Shoreline Change and Climatic Variability along the Moulay Bousselham Coast (Moroccan Atlantic)
ABSTRACT
Belrhaba, T.; Hakkou, M.; Rey, T.; Aangri, A.; Krien, Y.; Elmostafa, Z.; Leone, F., and Benmohammadi, A., 2024. Shoreline change and climatic variability along the Moulay Bousselham coast (Moroccan Atlantic).
Sandy coasts in Morocco are increasingly threatened by erosion. It is thus crucial to investigate the extent and causes of shoreline change to propose mitigation strategies to stakeholders. This study investigates the long-term shoreline dynamics of Moulay Bousselham (1949–2016) and its relationship with climate variability indices, namely the North Atlantic Oscillation, East Atlantic Oscillation, and West Europe Pressure Anomaly (WEPA). The analysis, which employs geospatial techniques and the Digital Shoreline Analysis System, reveals significant temporal and spatial variability in shoreline changes characterized by phases of erosion and accretion. The study identifies distinct periods of retreat and growth, highlighting the complex nature of coastal dynamics. The correlation analysis between wave parameters and climatic indices emphasizes the influential role of the WEPA index in controlling winter oceanic phenomena. The winter of 2013/14, marked by the highest WEPA, corresponds to the most energetic conditions in at least 70 years, underscoring the index’s significance in understanding local climate variability. While acknowledging methodological challenges and uncertainties inherent in shoreline displacement calculations, the study establishes a noteworthy correlation between the WEPA index and shoreline dynamics. This research contributes valuable insights into the intricate interactions between climate variability and coastal evolution, emphasizing the need for refined methodologies and a comprehensive understanding of factors influencing shoreline changes.
INTRODUCTION
Coastal zones are currently facing intensified natural and anthropogenic disturbances, with more than 80% of global beaches undergoing coastal erosion at rates ranging from 1 m/y to 30 m/y (Appeaning Addo, Walkden, and Mills, 2008). A robust and precise long-term delineation of the coastline is essential for efficiently managing coastal areas and making informed political decisions. However, shoreline changes manifest over various temporal scales, with interannual and decadal timeframes holding particular significance for coastal scientists because of their close association with global climate variability modes (Barnard et al., 2015).
At short to medium time scales, variations in wave patterns play a central role in shoreline change. Beaches react to intense storms, resulting in significant morphological variations within localized areas in a matter of hours (Dissanayake et al., 2015; Harley et al., 2017; Scott et al., 2015; Short and Trembanis, 2004). This response extends to seasonal fluctuations in wave conditions and changes in wave height over interannual to decadal periods (Castelle et al., 2018; Masselink and Pattiaratchi, 2001). Propelled by wave action, the processes of sediment transport (both transverse and longitudinal) drive changes in beach morphology and shoreline position. Although transverse processes typically dominate seasonal and annual coastal changes, longitudinal processes tend to exert greater influence on coastal dynamics over decadal times (Vitousek et al., 2017).
Variations over time in wave forcing are governed by large-scale weather patterns and their fluctuations. Within the Pacific Ocean basin, the primary driver of interannual climate variability is the El Niño-Southern Oscillation (ENSO); it is linked to the prevailing wave conditions (Boucharel et al., 2021). In the North Atlantic region, the North Atlantic Oscillation (NAO) emerges as the predominant mode of atmospheric variability, exhibiting a strong correlation with wave conditions. Positive phases of the NAO are linked to more energetic winter conditions in the NE Atlantic (Dodet, Bertin, and Taborda, 2010). Conversely, negative phases of the NAO coincide with energetic wave conditions along the southern coast of Spain and the Moroccan coast (Plomaritis et al., 2015). These prevailing atmospheric circulation patterns can be quantified using indices, which are typically derived from measurements of spatial variability in sea-surface temperature or atmospheric pressure, and can be statistically correlated with various wave parameters.
In the context of the NE Atlantic, Castelle et al. (2017) introduced a novel atmospheric index known as the West Europe Pressure Anomaly (WEPA). This index is derived from the sea-level pressure (SLP) gradient observed between Valentia (Ireland) and Santa Cruz de Tenerife (Canary Islands, located off the southern coast of Morocco). The positive phase of WEPA signifies an intensified and southward-shifted SLP contrast between the Icelandic low and the Azores high, prompting significant storms that channel high-energy waves toward Western Europe, specifically south of 52° N. The WEPA index has demonstrated a notably robust correlation with winter wave conditions spanning the entire European and North African Atlantic coast, extending from Ireland to Spain and along the Moroccan coast, significantly surpassing the NAO as a predictive indicator of wave height during the winter season.
