SST‑Gradient Driven Asymmetric Decline of Coastal Wind Power in Northwest Africa: Multi‑Model Attribution and Projections to 2100

This study quantifies the asymmetric decline in wind power density along Morocco's Mediterranean and Atlantic coasts and develops a multi-model projection framework. Analysis of ERA5 reanalysis data (1981–2020), evaluated against independent NASA POWER, MERRA-2, and NOAA OISST datasets, reveals a pronounced north–south contrast: the southern coast exhibits wind power density approximately 1.6 times higher than the northern coast, whereas the northern region shows 1.9 times greater inter-annual variability and an 11% faster rate of decline. A Random Forest attribution model identifies the Mediterranean–Canary Current sea-surface-temperature gradient as a dominant predictor of the north–south wind-speed differential, with the two SST predictors accounting for approximately 68% of the total permutation importance. The statistical relationship is consistent with a physical mechanism involving differential ocean warming and associated changes in the regional pressure gradient. Comparative analysis of constrained linear, logistic, quadratic, and CMIP6-informed projection approaches indicates that methodological choice introduces a 10–20-year range in projected threshold-crossing times. Five CMIP6 models were used to characterize structural model uncertainty under SSP1-2.6, SSP2-4.5, and SSP5-8.5. The results indicate substantially earlier loss of the 150 W m⁻² viability threshold in the northern region than in the southern region, although the precise crossing time depends strongly on the projection method and emissions scenario. These findings highlight the importance of physically informed, multi-model approaches for assessing climate-related risks to coastal wind-energy resources in Morocco.