., Zhu Y., Chelliah M., Ebisuzaki W., Higgins W., Janowiak J., Mo K.C., Ropelewski C., Wang J., Leetmaa A., Reynolds R., Roy Jenne R., Joseph D. (1996) – The NCEP/NCAR 40-year reanalysis project. Bulletin of the American Meteorological Society 77, 437-471.

Kamphuis J.W. (2000)Introduction to Coastal Engineering and Management. Advanced Series on Ocean Engineering Vol. 16. World Scientific Publishing, Singapore, 436 p.

Keim B.D., Muller R.A., Stone G.W. (2004) – Spatial and temporal variability of coastal storms in the North Atlantic Basin. Marine Geology 210, 7-15.

Kundu P.K. (1976) – Ekman Veering Observed near the Ocean Bottom. Journal of Physical Oceanography 6, 2, 238-242.

List J.H., Farris A.S., Sullivan C. (2006) – Reversing storm hotspots on sandy beaches: Spatial and temporal characteristics. Marine Geology 226, 3-4, 261-279.

Marcos M., Tsimplis M.N., Shaw A.G.P. (2009) – Sea level extremes in southern Europe. Journal of Geophysical Research-Oceans 114, C01007,doi:10.1029/2008jc004912.

Medina J.M. (1991) – La flecha de El Rompido en la dinámica litoral de la costa onubense. Ingeniería Civil 80, 105-110.

Menéndez M., Méndez F.J., Losada I.J, Medina R., Abascal A.J. (2004) – Variaciones del régimen extremal del clima marítimo en el litoral español en el periodo 1958-2001. In García Codron J.C., Diego Liaño C., Fdez. De Arróyabe Hernáez P., Garmendia Pedraja C., Rasilla Álvarez D. (Eds.) El Clima entre el Mar y la Montaña. Asociación Española de Climatología y Universidad de Cantabria, Santander, Serie A, 4, 73-84.

Morris B.D., Davidson M.A., Huntley D.A. (2001) – Measurements of the response of a coastal inlet using video monitoring techniques. Marine Geology 175, 1-4, 251-272.

Morton R.A., Gibeaut J.C., Paine J.G. (1995) – Meso-scale transfer of sand during and after storms: implications for prediction of shoreline movement. Marine Geology 126, 161–179.

O'Connor M.C., Cooper J.A.G., Jackson D.W.T. (2011) – Decadal Behavior of Tidal Inlet-Associated Beach Systems, Northwest Ireland, in Relation to Climate Forcing. Journal of Sedimentary Research 8, 1, 38-51.

Ratsimandresy A.W., Sotillo M.G., Carretero Albiach J.C., Alvarez-Fanjul E., Hajji H. (2008) – A 44-year high-resolution ocean and atmospheric hindcast for the Mediterranean Basin developed within the HIPOCAS Project, Coastal Engineering, 55, 827-842. doi:10.1016/j.coastaleng.2008.02.025.

Ribera P., Gallego D., Pena-Ortiz C., Del Río L., Plomaritis T.A., Benavente J. (2011) – Reconstruction of Atlantic historical winter coastal storms in the Spanish coasts of the Gulf of Cadiz, 1929-2005. Natural Hazards and Earth System Science 11, 6, 1715-1722.

Rusu L., Pilar P., Guedes Soares C. (2008) – Hindcast of the wave conditions along the west Iberian coast. Coastal Engineering 55, 906-919.

Sebastiao P., Guedes Soares C., Alvarez E. (2008) – 44 years hindcast of sea level in the Atlantic Coast of Europe. Coastal Engineering 55, 11, 843-848.

Short A.D., Tremblanis A.C., Turner I.L. (2000) – Beach oscillation, rotation and the Southern Oscillation, Narrabeen Beach, Australia. 27th International Conference on Coastal Engineering. ASCE, Sydney, 2439-2452.

Tanner W.F. (1960) – Florida coastal classification. Trans. Gulf Coast Assoc. Geol. Societies 10, 259-266.

Tsimplis M.N., Shaw A.G.P. (2008) – The forcing of mean sea level variability around Europe. Global and Planetary Change 63, 196-202.

Tsimplis M.N., Shaw A.G.P., Flather R.A., Woolf D.K (2009) – The influence of the North Atlantic Oscillation on the sea-level around the northern European coasts reconsidered: the thermosteric effects. Phil. Trans. R. Soc. A 2006, 364, 845-856. doi:10.1098/rsta.2006.1740.

Wettstein J.J., Wallace J.M. (2010) – Observed Patterns of Month-to-Month Storm-Track Variability and Their Relationship to the Background Flow. Journal of the Atmospheric Sciences 67, 1420-1437.

