Effects of Atlantic Ocean SST variability on African winter and summer precipitation

The goal of this analysis is to support the Famine Early Warning Systems Network (FEWS NET) activities through the CPC International Desks’ impact-based hazards outlooks, including effects of large-scale teleconnections leading to flooding, drought, and extreme events to support decision making.

The analysis focuses on seasonal prediction of precipitation over Africa using observational data, CFSv2 AMIP simulations, and NMME forecasts of Atlantic Ocean SSTs. The observational data include Optimal Interpolation version 2.1 SSTs (OIv2.1) and CPC unified precipitation data for the period of 1991-2022. Individual model’s ensemble mean was first calculated and then averaged with equal weight over the six NMME models (CanCM4i, CFSv2, GEM5-NEMO, GFDL-SPEAR, NASA-GEOS5v2, NCAR-CCSM4) to get a multi-model ensemble mean forecast over the 32-year (1991–2022) hindcast period. Multi-model ensemble anomaly was computed based on the 1991-2020 climatology. The seasonal correlation was computed for precipitation with Atlantic Nino (ATL3), Tropical North Atlantic (TNA), and TNA minus Nino34 (TNA-Nino34) SST indices for observation, CFSv2 AMIP and NMME 1 to 4-month forecasts.

The analysis focuses on interannual variability during winter (DJF) and summer (JJA) seasons. Through the observational diagnoses, AMIP simulation, NMME predictions, we identified the African winter and summer climate predictability and the SST source for the predictability. It is noted that tropical North (TNA) or equatorial (ATL3) Atlantic warming can result in above-normal rainfall in the tropical North Africa in boreal summer, while in boreal winter, positive gradient (TNA-Nino3.4) index can lead to wet Southern Africa. These results are useful in monitoring and forecasting of floods, droughts, and extreme events over Africa at the CPC international Desks.