The Souss-Massa basin is located in the center-west of Morocco covering an area of about 22,600 km², 25% of which is the plain area and 75% mountainous areas (Figure 1). It extends between 29° and 31° of latitude north and 7° and 10° of longitude west; bounded on the east by the Atlantic Ocean, on the north by the High Atlas Mountains, and on the south by the Anti-Atlas Mountains. The climate is classified as arid to semi-arid, with temperatures averaging around 14 °C in the north (High Atlas) and 20 °C in the south (Anti-Atlas), regarding temperature extremes, the maximum daily temperature can reach as high as 49 °C, while the minimum temperature can drop to as low as −3 °C (Hssaisoune et al., 2016). The precipitation levels exhibit significant variability both temporally and spatially (Abahous et al., 2018; Bouchaou et al., 2011). Indeed, rainfall demonstrates a notable diversity, ranging from 300 to 600 mm in the High Atlas to approximately 200 mm annually in the plain (Bouragba et al., 2011), with a wet period from October to April and a dry period from May to September. According to Hssaisoune et al. (2016), the Souss-Massa basin features two significant aquifers: a shallow aquifer (Plio-Quaternary) and a deep one (Cenomanian–Turonian). The shallow water table is the most important aquifer, flowing from east to west through recent sedimentary deposits. It plays a vital role in water supply for domestic and agricultural uses. The deep aquifer consists of Neogene conglomerates and Cenomanian–Turonian limestones, manifesting as artesian waters in specific boreholes. In the Souss plain, recharge primarily occurs from the High Atlas Mountains on the right side and a very light network of tributaries originating from the Anti-Atlas Mountains on the southern side, while the Massa basin system has only the headwaters from the Anti-Atlas Mountains (Bouchaou et al., 2008; El Malki et al., 2017; Danni et al., 2019; Hssaisoune et al., 2019, 2021). The Souss River and Massa River are the main drainage systems; Figure 1 presents a cross-section of the deep and shallow aquifer.

The Souss-Massa River basin is located in the southwestern part of Morocco. It is characterized by a semi-arid climate, a growing population, where water resources are vulnerable to scarcity (Mansir et al., 2021). The region’s economy heavily depends on agriculture, fishery, and tourism (Abahous et al., 2018; Bouchaou et al., 2017; Guemouria et al., 2023).

Actually, Souss-Massa is a pivotal area for Morocco’s agriculture, accounting for approximately 90% of fruit and vegetable national exports (Hirich et al., 2015; Abou Ali et al., 2023). This region is not only a significant contributor to Morocco’s economy but also plays a vital role in the global food supply. In particular, Souss-Massa produces more than half of Morocco’s citrus exports and 85% of its vegetable exports, highlighting its importance in both local and international markets (Abou Ali et al., 2024; Choukr-allah et al., 2016).

Despite its agricultural importance, Souss-Massa faces critical challenges due to severe groundwater depletion and climate variability (Milewski et al., 2020).

The selection of the Souss-Massa basin as a case study is particularly relevant due to its status as one of Morocco’s most agriculturally developed regions, which significantly contributes to the national economy (Mansir et al., 2018), in addition to its vital role in the international markets through exportation of citrus, fruits, and vegetables (Choukr-allah et al., 2016). However, the region faces critical challenges related to water resource management, exacerbated by climate change and increasing agricultural demands. Previous studies have highlighted the impacts of drought on water quality and availability. Yet, there remains a notable gap in comprehensive research specifically addressing the effects of drought on water resources at the basin scale (Mansir et al., 2018). Our study aims to fill this gap by providing a detailed analysis of drought impacts on water resources.