Woolf D.K., Challenor P.G., Cotton P.D. (2002) – Variability and predictability of the North Atlantic wave climate. Journal of Geophysical Research 107, C10, 3145, 9-1 – 9-14, doi:10.1029/2001JC001124.

Woolf D.K., Shaw A.G.P., Tsimplis M.N. (2003) – The influence of the North Atlantic Oscillation on sea-level variability in the North Atlantic region. The Global Atmosphere and Ocean System 9, 145-167.

Haut de page

Annexe

Version française abrégée

Les événements de tempête sont considérés comme étant la cause principale de l'érosion de la ligne de rivage dans beaucoup de pays (Fenster et al., 2001) et jouent ainsi un rôle majeur dans l'évolution des côtes sur le long terme (décennal à centennal). Des tendances spécifiques concernant les hauteurs de vague sont observées dans l'Atlantique Nord (Carter and Draper, 1988; Bacon and Carter, 1991; Dupuis et al., 2006). Par ailleurs, Bacon and Carter (1993) ont observé une corrélation entre le gradient de pression atmosphérique de la partie méridionale de l'Atlantique Nord et la hauteur de vague. Woolf et al. (2002) ont établi une relation entre des anomalies de la hauteur de vague et des modèles de pression atmosphérique à petite échelle dans l'Atlantique du Nord-Est, sur la base d'une altimétrie satellitaire. Plus précisément, ils attribuent une partie de la variabilité à l'oscillation nord-atlantique (NAO) et une autre partie au modèle de l'Atlantique Est (EA). Le travail présenté concerne la région côtière de Cadix et analyse l'occurrence de phénomènes combinant l'action des vagues et du niveau moyen de la mer ainsi que leur contribution à l'aggravation croissante des risques côtiers.

Le golfe de Cadix est le sous-bassin qui relie l'océan Atlantique avec la mer Méditerranée, à travers le détroit de Gibraltar. À l'échelle locale, l'orientation de la ligne de côte et les caractéristiques physiographiques favorisent un effet d'abri des vents de composante Nord et Est. Le vent dominant et les champs de vagues, composés de vagues de mer et de houles, sont de direction OSO, engendrant un courant de dérive littoral dominant vers l'E et le SE. Dans ce travail, les données du modèle de simulation de Puerto del Estado (Spain) pour le littoral de Cadix ont été utilisées (fig. 1). La base de données (données HIPOCAS ci-dessous) est issue d'une nouvelle analyse de 44 ans de variables météorologiques, vagues et niveau de la mer, entre janvier 1958 et décembre 2001(Guedes-Soares et al., 2002).

Les données de la bouée côtière de Puertos del Estado ont été utilisées pour la validation (fig. 1). Dans le but d'obtenir une série de données statistiquement indépendantes des conditions de tempête, une analyse de pics au-dessus d’un seuil (POT) a été conduite sur la période, avec une utilisation simultanée de données de modèle et d'observation (Kamphuis, 2000). L'analyse a produit une correction basée sur le pic des valeurs de chaque tempête et a été ainsi appliquée à l'ensemble du jeu de données. Bien que le chevauchement des données portant sur la direction des vagues entre la bouée de Cadix et les données de HIPOCAS n'existe que sur l'année 2001, la période de temps est assez longue pour couvrir à la fois une saison calme et une saison orageuse. La concordance des deux séries temporelles a été testée par le vecteur de corrélation Kundu (1976) entre les deux ensembles de données. Cette méthode de corrélation produit un coefficient entre les hauteurs des vagues directionnelles des deux séries temporelles et l'angle principal par lequel la première série devrait être soumise à rotation dans le sens antihoraire pour correspondre à la direction de la deuxième série.

La série chronologique corrigée a été utilisée pour calculer les valeurs mensuelles, saisonnières et annuelles des hauteurs de vagues, directions et niveau de la mer afin d'identifier la climatologie des vagues dans l'aire d'étude. En outre, les caractéristiques des tempêtes ont été calculées sur une base mensuelle afin d'être comparées avec les indices atmosphériques à grande échelle. Ces caractéristiques étaient le nombre total d'heures lorsque le seuil de la tempête a été dépassé (Indice de tempête), l’énergie totale de la tempête par mois et le nombre de tempêtes par mois. Des corrélations des valeurs décrites ci-dessus avec les indices atmosphériques ont été effectuées. Pour le calcul de probabilité jointe (JP), une analyse de POT a été effectuée afin de construire un tableau de contingence avec les événements de tempête et les surcotes associées.