Recent research indicates that winter wave conditions near the United Kingdom and Ireland are associated with climate variability modes in the North Atlantic, notably the North Atlantic Oscillation (NAO), West Europe Pressure Anomaly (WEPA), and the East Atlantic Oscillation (EA). These indices exhibit robust correlations with both total and directional winter wave power, as highlighted by Scott et al. (2021). On a regional scale, Wiggins et al. (2020) explored the characteristics of bidirectional wave climates along the SE coast of England. Their findings revealed a strong correlation between wave power and WEPA+ and a corresponding correlation between wave power and NAO−.
Given the strong connection between wave-height variability and coastal dynamics, it can be hypothesized that climatic indices are intertwined with coastal changes. Almar et al. (2023) expanded this hypothesis on a global scale by constructing a comprehensive conceptual model based on satellite-derived shoreline (SDS) positions spanning from 1993 to 2019 and on various reanalysis products. They suggest that global interannual shoreline changes are predominantly directed by distinct ENSO regimes and their intricate inter-basin teleconnections, although reported correlations for the NE Atlantic exhibit relative weakness. Despite this, indisputable links between powerful El Niño events and extreme coastal erosion along the west coast of the United States have been established (Barnard et al., 2011; Barnard et al., 2017; Young et al., 2018).
In a recent study, Vos et al. (2023) used 38 years of Landsat images to delineate shoreline variability across the Pacific and identified consistent patterns of erosion and beach accumulation influenced by ENSO, albeit with regional variations. Parallel works have aimed to reveal associations between climatic indices and coastal changes on the Atlantic coast of Europe. Based on a decade-long analysis of video-monitoring data from a sandy beach in SW England, Masselink et al. (2014) indicated that a correlation between beach state and bar morphology with the NAO occurred. Castelle et al. (2022) examined the temporal and spatial evolution of a 269 km high-energy, meso-macrotidal sandy coast in SW France from 1984 to 2020 using SDS data. They found a strong correlation between interannual variability in shoreline positions and winter WEPA, surpassing conventional teleconnection pattern indices such as NAO. Notably, the winter of 2013/14, characterized by the highest winter WEPA value since 1948, was marked by exceptional wave conditions, leading to substantial erosion along the entire Atlantic coast of Europe (Masselink et al., 2016) and Morocco (Hakkou et al., 2019). The tantalizing prospect of identifying causal links between atmospheric indices and shoreline changes holds the potential to streamline future shoreline dynamics modeling efforts, eliminating the need for explicit wave modelling (Robinet et al., 2016), especially in the pursuit of seasonal forecasts of coastal changes (Scott et al., 2021).
For effective coastal zone management, it is imperative to develop a thorough understanding of shoreline change trends (Anfuso, Gracia, and Battocletti, 2013; Appeaning Addo, Walkden, and Mills, 2008; Mills et al., 2005; Ozturk and Sesli, 2015; Rangel-Buitrago, Anfuso, and Williams, 2015). Over the past seven decades, shoreline-mapping science has undergone major advancements because of various technological progressions (Appeaning Addo, Walkden, and Mills, 2008). However, shoreline mapping with the least percentage of error remains uncertain because of its variable and dynamic nature, heavily influenced by various short-term effects (e.g., tides) and long-term effects (e.g., sea-level rise). Thus, accurately calculating historical shoreline change rates remains a challenging task. Globally, the qualitative and quantitative analysis of spatiotemporal variations in the shoreline has been addressed by several studies, including Appeaning Addo, Jayson-Quashigah, and Kufogbe (2012); Kabuth, Kroon, and Pedersen (2014); Murali et al. (2015); Nandi et al. (2016); Nassar et al. (2019); and Tran Thi et al. (2014). Locally, the erosion hazard combined with human pressure has motivated several studies on regional sediment dynamics and coastline evolution on the Moroccan Atlantic coast (Aangri et al., 2022; Ahizoun et al., 2009; El Habti et al., 2022; Ennouali et al., 2023; Hakkou et al., 2018; Moussaid et al., 2015; Tahri et al., 2017).
This study focuses on analyzing the shoreline kinematics of Moulay Bousselham over a period of 67 years (1949–2016) using geospatial techniques, GIS, and automated calculations (Digital Shoreline Analysis System [DSAS]). The main objectives of this article are (1) to map and quantify shoreline erosion and accretion rates using various statistical approaches integrated into DSAS, including the Linear Regression Rate, the end-point rate (EPR), and the net shoreline movement (NSM) along the Moulay Bousselham coast, and (2) to investigate the role of climate variability (expressed by the NAO, EAO, and WEPA) on significant wave heights (Hs) and their energies and also the mobility of the Moulay Bousselham shoreline and the positive and negative phases of the three climate indices. This study is of great importance for the sustainable development and environmental protection of the Moulay Bousselham site, designated as a RAMSAR site in 1980.