The precipitation datasets were collected from local meteorological stations under the authority of the Hydraulic Basin Agency of Souss-Massa (ABHSM). Only 7 stations were selected for this study (Table 1). This selection is based on 3 criteria: (i) the distribution of meteorological stations over the Souss-Massa basin, (ii) the length of the rainfall data series, and (iii) the quality of the rainfall series data (fewer missing data). Definitely, for all hydrological research, it is important to analyze the spatiotemporal dynamics of precipitation. Certainly, the rain gauges of precipitation offer data quality at the locally scale. However, they usually do not provide global coverage; consequently, the time-series data may contain missing values. We found that data gaps in observed precipitation are quite low, ranging from 0.2 to 8% of the monthly records during the study period (1980–2023). To fulfill these, we incorporated data from The Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (GPM IMERG-Final V06), denoted as GPM-F. This choice was recommended by Rachdane et al. (2022) as they confirmed the accuracy of GPM-F compared with the data observed in the Souss-Massa basin. Moreover, they highlighted the good performance of GPM-F in arid regions and recommended its use for diverse hydrological studies, including drought assessment, flood monitoring, and landslide forecasting. The GMP-F data were obtained from NASA’s website¹.

The piezometric level time series, used in this study, was also provided by the ABHSM. Groundwater level analysis was conducted using monthly data from 18 piezometers (Table 2), distributed over the Souss-Massa plain (Figure 2), covering a period from 1960 for the Souss basin and from 1970 for the Massa basin.

The climate change assessment is based on precipitation trends analysis (moving average) and the standardized precipitation index (SPI) is proposed by Mckee et al. (1993) to identify the rainfall trends, and the drought frequency and duration, using only rainfall data.

The SPI is defined as the ratio of the difference of precipitation from the mean for a specified period and the standard deviation (Equation 1). Where the mean and standard deviation are determined from records (Mckee et al., 1993):

With P = monthly precipitation, P _ = long-term mean precipitation, σP = standard deviation of precipitation.

A functional and quantitative definition of drought can be established for different time scales by utilizing the SPI as an indicator. According to the SPI, a drought event is identified as a continuous negative value. The onset of drought occurs when the SPI falls below zero for the first time, and it ends when the SPI becomes positive again. The intensity of the drought event is determined based on arbitrary categorizations associated with SPI values. These categories can be found in Table 3.

Hydrological droughts are initiated just like other droughts, by a shortage of precipitation and therefore a decline in groundwater recharge. The intensification of these droughts is exacerbated by mismanagement and excessive exploitation of aquifer resources, which exacerbates their impacts. In this study, the SWI proposed by Bhuiyan (2004) was used to evaluate groundwater deficit (Equation 2). The SWI builds on the SPI framework to compute standardized differences in groundwater depth (GWD), thereby quantifying anomalies in groundwater levels in relation to their long-term mean. SWI is often computed for its simplicity as follows:

Where: GWDi: the seasonal water level for observation (i), GWD _ : seasonal mean of GWD, and 𝜎_GWD the standard deviation of GWD.

The SWI was computed on a seasonal basis, focusing on the summer months (June–September), when groundwater stress is typically highest in the Souss-Massa basin. Each season includes four monthly groundwater depth observations, and the seasonal mean was used as the representative value for each year (Bhuiyan, 2004; Sadeghfam et al., 2018).

SWI ranges are designated to define drought bands to trigger appropriate management actions. As groundwater level is assessed downwards from the Earth’s surface, positive anomalies indicate drought, while negative anomalies reflect normal or wet conditions. Bands classified by Bhuiyan (2004) are given in Table 4, and their designated ranges are also mentioned.

The evaluation of three-month accumulation periods reveals cyclical fluctuations that are not suitable for the precise prediction of drought events; such short periods may capture short-lived drought spells but can be less effective in characterizing prolonged and severe drought conditions. Instead, we used a longer-term cumulative rainfall approach (q = 12) for a more comprehensive analysis of drought characteristics in the basins.

