This study analyzes perceptions and representations of drought among residents of the Oued El Abid watershed in Morocco, based on a structured field survey conducted with 150 households. The results reveal a high level of awareness of drought, with 95% of respondents acknowledging its occurrence and 80% considering their area to be directly affected. Nearly 75% of the surveyed population report having personally experienced its impacts, particularly reduced precipitation, water scarcity, and the drying up of springs and streams. Local populations perceive a significant deterioration in climatic conditions, as 58.4% currently consider precipitation to be low compared to 45.6% who previously described it as abundant. In addition, 72.2% of respondents report an increase in temperatures, and 70.4% identify recent decades as the driest. Drought is described as a multidimensional phenomenon encompassing climatic, hydrological, agricultural, and socio-economic aspects, with direct repercussions on water resources, agriculture, and livestock. The study highlights critical impacts on water supply, marked by reduced flows in springs, rivers, and wells, as well as significant economic consequences, including crop losses, declining agricultural yields, and livestock mortality, thereby increasing the socio-economic vulnerability of local populations. Furthermore, adaptation strategies such as the semi-nomadic mobility of herds illustrate the capacity of communities to adjust their practices in response to environmental constraints. Overall, these findings emphasize that drought represents a major environmental and socio-economic challenge in the Oued El Abid watershed, and that integrating local perceptions is essential for developing sustainable water resource management strategies and strengthening community resilience to climate change.

This study proposes a segmentation & GPLM framework to estimate nonstationary rating curves, and their uncertainties at the Tamgnounte hydrometric station, Moroc-co. Based on 176 stage discharge gaugings from 2000 to 2020, a residual based PELT & CROPS procedure was first used to detect homogeneous periods, and GPLM was then fitted independently, within each segment. Results identified robust changepoints at positions 40, 80, and 105, while an additional changepoint at 150 was detected only under a lower CROPS penalty range. Segment specific GPLM calibrations showed sat-isfactory fits, with most observations within the 95% posterior predictive intervals and residuals generally centered around zero. These results confirm the nonstationarity of the stage discharge relationship and support the use of one rating curve per homoge neous period rather than a single global curve.