Study Area
The Moulay Bousselham coast, situated approximately 80 km north of the city of Kenitra on the Moroccan Atlantic, is characterized by a straight sandy shoreline bordered by dunes, interrupted by the Merja Zerga lagoon inlet (Figure 1). Covering an area of 35 km2, the Merja Zerga lagoon is recognized as one of the most significant biological reserves in the northern Moroccan Atlantic and has held Ramsar Convention status since 1980.
Citation: Journal of Coastal Research 40, 5; 10.2112/JCOASTRES-D-23-00082.1
The foreshore of the beach measures 100–200 m in width, narrowing to 50–100 m as it approaches the inlet. A wide dune, ranging from 10 to over 30 m in height, crowns the beach. This dune disappears to the north near the inlet, giving way to a consolidated old dune that overlooks the cliff beach (Carruesco, 1989; Mhamdi Alaoui, 2009). According to the general census of Morocco, the city of Moulay Bousselham is populated by 7335 inhabitants (https://www.hcp.ma/region-kenitra/attachment/1519457/); it is situated on the fossil dune. Residences line the coast, extending for approximately 2 km, constituting the only urban development in this pristine landscape; however, ongoing projects are currently under development (Figure 2).
Citation: Journal of Coastal Research 40, 5; 10.2112/JCOASTRES-D-23-00082.1
Because of lack of data before 1958, the analysis focused on wave climatology for the period from 1958 to 2016, using data from the SIMAR database point 1054040, is located at 35.00° N, 6.50° W (Figure 1a). The wave rose diagram illustrates prevailing NW waves, with the most significant wave height ranging between 0.5 and 2 m with an average height of 1.26 m. During severe storms, the significant wave height can reach 7 to 9 m (Benmohammadi et al., 2007). The predominant wave directions generate a north-to-south longshore drift current, with a higher flow in the south (2.105–8.105 m³/y) than in the north (3.104–3.105 m³/y; Mhamdi Alaoui, 2009).
Additionally, the tide exhibits mesosemidiurnal patterns with a tidal range of 0.6 to 3.8 m (Carruesco, 1989; Mhamdi Alaoui, 2009) Tide-induced currents on the continental shelf range from 0.2 to 0.3 m/s (Charrouf, 1989; Jaaidi, 1981), generally flowing northward during ebb and southward during flow. In the nearshore zone of Moulay Bousselham, coastal and tide-induced currents are negligible compared with wave-induced currents, classifying the beaches as wave dominated. The Drader River and an artificial canal (Nador Canal) conclude their course in the Merja Zerga lagoon (Figure 2); however, no studies have calculated the sediment transfer between the lagoon and the ocean.
The study conducted on this coast focuses on a sandy section spanning 12 km on either side of the Moulay Bousselham lagoon inlet.
METHODS
Two steps were followed in this work: (1) analyzing and interpreting aerial photographs data to deduce the dynamics of the coastline and (2) analyzing the variability of wave height and wave-energy flux during winter and correlating it with NAOs, EAOs, and WEPA. Details of this approach are provided here.
Historical Shoreline Mapping and Change-Rate Computation
The schematic approach to the study of coastline evolution is illustrated in Figure 3.
Citation: Journal of Coastal Research 40, 5; 10.2112/JCOASTRES-D-23-00082.1
Technical conventions for determining the rate of historical shoreline change involve comparing the shoreline position over time. This includes field measurements of water-level position, with position shoreline traced from aerial photographs, satellite images, and topographic and bathymetric maps using several methods such as EPR (Fenster, Dolan, and Elder, 1993), average of rates (AOR), Linear Regression Rate, and jackknife (JK) (Dolan, Fenster, and Holme, 1991).
The historical evolution of the shoreline was assessed by delineating shoreline positions and analyzing changes in rates over the past 67 years. The reference line used for the extraction of shoreline positions is the high-water line (HWL), digitized from an assortment of aerial photographs and satellite images spanning from 1949 to 2016 (Table 1). This boundary is discerned by a shift in color or gray tone, a consequence of the disparity in the water content of the sand on either side of the HWL. However, it is critical to note that this line may not always be distinctly visible and is influenced by factors such as wind, wave, and tide conditions at the time of observation, as well as the quality of the aerial photographs used for extraction (Boak and Turner, 2005).
Georeferencing of aerial photographs, shoreline digitization, error estimation, and determination of shoreline change rates were performed using ArcGIS version 10.5 (ESRI, 2017), developed by ESRI using the Lambert Conformal Conic Zone 1 Datum Merchich Spheroid Clarke 1880 projection (EPSG code: 26191).