The evolution of the SPI index for the seven selected rainfall stations over the period 1980–2020 is presented in Figure 4 (The full SPI series for all seven stations are given in Supplementary Figure S1). From the results, it is evident that most stations exhibit rainfall deficiency, with values reaching as low as −2 in some stations. The period from 2000 to 2010 is characterized by a deficit in rainfall at all stations, with SPI index values dropping below −1 (indicating severe drought), while positive values are slightly more frequent. After 2010, it is clear that all stations experienced an increase in humidity, with values greater than 2 observed in some stations. Nevertheless, rainfall subsequently becomes systematically deficient until the end of the observation period. Further comparative analysis of drought patterns between the Massa and Souss basins displays different intensities and durations, with the Souss basin showing higher SPI values, indicating lower drought intensity and longer wet periods. To highlight this evolution the mean SPI for the entire region was examined to offer an overview of the meteorological drought pattern in the research area (Figure 5). Since the end of the ‘80s, the Souss-Massa basin has experienced significant changes, manifested by alternating dry and wet periods; there are seven moderately wet periods with an average of SPI = 0.31 and one very wet event in 2010. The longest wet period ranged from 1987 to 1990, with an SPI value average not exceeding 1, followed by the longest dry period, which was extended from 1990 to 1995. Since 2010, climate variability has been well noticed and highlighted by successive increases and decreases of SPI values reflecting sometimes-dry conditions and sometimes-wet conditions.

From the available data by the ABHSM, a total of 18 piezometers: 14 from Souss and 4 from Massa aquifer were selected. The choice of the piezometers to be investigated was based on their location and the availability of reliable long-term data, Figure 6 shows only 4 representative piezometers from this selection.

The piezometric level variation in the Souss basin can be divided into two groups. The first group exhibits minimal variations and after a decrease, they return to their initial position (Figure 6a). This includes for example piezometer 701/62. As for the other piezometer, 859/62 despite its location in the Upstream Souss, it experienced a considerable decline in groundwater levels, but during 1996 there was a notable increase in groundwater level. The rapid drop rates of 1.9 m/year (1998–2008) and 3.5 m/year (2012–2020) indicate a significant influence (Figure 6a).

The other borehole 4810/70 is located in the Middle Souss; its piezometric level continuously decreases in the long term (Figure 6b), and there is a temporary light rise in groundwater levels that should be noted.

Concerning the Massa basin (Figure 6d), an important agricultural area, the boreholes consistently show a gradual decline in water levels over the years. This decline averaged one meter per year in last decades. It is worth noting that some piezometers in the Massa basin were already dry by 2018.

The Groundwater level is measured from the surface, and SWI was calculated for groundwater drought assessment, with positive anomalies indicating drought conditions and negative anomalies indicating normal conditions. A cross-correlation analysis was performed on the SPI and the SWI while varying the duration of the SPI at different time scales.

As shown in Figure 7, since 1965 in the Souss Basin, the trend in the SWI values has been generally toward negative values, indicating wet periods. However, from the early 2000s onwards, the region has experienced a change toward mild drought (0.5 < SWI < 1). The correlation between the SWI and SPI was found to be maximum when the accumulation period is 1 month.

In the Massa basin, a consistent trend toward dryer conditions has been observed since 1999 (Figure 8). The years 2019 and 2020 were marked by severe groundwater drought intensity (SWI = 1.8). The results showed that the accumulation period required to achieve the maximum correlation between the SPI and SWI time series is 4 months for the Massa basin. However, the correlation between the SPI and SWI remains very low (r = 0.07).

Figure 9 illustrates the spatial distribution of resistance across various drought bands, and it can serve as a drought vulnerability.

Our results show that the Massa aquifer demonstrates low resistance, withstanding only a maximum of 54 meters of variation under mild, moderate, and extreme drought conditions. An 18.5% increase in resistance is observed during extreme drought, reaching 63 meters. For the middle Souss region, resistance to loads went to a maximum of 146.04 m under mild drought, 151.56 m for moderate drought, 202 m for severe drought, and 254.09 m under extreme drought. As in the upstream Souss, resistance values vary from 7.7 to 164.98 m across different drought bands. Thereafter, the middle and upstream Souss regions exhibit diverse resistance values, indicating varying sensitivities to groundwater fluctuations.