Le cycle annuel moyen pour la hauteur de vague significative corrigée (Hs_c) et les directions de vague associées pour la côte de la ville de Cadix sont présentés sur la figure 5. Les résultats (tab. 1) montrent une faible corrélation entre le Hs_c annuel et la NAO qui augmente à -0,58 pour les mois de la saison de tempête (fig. 6A). Les corrélations entre les anomalies mensuelles moyennes et la NAO montrent une augmentation significative (p < 0,02) seulement pour les valeurs annuelles, laissant penser que la NAO a un effet sur la variabilité des vagues qui prévaut dans des conditions calmes. Pour le cas de l'EA, les paramètres ci-dessus montrent des corrélations faibles mais significatives (tab. 2) dont la plus prononcée est celle de la hauteur significative des vagues pendant les mois de tempête. Dans tous les cas d'EA, les coefficients de corrélation sont plus petits que ceux avec la NAO. Des corrélations légèrement plus élevées (-0,67, p < 0,01) ont été obtenues entre l'indice de tempête et la NAO pour la saison des tempêtes (fig. 6B), et des corrélations très élevées ont été obtenues entre l'indice de tempête et Hs_c (0,90, p < 0,01, fig. 6D). Cependant, le maximum mensuel de la hauteur de vague obtenu lors des tempêtes individuelles donne une faible corrélation (-0,41, p < 0,01) avec la NAO pendant la saison des tempêtes. Des résultats similaires ont été obtenus pour l'énergie totale des vagues de tempête pour chaque mois où le coefficient de corrélation avec la NAO était de -0,50 (p < 0,01).

Pour chaque événement, les pics Hs_c et RSL ont été utilisés pour construire un tableau de contingence. Pour ces événements, le RSL a montré une répartition large qui est principalement concentrée dans des valeurs positives entre 0 et 15 cm. Pour les événements restants, une tendance claire apparait quand de grandes hauteurs de vagues sont observées de concert avec un RSL positif, avec des valeurs supérieures à 30 cm pour les événements extrêmes de hauteur de vague de 4-6 m qui ont une période de retour dans la région de Cadix comprise respectivement entre 3 et 4 ans. En accord avec les résultats présentés ci-dessus, entre la tempête et les corrélations de RSL avec la NAO, l'analyse JP a été effectuée séparément pour les événements de NAO positive et négative. Pour les phases de NAO avec un index plus grand/plus petit que +/- 1 et +/- 1,5, on peut voir que les phases de NAO négative montrent presque deux fois plus d'événements que les phases de NAO positive.