This research employed polynomial georeferencing, a technique commonly used for extensive sets of aerial photographs (Aangri et al., 2022; Hughes, McDowell, and Marcus, 2006; Kumar, Narayana, and Jayappa, 2010; Maanan et al., 2014; Moussaid et al., 2015). This method is compatible with most commercially available GIS software and is widely adopted for the registration of digitized aerial photos. Following the scanning of aerial photographs, the polynomial georeferencing process involved three main steps: (1) matching ground control points (with more than 50 selected on the scanned image and the base layer), (2) transforming the coordinates of ground control points on the scanned image from a generic raster dataset to a geographic projection and coordinate system, and (3) pixel resampling.
The application of polynomial georeferencing ensures a minimal spatial error in the geometric properties of remotely sensed images, particularly evident in the case of a linear shoreline (Hakkou et al., 2011). Owing to its inherent robustness in correcting the geometric aspects of remotely sensed data, this method was employed in the present study (Aangri et al., 2022; Avinash, Deepika, and Jayappa, 2013; El Habti et al., 2022; Ennouali et al., 2023; Hakkou et al., 2018; Jolivet, Gardel, and Anthony, 2019; Rocchini and Di Rita, 2005).
The rate of shoreline change was determined using the DSAS (DSAS v5; Himmelstoss et al., 2018). This program establishes a baseline, generates orthogonal transects along the coastline relative to this baseline, and computes rates of change using various methods such as Linear Regression Rate, EPR, and JK.
In the present study, a total of 704 transects were automatically generated by the DSAS program at regular 15 m intervals along the coast. These transects, each spanning 400 m in length, collectively cover the entire extent of the study coastline (Figure 4).
Citation: Journal of Coastal Research 40, 5; 10.2112/JCOASTRES-D-23-00082.1
Relative shoreline changes were estimated using the EPR method, NSM, and Linear Regression Rate for each set of data obtained for all transects. The EPR is calculated by dividing the distance separating two shorelines by the time elapsed between the dates of two shorelines, giving the rate of shoreline change (m/y). The NSM is the distance between two shorelines (oldest and the youngest shoreline for each transect), giving the variation in beach width (m) (Thieler et al., 2009). If the NSM is divided by the number of years between the two shoreline positions, the result is the EPR. (NSM and EPR are essentially the same thing.) The Linear Regression Rate is the most common statistical technique used to express shoreline movement and estimate rates of change (Crowell, Douglas, and Leatherman, 1997). Positive EPR, NSM, and LLR values indicate seaward movement (rate of accretion), whereas negative values represent landward movement of the shoreline (rate of erosion) (Moussaid et al., 2015).
Error Evaluation
To improve the precision of computed rates, uncertainties were systematically evaluated and accounted for. Several studies provided assessments of standard measurement errors linked to mapping methodologies and the digitization of shorelines (Aangri et al., 2022; Moore, 2000; Morton, Miller, and Moore, 2004; Ruggiero and List, 2009; Thieler and Danforth, 1994). The error in historical shoreline change (Esp) is determined by integrating measurement errors using Equation (1) (Morton, Miller, and Moore, 2005):
where, Ed represents the digitizing error, Er is the root-mean-square error of the georeferencing method, Ep is the pixel error, and EHWL accounts for variations in the position of the HWL induced by tide oscillations.
Shoreline Changes and Climatic Variability
The influence of global climate variability on wave height and wave direction has a significant effect on wave-energy flow (wave power), which is an indicator of the intensity of sediment transport that controls shoreline mobility (Charles et al., 2012). To assess the influence of climate variability on Moulay Bousselham’s coastal process, an analysis was conducted on the relationship between three climate indices (NAO, EAO, and WEPA) associated with the main atmospheric modes in the NE Atlantic and the influence on wave climate, particularly on the Moroccan Atlantic coast. The purpose of this part of study is to understand the relationship between the NAO, EAO, and WEPA teleconnection on winter climate and wave climate variability.
Data
Hourly sea-wave data were acquired from the SIMAR database point 1054040, situated at 6.50° W, 35° N (Figure 1a). Access to these data can be requested from Puertos del Estado (2024). The dataset covers the period from 1958 to 2016 and was generated using numerical simulations employing an advanced version of the WAM spectral model (Wamdi Group, 1988). The spatial resolution of SIMAR data is 1/24 of a degree. Beginning in 2001, the data is sourced from the WANA database. These data are the outcome of simulations performed using the WAM code, with an accuracy level of approximately 98%.
Monthly teleconnection indices for NAO and EAO have been available since January 1950 and are based on the empirical orthogonal function (EOF) analysis used by Barnston and Livezey (1987). These data were taken from the National Oceanic and Atmospheric Administration (NOAA) Climate Prediction Center (NOAA and National Weather Service, 2024).