Figure 10 illustrates the spatial distribution of Groundwater Drought Risk (GDR) across various drought severity bands (Mild, Moderate, Severe, and Extreme) in the Souss-Massa basin. These maps offer a detailed visualization of how drought risk varies geographically and across different levels of drought severity. Under the mild drought band, the GDR values range from 19.2 to 57.94%, we note that the Massa basin exhibits the highest GDR values within this category, suggesting a significant susceptibility to mild drought conditions. These high GDR values in the Massa basin (indicated by the red and orange areas) imply a higher probability and potential frequency of experiencing mild droughts. For the moderate drought band, the values range from 3.07 to 45.64%. As the severity of drought increases to moderate, there is a noticeable decrease in GDR values, reflecting a reduced frequency but potentially greater impact of each event. The downstream and middle Souss regions display moderate GDR values (green to yellow zones), indicating a balanced risk, while lower than in the Massa basin (Red color). The severe and extreme drought categories show a further reduction in GDR values; 1.2 to 40.86% and 0.57 to 36.62% respectively; highlighting the infrequent but potentially destructive nature of these droughts. Despite their low frequency, these droughts could have critical impacts, particularly in areas like the Massa basin where GDR values are still relatively high compared to other regions.

Figure 11 illustrate the spatial distribution of EHI within Souss-Massa plain. The results showed elevated EHI values, particularly in the Massa basin, where drought risk reached over 80% under mild drought conditions. Downstream Souss and Middle Souss regions exhibited drought risks between 60 and 70%. The Upstream Souss region had a comparatively lower risk, with values below 40%. For other drought bands, the results indicated a concerning situation in the Massa basin, with a sharp change in groundwater levels.

The Souss-Massa basin, a vital agricultural area in Morocco, has experienced significant changes in climatic and hydrological conditions. The spatial and temporal variability in rainfall observed in our results is consistent with previous studies. Bouchaou et al. (2011) and Ait Brahim et al. (2016) reported substantial differences in rainfall distribution across the Souss Massa basin. Similarly, Abahous et al. (2018) documented a long-term decline in rainfall since the 1960s, highlighting a shift toward more arid climatic conditions. These changes have important implications for both water availability and agricultural productivity.

The analysis of meteorological drought conditions in the Souss-Massa basin reveals a clear upward trend in both frequency and severity of drought events in recent decades. This is evidenced by SPI values frequently falling below normal, which denote severe drought conditions. Our findings are consistent with the observations made by Balaghi et al. (2010), who documented a marked increase in drought occurrences in Morocco since 1980, highlighting a shift from experiencing one dry year every 15 years to one every 3 years. A local study by Ait Brahim et al. (2017), highlighted a significant rise in the frequency of dry years since the 1970s in the Souss-Massa basin, emphasizing the climatic shifts that threaten sustainable water management practices in the region. Moreover, our SPI-12 analysis indicates that the drought conditions observed are not episodic or isolated but part of an enduring pattern of rainfall deficits. This aligns with the findings of Bijaber et al. (2018), who emphasized the impact of climate change on the frequency and intensity of droughts in Morocco, all in the same way, previous studies conducted by Bouchaou et al. (2011); Mansir et al. (2018); and Seif-Ennasr et al. (2017) and Hssaisoune et al., 2021 confirm the recurrent fluctuation of wet and dry cycles in the Souss-Massa basin. The historical patterns described by Abahous et al. (2018) further illustrate alternating cycles of drought since 1980s.

The comparative analysis of drought patterns between the Massa and Souss basins further elucidates the spatial variability in drought intensity and duration, in agreement with the findings of Yoo et al. (2020), who highlighted that spatial heterogeneity plays a key role in the propagation and persistence of drought events. Rainfall variability in the Souss-Massa basin is influenced by several geographic factors. The High Atlas Mountains enhance rainfall through orographic effects, while the inland plains are drier due to the rain shadow effect. Elevation, valley orientation, and the influence of the Atlantic Ocean further explain the spatial contrasts observed (Hssaisoune et al., 2017).

