Evaluation of the Skill of Monthly Precipitation Forecasts from Global Prediction Systems over the Greater Horn of Africa

Abstract The skill of precipitation forecasts from global prediction systems has a strong regional and seasonal dependence. Quantifying the models for different regions timescales is important, not only to improve forecast skill, but enhance effective uptake information. sub-seasonal (S2S) database contains near real-time re-forecasts 11 operational centres provides great opportunity evaluate compare S2S systems. This study evaluates these state-of-the-art in predicting monthly over Greater Horn Africa. comprehensive evaluation was performed using deterministic probabilistic verification metrics. Results analysis showed that varies with months region. Generally, show high during start rainfall season March lower peak April. ECCC, ECMWF, KMA, NCEP UKMO better region most compared rest models. Conversely, BoM, CMA, HMCR ISAC poor Overall, ECMWF model performs best among analyzed. Importantly, this serves as baseline assessment findings helping inform how subset could be selected construct an objectively consolidated multi-model ensemble products Africa region, recommended by World Meteorological Organization.