The same approach used by Castelle et al. (2017) was used for the WEPA indices. Briefly, the winter WEPA was calculated as the difference anomaly December, January, February, and March (DJFM) SLP between the stations of Valentia (Ireland) and Santa Cruz de Tenerife (Canary Islands) obtained from the National Centers for Environmental Prediction/National Center for Atmospheric Research reanalysis project from January 1948 to April 2016 (Kalnay et al., 1996).
The winter wave-energy flux (wave power) for Moulay Bousselham’s coast from 1958 to 2016 can be calculated with the help of the following formula:
where, HS(DJFM) is the significant wave height averaging the monthly values for the boreal winter (DJFM), ρ is the sea-water density (1030 kg/m3), g is the gravitational acceleration (9.81 m/s2), and TS(DJFM) is the significant wave period averaging the monthly values for the boreal winter. Equation (2) can be simplified to Equation 3 as follows:
RESULTS
The temporal and spatial changes in the shoreline provide information on coastal landform dynamics (Aangri et al., 2022; Hakkou et al., 2018; Maiti and Bhattacharya, 2009). Accelerated accretion or decelerated erosion results from the interaction between natural and human processes. Therefore, accurate detection, frequent monitoring of shorelines, and analysis of variations in wave heights and wave-energy flux are essential to understand the coastal processes and dynamics of various coastal features.
The results obtained in this study through the diachronic analysis of the shoreline and the analysis of the variability of wave climate with climatic indices are presented in Table 2 and Figures 5–7.
Citation: Journal of Coastal Research 40, 5; 10.2112/JCOASTRES-D-23-00082.1
Citation: Journal of Coastal Research 40, 5; 10.2112/JCOASTRES-D-23-00082.1
Citation: Journal of Coastal Research 40, 5; 10.2112/JCOASTRES-D-23-00082.1
Historical Shoreline Changes
Using DSAS to analyze 704 transects along the Moulay Bousselham coast, the linear regression rate method revealed significant spatial variability in the study area over the past six decades (1949–2016). This variability is characterized by erosion in the northern part of the study area from the lagoon inlet (−0.42 m/y) and accretion in the southern part of the study area from the lagoon inlet (+0.10 m/y; Figure 5). Considering a margin of error of approximately ±0.32 m/y, in general, shoreline fluctuations manifest in three distinct phases of erosion (1949–88, 1991–94, 2004–07) and three phases of accretion (1988–91, 1994–2004, 2007–16). The results of these observations are consolidated in Table 2 and visually represented in Figure 5.
Period 1949–63
For the earliest period, spanning from 1949 to 1963 (14 years), the average shoreline mobility rate has been calculated at −2.78 m/y, representing the highest retreat value after the period 1991–94. Out of the 543 transects analyzed during this period, 424 transects are experiencing erosion (78.15%), whereas 119 are identified as accretive transects (21.8%). Erosion is observed from transect 184, with a maximum recorded erosion rate of −19.83 m/y. In contrast, from transect 1 to 183, positive values are observed, indicating beach accretion with a maximum recorded accretion rate of +3.58 m/y.
Period 1963–72
The 1963–72 period closely resembles the previous one, with an average rate estimated at −1.53 m/y. Out of the 543 analyzed transects, 419 transects (77.17%) were undergoing erosion, whereas 124 transects (22.82%) were experiencing accretion. The entire northern part of the study area from the inlet has regressed, with a maximum retreat of −57.07 m recorded on transect 178. Conversely, the examination of shoreline positions in the southern part of the study area from the inlet revealed an alternation between sectors experiencing accretion and others undergoing erosion.
Period 1972–88
During the 16-year period 1972–88, the average shoreline mobility rate was calculated at −0.69 m/y. Of the 704 analyzed transects, 496 transects (70.33%) exhibited erosion, whereas 208 transects (29.33%) showed accretion. The most significant retreat occurred at transect 298, with a rate of −181.74 m.
The three periods between 1949 and 1988 are thus comparable, depicting a predominantly negative trend with minimal fluctuation. The coastal system underwent an extended duration marked by continual retreat in the northern part of the study area from the inlet and alternating sections of accretion and erosion in the southern part of the study area from the inlet.
Period 1988–91
During the 3-year period from 1988 to 1991, the average rate of shoreline mobility was calculated at +2.01 m/y, indicating a substantial accretion along the coast during this period. The majority of the 704 analyzed transects exhibited accretion, with 617 recording accretion (87.55%), except for transects 321–336 and 528–586, where a retreating process was observed. The maximum accretion rate of 49.97 m/y was recorded at transect 301, north of the inlet. Interestingly, according to Ameur (1994), the inlet completely closed during this period and was artificially reopened in 1991.