These meteorological droughts have a direct impact on groundwater resources, with effects that vary depending on the location and land use. Groundwater in the Souss basin is recharged primarily by the High and Anti-Atlas Mountains (Bouchaou et al., 2008; Hssaisoune et al., 2017). In recharge-favored zones, such as near High Atlas, the aquifer response to droughts is buffered by natural infiltration (Hssaisoune et al., 2020). Piezometer 701/62 (Upstream Souss), for example, exhibits recovery during wet periods and dam releases like those in 2004 (Gouahi et al., 2024; Bouragba et al., 2011), reflecting the aquifer’s resilience to meteorological variability. In contrast, Middle and downstream areas in addition to the Massa plain, saw a rapid and prolonged groundwater level declines, due to limited recharge and high abstraction rates. These spatial variations are further explored through the Pearson correlation between SPI and SWI values. Although this correlation is relatively weak, this does not negate the impact of rainfall shortage on the recharge process (Bouimouass et al., 2020). On the contrary, the water’s decline still reflects the cumulative effect of reduced rainfall combined with the anthropogenic pressures.

The methodological framework used in this study is built upon the reliability concept, where aquifer resistance and external loads are integrated to evaluate the probability of failure and quantify the Groundwater Drought Risk (GDR). This approach allows for distinguishing areas that are more likely to experience groundwater drought under different stress conditions. The Environmental Hazard Index (EHI), on the other hand, serves as an important metric for the assessment of the increasing risk over time. Together, these indicators provide complementary insights into the spatiotemporal dynamics of groundwater stress in Souss-Massa.

Our results clearly illustrate that the Massa aquifer exhibits relatively low resistance, making it more vulnerable to groundwater depletion and drought. The resistance values, particularly during extreme drought conditions, are concerning, as the aquifer can only withstand groundwater fluctuations of up to 63 meters. This low level of resilience is further highlighted by the complete drying-up of two boreholes in the region, a powerful indicator of the aquifer’s inability to cope with prolonged drought conditions. These results highlight the critical state of the Massa aquifer and underscore the urgent need to adopt water management strategies to mitigate the effects of future droughts. In contrast, the middle and upstream Souss regions display a broader range of resistance values, suggesting a more diverse response to groundwater fluctuations. For example, in the middle Souss, resistance reached up to 254.09 meters under extreme drought conditions, reflecting a higher capacity to withstand significant groundwater stress. This suggests that certain areas of the Souss aquifer have a greater ability to maintain stability, even in the face of extreme changes in groundwater levels.

The variation in resistance values between the Souss and Massa regions underscores the heterogeneity of the aquifer system. The higher resistance observed in the Souss suggests a more robust system capable of withstanding larger changes in groundwater depth, while the Massa aquifer remains more fragile and susceptible to drought impacts. These findings align with those of Mansir et al. (2021), contributing to a growing body of research that deepens our understanding of the vulnerability of the Souss-Massa basin to water scarcity.

Building on that concept of vulnerability (resistance), GDR quantifies the risk of groundwater drought by evaluating aquifer resistance to the intensity of hydrological stress. When the stress exceeds the resistance, the risk of drought increases. This allows GDR to spatially highlight areas where groundwater systems are most likely to fail. Higher GDR values indicate a greater probability of experiencing severe droughts, although these events tend to occur less frequently. Conversely, lower GDR values suggest a lower susceptibility to severe drought, but these areas may experience more frequent drought events, albeit with less severe impacts. This inverse relationship between the frequency and severity of drought events, as indicated by the GDR values, has been corroborated by earlier research, including studies by Roshni et al. (2019) and Xu et al. (2021).

The spatial distribution of GDR values across the Souss-Massa basin illustrates a clear distinction between the Souss and Massa regions. The Souss upstream region exhibits relatively low GDR values, suggesting a lower risk of groundwater drought. This area appears to be more resilient, likely due to its higher resistance to groundwater fluctuations. As we move downstream in the Souss basin, GDR values gradually increase but remain within manageable levels, indicating that the region is still relatively resilient to drought conditions. In contrast, the Massa basin stands out with significantly higher GDR values, indicating a much greater risk of groundwater drought. The higher GDR values in Massa suggest that the region is more vulnerable to severe drought impacts. This increased risk is likely due to the lower resistance values observed in the Massa aquifer, which struggles to cope with the stress imposed by prolonged drought conditions. The drying-up of boreholes, as previously noted, underscores this vulnerability.

