On the Effect of Classical Versus Geopotential-jet Weather Regimes on Wintertime Rainfall Variability: Case of Morocco

Rainfall variability in North-West Africa, in particular Morocco, has profound socioeconomic impacts, with climate projections indicating a continued decrease in precipitation. However, interpreting rainfall projections is challenging due to substantial biases in current climate models, partly resulting from their low resolutions compared to precipitation scales. Large-scale North Atlantic atmospheric dynamics, e.g., the NAO, significantly impact Moroccan wintertime rainfall variability. As models resolve these large-scale dynamics comparatively well, we investigate their use as dynamical proxies for Moroccan rainfall variability, from the perspective of Euro-Atlantic weather regimes (WRs). The four classical WRs have previously shown limitations when used for downscaling Moroccan precipitation in the context of climate change (future climate simulations). Here we adopt recently-introduced 'Geopotential-Jet Regimes' (GJRs), using three and seven clusters, and compare their connection to observed Moroccan rainfall to those of classical WRs. We highlight that the NAO- regime is the main driver of winter rainfall in northwestern Morocco, producing rainfall levels approximately twice the climatological average, and that Scandinavian and European blocking have dramatically different rainfall teleconnections in North Africa. By comparing station data with regime-based reconstructions, this study finds significant correlations between North Atlantic WRs and winter rainfall in Morocco's most rainy and populous Northwestern region. Correlations average is around 0.6 for the three-GJR framework in the case of mean rainfall and the wet days fraction, it reaches 0.7 at some stations. The simpler three-regime GJR framework proves at least as effective as the four classical WRs in the historical period explaining about 36% to 49% of the total precipitation variance in nearly all the stations of the Northwest. This highlights their usefulness when combined with their previously demonstrated stable, well-reproduced regimes in CMIP6 simulations. This work therefore suggests GJRs may offer unique insights for improving projections of future rainfall changes in North-West Africa which we will pursue in future work.