Period 1991–94
During this period, the entire coast experienced erosion in 96% of the studied areas. This phase is characterized by the most pronounced retreat rate for the entire zone, reaching −4 m/y (Table 2). The highest erosion rates were observed in the northern part of the study area from the inlet, with a maximum retreat of −251.71 m. In accordance with the observations of Mhamdi Alaoui (2009), this period coincides with a substantial increase in longshore sediment transport.
Period 1994–97
Following a pronounced retreat phase in the previous period, the period between 1994 and 1997 (3 years) is characterized by the highest accumulation among all the studied periods, with an average of +4.56 m/y, affecting 98% of the analyzed transects. The aerial imagery from 1997 attests to the agglomeration of foreshore bars that directly supplied the intertidal zone to the north and south of the inlet, probably contributing to the advancement of the shoreline.
Period 1997–2004
Over the 7-year span from 1997 to 2004, the average coastal zone mobility rate was calculated at more than +15.14 m, equivalent to +2.16 m/y. The examination of shoreline positions indicates a cyclic pattern alternating between sectors in accretion and others in erosion. Among the 363 analyzed transects, 241 transects are experiencing accretion (67.96%), whereas 122 transects are undergoing erosion (33.59%).
Period 2004–07
Between 2004 and 2007 (3 years), among the 704 analyzed transects, 535 transects were in an erosion phase (76.03%), whereas 169 transects were in an accretion phase (23.96%). The calculation of the overall average rate established itself at −0.89 m/y. The analysis of the two shorelines during this period allowed the identification of two main accumulation sectors, located between transects 334 and 369 on the one hand and transects 663 and 704 on the other hand. This observation may be linked to a global readjustment of the coastal system in response to the significant accumulation observed during the two previous periods.
Period 2007–16
Over this 9-year period, the annual shoreline mobility rate is slightly positive, with an average of +0.81 m/y. The maximum accretion rate during this period is +5.83 m/y, observed at transect 457, in the southern part of the study area from the inlet, which did not record erosion rates in this period. In contrast, the northern part of the study area from the inlet exhibited an alternation between areas of accretion and erosion.
Atlantic Climate Variability
The data collected between 1958 and 2016 offer detailed insights into the seasonal variations of average wave height and wave-energy flux in the study area (Figure 6). The winter of 2014 was particularly notable, with a maximum wave height of about 2.32 m, whereas the winter of 2000 recorded the minimum value of about 1.07 m. These extremes underscore the importance of understanding annual fluctuations influencing oceanographic parameters.
The maximum wave-energy flux surpassed 35 kW/m during the study period, with an average of 15.6 kW/m. These values offer valuable indications of the energy involved in the winter oceanic phenomena off Moulay Bousselham, highlighting the significance of monitoring these parameters to better comprehend coastal conditions.
An in-depth analysis of the relationships between wave height and climate indices (NAO, EAO, and WEPA) was conducted (Figure 7). The results indicate a negative correlation with the NAO index, suggesting limited influence on the observed variability. Conversely, a moderate correlation with the EAO index was noted, whereas the WEPA index exhibited a significant correlation. This implies that variations in average wave height and wave-energy flux in winter are probably controlled by the WEPA index, emphasizing the effect of these specific climatic conditions on observed ocean dynamics.
In summary, this analysis provides a comprehensive perspective on winter oceanic phenomena off Moulay Bousselham, highlighting the complexity of interactions between climatic conditions and oceanographic parameters. These findings have significant implications for understanding long-term changes in the region and underscore the importance of the WEPA index as an indicator of local climate variability.
Climate Indices and Shoreline Dynamics
After analyzing the kinematic evolution of the coastline of Moulay Bousselham over a period of 67 years (1949–2016), it proved insightful to parallel the phases of Moulay Bousselham’s coastline evolution (advance/retreat) with the different phases (±) of climatic indices.
Several trends are seen in Figure 8. The first trend reveals a significant retreat during three periods from 1949 to 1963, from 1963 to 1972, and from 1972 to 1988, closely corresponding to predominantly negative phases of NAO and EAO, whereas WEPA shows strongly positive indices. The second trend, from 1988 to 1991, represents an advance in the coastline of Moulay Bousselham with positive phases of NAO and EAO, along with an alternation of positive and negative phases of WEPA. For the third phase (1991 to 1994, recording the most significant erosion of the studied coastline across all periods), the NAO index, although positive during this period, contradicts the results of the coastline evolution. For EAO, it alternates between positive and negative phases, whereas WEPA records a similar change, shifting from a negative dominance to a positive dominance, with the second-highest WEPA measured during the winter of 1993/94 (+1.945).
