Water scarcity food security and economic growth in Africa using an integrated CGE machine learning framework for Morocco

Water scarcity is a structural constraint on economic development and food security in Africa, with Morocco particularly exposed because of its semi-arid climate and water-dependent agriculture. This study develops an integrated computable general equilibrium (CGE)-machine learning framework for Morocco and anchors its empirical accounting benchmark in official Social Accounting Matrix (SAM) workbooks published by the Haut-Commissariat au Plan (HCP). The 2023 SAM is used as the principal calibration benchmark, while the 2020 and 2021 matrices provide historical accounting-consistency checks. Between 2020 and 2023, domestic production rises from MAD 1,873,506 million to MAD 2,516,189 million and gross value added from MAD 1,036,049 million to MAD 1,340,158 million. In the study's 30% water-reduction scenario, real GDP declines by 4.1%, agricultural value added by 13.2%, and rural household income by 6.5%; the associated GDP uncertainty range is reported as a central 90% model-based interval of − 4.9% to − 3.3%. Machine learning is used as an emulator and sensitivity layer around CGE-generated scenarios rather than as a direct estimator of structural elasticities from the SAMs. A versioned Python reproducibility package is available from the corresponding author upon reasonable request and documents HCP-SAM extraction, model structure, closure, random seeds, ML settings, validation procedures, and figure generation, while explicitly separating SAM-derived evidence from external hydrological and behavioral inputs. The framework supports integrated water, food, economic-resilience, and climate-adaptation analysis linked to SDGs 2, 6, 8, and 13.