Haut de page

Table des illustrations

Titre Fig. 1 – Bathymetric map of the study area showing the HIPOCAS data grid cell and the location of the coastal buoy of Cadiz managed by Puertos del Estado. Fig 1Carte bathymétrique de la zone d'étude indiquant la cellule de la grille des données HIPOCAS et l'emplacement de la bouée côtière de Cadix géré par Puertos del Estado.
URL http://geomorphologie.revues.org/docannexe/image/10728/img-1.png
Fichier image/png, 82k
Titre Fig. 2 – Comparison between modeled (HIPOCAS), measured (buoy), and corrected data using POT analysis for significant wave height (Hs). Fig. 2Comparaison entre des données modélisées (HIPOCAS), mesurées (bouée) et corrigées en utilisant l'analyse POT pour la hauteur significative des vagues (Hsig).
URL http://geomorphologie.revues.org/docannexe/image/10728/img-2.png
Fichier image/png, 139k
Titre Fig. 3 – Comparison of the difference between mean wave direction for the corrected (HIPOCAS) and measured (Buoy) data for 2001. Fig. 3Comparaison de la différence de la direction moyenne des vagues pour les données corrigées (HIPOCAS) et mesurées (bouée) pour 2001.
Légende Grayscale presents the data density. L’échelle des niveaux de gris montre la densité des données.
URL http://geomorphologie.revues.org/docannexe/image/10728/img-3.png
Fichier image/png, 123k
Titre Fig. 4 – Comparison between modeled (HIPOCAS), measured (buoy), and corrected data using POT analysis for (A) the zonal (East-West) and (B) the meridional components of the significant wave height (Hsig). Fig. 4Comparaison entre des données modélisées (HIPOCAS), mesurées (bouée) et corrigées en utilisant l'analyse POT pour (A) la composante de la zone (Est-Ouest) et (B) la composante méridienne de la hauteur significative des vagues (Hsig).
URL http://geomorphologie.revues.org/docannexe/image/10728/img-4.png
Fichier image/png, 223k
Titre Fig. 5 – Average seasonal cycle for the entire reanalysis period of (A) Significant wave height and (B) Mean wave direction. Fig. 5Cycle saisonnier moyenne pour la période entière de réanalyse de (A) la hauteur significative des vagues et (B) la direction moyenne des vagues.
Légende Note: correction values for the Gulf of Cadiz described in section "Model validation" were applied in both cases. Note : les valeurs de correction pour le Golfe de Cadix décrites dans la section "Validation du modèle" ont été appliquées dans les deux cas.
URL http://geomorphologie.revues.org/docannexe/image/10728/img-5.png
Fichier image/png, 41k
Titre Fig. 6 – Correlations between NAO index and (A) mean monthly significant wave height, (B) Storm Index (number of hours that the wave height exceeds the storm threshold) and (C) residual monthly mean sea level, for the months of the storm season (November - March). (D): Correlation between the Storm Index and the mean monthly significant wave height. Fig. 6Corrélations entre l'indice NAO et (A) la hauteur moyenne mensuelle de vague significative, (B) le Storm Index (nombre d'heures que la hauteur des vagues dépasse le seuil de tempête) et (C) le résidu du niveau moyen de la mer mensuel, pour les mois de la saison des tempêtes (Novembre - Mars). (D) Corrélation entre le Storm Index et la hauteur significative des vagues moyenne mensuelle.
Légende Grayscale represents the data density.L’échelle de nuances de gris représente la densité des données.
URL http://geomorphologie.revues.org/docannexe/image/10728/img-6.png
Fichier image/png, 261k
Titre Fig. 7 – Observed joint probability distribution of storm wave heights and residual mean sea level obtained from POT analysis of the corrected HIPOCAS data. Fig. 7Distribution observée de la probabilité conjointe des hauteurs des vagues de tempête et du résidu du niveau moyen de la mer résultant de l'analyse POT des données HIPOCAS corrigées.
Légende Greyscale indicates the number of events. L’échelle de nuances de gris indique le nombre d'événements.
URL http://geomorphologie.revues.org/docannexe/image/10728/img-7.png
Fichier image/png, 71k
Titre Fig. 8 – Observed joint probability distribution of storm wave heights and residual mean sea level obtained from POT analysis of the corrected HIPOCAS data for (A) storm events during NAO > +1.5 and (B) storm events during NAO < -1.5. Fig 8Distribution observée de la probabilité conjointe des hauteurs des vagues de tempête et du résidu du niveau moyen de la mer résultant de l'analyse POT des données HIPOCAS corrigées pour (A) les tempêtes au cours d'une NAO > +1.5 et (B) les tempêtes au cours d'une NAO < -1.5.
Légende Grayscale indicates the probability of occurrence. L’échelle de nuances de gris indique la probabilité d’occurrence.
URL http://geomorphologie.revues.org/docannexe/image/10728/img-8.png
Fichier image/png, 90k
Haut de page

Pour citer cet article

Référence papier

Theocharis A. Plomaritis, Javier Benavente, Irene Laiz et Laura del Río, « Storminess interannual variability and coastal hazards over the south-western Spanish coast: links to large scale atmospheric forcing », Géomorphologie : relief, processus, environnement, vol. 20 - n° 3 | 2014, 275-286.

Référence électronique

Theocharis A. Plomaritis, Javier Benavente, Irene Laiz et Laura del Río, « Storminess interannual variability and coastal hazards over the south-western Spanish coast: links to large scale atmospheric forcing », Géomorphologie : relief, processus, environnement [En ligne], vol. 20 - n° 3 | 2014, mis en ligne le 01 janvier 2016, consulté le 20 octobre 2017. URL : http://geomorphologie.revues.org/10728 ; DOI : 10.4000/geomorphologie.10728

Haut de page

Auteurs

Theocharis A. Plomaritis

Centro de Investigação Marinha e Ambiental (CIMA) – Faculdade de Ciencias y Tecnologia – Campus Universitário de Gambelas – Universidade do Algarve – 8005-139 – Faro – Portugal (tplomaritis@ualg.pt).

Javier Benavente

Centro de Investigação Marinha e Ambiental (CIMA) – Faculdade de Ciencias y Tecnologia – Campus Universitário de Gambelas – Universidade do Algarve – 8005-139 – Faro – Portugal.

Irene Laiz

Department of Applied Physics – Faculty of Marine and Environmental Science – University of Cadiz – Poligono Rio San Pedro S/N – 11510, Puerto Real, Cadiz – Spain.

Laura del Río

Centro de Investigação Marinha e Ambiental (CIMA) – Faculdade de Ciencias y Tecnologia – Campus Universitário de Gambelas – Universidade do Algarve – 8005-139 – Faro – Portugal.

Haut de page

Droits d’auteur

© Groupe français de géomorphologie

Haut de page
  • Logo CNRS - Institut des sciences humaines et sociales
  • Logo Groupe français de géomorphologie
  • Revues.org