Citation: Journal of Coastal Research 40, 5; 10.2112/JCOASTRES-D-23-00082.1
The fourth trend is marked by the most significant advance of the coastline from 1994 to 1997 and from 1997 to 2004; it corresponds to alternating of positive and negative phases for all three indices (NAO, EAO, and WEPA), with a positive dominance of NAO and EAO and a negative dominance of WEPA. For the trend from 2004 to 2007, characterized by a retreat of the beach, the phase shows fluctuations in NAO and EAO, alternating between positive and negative phases with a negative dominance. The WEPA index, however, records predominantly negative phases, except for a positive peak in 2007. For the final trend (2007 to 2016, which is marked by an accumulation on the beaches of Moulay Bousselham), despite the winter of 2013/14 being the most energetic on the Moroccan coast since at least 1948 (Hakkou et al., 2019), alternating of positive and negative phases is recorded for all three climatic indices. It is noteworthy that only the WEPA, capturing the winter of 2013/14, corresponds to the highest WEPA in the series (+2.65), the most energetic winter in at least 70 years, which has significantly affected the Moroccan coastline.
DISCUSSION
The shoreline represents one of the most dynamically changing landforms in coastal regions, playing a pivotal role as a geoindicator in coastal evolution and providing crucial insights into coastal landform dynamics (Maiti and Bhattacharya, 2009). The position of the shoreline serves as a reflection of the coastal sediment budget, with alterations potentially indicating natural or anthropogenic influences alongshore or in neighboring river catchments. Therefore, precise detection and frequent monitoring of shorelines are imperative for comprehending coastal processes and the dynamics of various coastal features. Drastic changes in shoreline position can have significant and costly implications, directly affecting transportation routes, coastal infrastructure, communities, and ecosystems.
The outcomes derived from the diachronic analysis of the shoreline, using an automatic method, are detailed in Table 2 and Figure 5. The utilization of DSAS statistical parameters in this study facilitates an exploration of the temporal and spatial dynamics of coastal change and the geomorphic variability along the beach. These parameters encompass the incorporation of all shoreline positions, cumulative shoreline movement, and time variations, providing a comprehensive analysis of the historical dataset.
The shoreline changes statistics, including EPR and Linear Regression Rate disclosed in this study, unveil the large-scale patterns of retreat and growth in the shoreline of the case study. Throughout the period from 1949 to 2016, coastal evolution underwent diverse phases, comprising multiple intervals of erosion and accretion. The analysis reveals a prevailing trend of coastal regression between 1949 and 1963, characterized by 78.15% of transects experiencing erosion. This trend persists from 1963 to 1972 (77.17% in erosion) and from 1972 to 1988 (70.33% in erosion). However, a noteworthy transition occurred between 1988 and 1991, marked by a substantial accretion phase (87.55% of transects in accretion). This dynamic reverses between 1991 and 1994, signifying widespread erosion in 96% of the studied areas, followed by an exceptional accretion period between 1994 and 1997 (98% of transects in accretion).
The subsequent years up to 2016 exhibit cyclical variations, with sectors alternating between erosion and accretion, illustrating the complexity of coastal changes over time. Given the intricate nature, the shoreline displacement is influenced by various nonlinear physical processes interacting through complex feedback mechanisms across a wide range of spatial and temporal scales (Robinet et al., 2016; Stive et al., 2002). Several interacting factors, such as sea-level variations, wave energy, tidal currents, topography, geological configurations, marine sand extraction, depletion of groundwater, decrease in river discharge, and geologic phenomena such ad land uplift, contribute to shoreline changes in the medium term and long term (Aangri et al., 2022; Nicholls and Cazenave, 2010; Robinet et al., 2016; Toimil et al., 2020). In the short term, recent studies on coastal areas suggest that no direct relationship has been observed between the impact of storm waves on a decadal scale and coastal evolution (Chaverot, Héquette, and Cohen, 2006; Vespremeanu-Stroe et al., 2007).
This research established a correlation between climatic indices and shoreline displacement. Three climatic indices—NAO, EAO, and WEPA—were carefully chosen to explore the potential connection between shoreline movement and climatic conditions. Before delving into this relationship, a correlation analysis was conducted between wave height and wave-energy flux in the study area using data collected from 1958 to 2016. The findings revealed a negative correlation with the NAO and a moderate correlation with the EAO indices, suggesting a limited influence on the observed variability, whereas the WEPA index exhibited a significant correlation. This suggests that variations in average wave height and wave-energy flux in winter are likely influenced by the WEPA index. Consequently, the focus shifted to exploring the relationship between the three climatic indices (NAO, EAO, and WEPA) and shoreline displacement. The results highlight a clear relationship during the first three periods: from 1949 to 1963, from 1963 to 1972, and from 1972 to 1988. It is noteworthy that only the WEPA, during the period from 1972 to 1988, records transitions from a positive index to a negative index from one year to another, which could explain the low rate of erosion recorded during this period compared to previous phases.