When examining the drought categories, the GDR values reflect the region’s vulnerability across different severity levels. Mild and moderate droughts are relatively frequent across the basin, with GDR values indicating that these events are likely to occur regularly. However, these droughts tend to have less severe impacts, particularly in the Souss region, where resistance to groundwater fluctuations is higher. On the other hand, severe and extreme droughts are much less frequent but can have catastrophic effects, particularly in the Massa basin, where groundwater resources are more vulnerable. The relationship between GDR and resistance is also evident in these findings. Areas with small resistance values, such as the Massa basin, tend to have higher GDR values, reflecting their increased risk of drought. This is a normal correlation in groundwater systems: lower resistance implies a lower capacity to withstand fluctuations in groundwater levels, leading to a higher probability of system failure during drought conditions (Hssaisoune et al., 2016; Sadeghfam et al., 2018; Roshni et al., 2019).

EHI complements this assessment by tracing the evolution of drought risk over time. The index shows a dramatic increase in groundwater stress in the Massa basin over the past two decades, with EHI values exceeding 80% under mild drought conditions. This observation aligns with the drying-up of two boreholes in the area by 2018, further underscoring the severity of the situation. The Souss basin, particularly the Downstream and Middle Souss regions, shows moderately high EHI values. Although the risk is not as severe as in the Massa basin, the steady increase in drought risk suggests that these regions are also becoming more vulnerable to groundwater depletion and associated environmental hazards. Meanwhile, upstream Souss remains more resilient, with GDR and EHI values both below 40%. However, changes in climate or water use could alter this status in the future. There are also contrasting patterns, such as around piezometer 937/69, where high GDR but low EHI suggests temporary drought exposure with good recovery potential, likely due to seasonal recharge. Conversely, piezometers 04/69, 859/69, and 866/59 report high GDR and EHI, signifying chronic drought risk and heightened hazard. This includes the threat of saltwater intrusion in coastal zones, as confirmed by Ez-zaouy et al. (2022, 2023) and Malki et al. (2016).

Despite the diversity of drought indices, SPI remains the most used in drought monitoring for its robustness and its low data inputs (World Meteorological Organization (WMO) and Global Water Partnership (GWP), 2016). Thus, this index is highlighted by WMO as the starting point of meteorological drought and serves as an early warning signal. Furthermore, its multi-scalar nature allows the evaluation of droughts over different timescales (e.g., 1, 3, 12 months), which is essential to capture short-term meteorological versus long-term hydrological droughts. However, SPI does not account for evapotranspiration, which is critical in semi-arid regions like Souss Massa where water loss through ET is high. In addition, SPI relies on the assumption that precipitation follows a stationary probability distribution, which may be invalid under shifting climate conditions or long-term trends (Stagge et al., 2015). This limitation was acknowledged in our study, and the integration of SPEI is highlighted as a perspective for future research.

In this study, the Standardized Precipitation Index (SPI) was correlated with groundwater levels using the Standardized Water-level Index (SWI) to explore the linkage between meteorological and groundwater droughts (Figures 7, 8). While this approach offers insights into the broader trends of drought propagation, the observed low correlation (r = 0.07) likely reflects multiple interacting factors (Schreiner-McGraw and Ajami, 2021). In particular, ABHSM reports indicate piezometric declines of 1–2.5 m/year, with cumulative drops of up to 30 m in 30 years, reflecting a structural deficit of 260–300 Mm³/year (ABHSM, 2018). Irrigation accounts for more than 85–90% of withdrawals, mainly citrus and vegetable crops (Mansir et al., 2018), and groundwater abstraction increased dramatically from 747 Mm³ in 2007 to 4,884 Mm³ in 2020 (Guemouria et al., 2023). Isotopic studies (Hssaisoune et al., 2017) also confirm that modern recharge is spatially limited and delayed. Together, these studies provide proof that explains why groundwater drought in Souss-Massa is not solely climate-driven but strongly shaped by anthropogenic demand.

Resistance and GDR values and their spatial distribution demonstrate high spatial variability across the basin, reflecting heterogeneous responses to drought events. Overall, the risk and vulnerability to groundwater drought are disproportionately concentrated in middle and downstream regions (Figures 9, 10), which can be attributed to several interconnected factors: