Advanced hydrogeological modeling to reduce uncertainty in groundwater recharge estimation in semi-arid aquifers: Case of Al-Haouz-Mejjate aquifer in Morocco

The Al-Haouz-Mejjate aquifer, a vital water resource in Morocco's semi-arid region, faces increasing pressure from agricultural intensification and climate variability. Despite its strategic importance for regional water security, previous studies have been limited by insufficient integration of multiple data sources and advanced computational methods, resulting in high uncertainty in recharge estimates. This study presents an integrated approach to constrain recharge estimation uncertainty by combining field data, geostatistics, machine learning, numerical modeling, and remote sensing to characterize aquifer dynamics over the period 1971–2024. The methodology incorporates 518 field pumping tests and lithological boreholes, Markov chain geostatistics, machine learning algorithms, and a calibrated 3D MODFLOW model, along with climate change scenarios and management strategies projected to 2040, to evaluate impacts on aquifer sustainability. Results indicate that total aquifer recharge increased from 343.05 Mm 3 /year in 1971 to 486.47 Mm 3 /year in 2024, while total discharge rose dramatically from 344.06 Mm 3 /year to 995.53 Mm 3 /year. The contribution from the High Atlas Mountains remained dominant at 65.3 % of total recharge, increasing to 317.69 Mm 3 /year, whereas rainfall infiltration and irrigation return flow reached 126.88 Mm 3 /year. Model simulations further reveal that aquifer storage depleted from approximately 17.15 billion m 3 to approximately 8.08 billion m 3 between 1971 and 2024, with maximum drawdowns exceeding 90 m. Future scenarios suggest that under current practices, storage could decline to approximately 4.2 billion m 3 by 2040, with drawdowns reaching 100 m; however, a management scenario involving a 25 % reduction in groundwater extraction could decrease the maximum drawdown from 100 m to 82 m and reduce the area affected by dry conditions from 4 % to 1.5 % of the total aquifer area. Alternative interventions, such as artificial recharge (adding 221.79 Mm 3 /year) show similar potential benefits. These findings demonstrate that reducing extraction rates through efficient irrigation technologies, modified cropping patterns, and improved water management practices have the potential to effectively mitigate both current aquifer stress and potential climate change impacts.