Furthermore, the WEPA is the only index that shows a strong correlation with shoreline displacement phases from 1988 to 2016, capturing notably the winter of 2013/14 (Figure 8), which has proven to be the most energetic along the Moroccan coast since at least 1948 (Hakkou et al., 2019). Indeed, the winter of 2013/14 corresponds to the highest WEPA in the series (+2.65), whereas the second-highest WEPA was measured during the winter of 1993/94 (+1.945).
The observed relationship may appear modest, but it nevertheless represents a significant result, albeit subject to debate because of the methodological complexities inherent in the study. The calculation of shoreline displacement introduces several sources of uncertainty, ranging from measurement errors to the inherent complexity of coastal morphology. The study’s methodology reveals numerous challenges in accurately determining the shoreline’s movement, highlighting the need for careful consideration of potential uncertainties in the results.
Uncertainties arise not only from the employed measurement techniques but also from the unique characteristics of the study site. Coastal areas often exhibit complex morphologies, influenced by various factors such as local topography, sediment composition, and human interventions. These factors contribute to the dynamic nature of the coastline and introduce additional complexities into the calculation of shoreline displacement. Consequently, the observed relationship between the WEPA climatic index and shoreline movement, although significant, should be interpreted with caution.
Recognizing these methodological challenges and inherent uncertainties is crucial for accurately interpreting the study’s findings. Future research endeavors should focus on refining measurement techniques, accounting for site-specific complexities and incorporating advanced modeling approaches to enhance the robustness of results. Considering the dynamic nature of coastal environments, a more comprehensive understanding of the interactions among climatic factors, geological processes, and human activities is essential for a nuanced interpretation of shoreline dynamics. Although the identified link between the WEPA index and shoreline displacement is intriguing, ongoing research efforts should strive to address methodological limitations and refine the understanding of the intricate factors influencing coastal evolution.
CONCLUSIONS
This study explores the dynamic interplay between climate variability and shoreline changes in the Moulay Bousselham region over a 67-year period (1949–2016). The primary objectives were to map and quantify shoreline erosion and accretion rates and to investigate the influence of climate variability, as expressed by NAO, EAO and WEPA, on wave heights, wave-energy flux, and the mobility of the Moulay Bousselham shoreline.
The diachronic analysis of the shoreline reveals significant temporal and spatial variability, marked by distinct phases of erosion and accretion, illustrating periods of retreat and growth. The study identifies complex interactions between climate indices and shoreline dynamics, with the WEPA index emerging as a key indicator, demonstrating a robust correlation with shoreline displacement.
Correlation analysis between wave height, wave-energy flux, and climatic indices highlights the substantial influence of WEPA on winter oceanic phenomena off Moulay Bousselham. The study suggests that variations in average winter wave height and wave-energy flux are likely controlled by the WEPA index for the region of interest, emphasizing its importance in understanding local climate variability. These findings align with previous research, supporting the explanation of winter wave activity by the WEPA climate index at latitudes along the west European coast (R = 0.81, 0.76, and 0.6, respectively; Autret et al., 2018; Castelle et al., 2018; Castelle et al., 2022; Stéphan et al., 2018).
Ultimately, based on the obtained results, this study provides a fundamental assessment of the WEPA index in terms of its climate impact on winter wave activity along the Moulay Bousselham coast.
The observed relationships between climatic indices and shoreline displacement are subject to methodological complexities and uncertainties, but they also offer valuable insights. The study contributes to the broader understanding of coastal dynamics, emphasizing the need for nuanced interpretations considering the intricate factors influencing shoreline evolution.
Location of the study area. (a) Regional setting and wave rose diagram established from data obtained by the Puertos Datos record station at SIMAR calculation point 1054040 over the period 1958–2016. (b) Aerial photography of Moulay Bousselham and Merja Zerga lagoon, with field photographs (A, B, and C) showing the limit between dry and wet sand (watermark).
Simplified geomorphological map of the study area.
Flowchart showing the hierarchical structure of the shoreline change analysis.
Multitemporal shoreline positions north and south Merja Zerga inlet with the details of transect line (x) and shorelines intersection.
Shoreline evolution between 1949 and 2016 (erosion and accretion), calculated by the Linear Regression Rate and net shoreline movement (NSM) method.
Time series data for winter (DJFM) wave height (Hs), wave-energy flux (E) with NAO, EAO, and WEPA indices from 1958 to 2016.
The correlation of North Atlantic Oscillation (NAO), the East Atlantic Oscillation (EAO), and the West Europe Pressure Anomaly (WEPA) indices with wave height (Hs) and wave-energy flux (E).
Interannual variability of the NAO, EAO, and WEPA indices from 1950 to 2015 with periodic shoreline changes. Gray areas indicate periods of shoreline retreat, and white areas signify periods of shoreline accretion.
Contributor Notes
