Performance Metrics of Medium-Range Precipitation Forecasts derived from the Global Forecast System Over Major River Basins in Africa
Abstract
This study evaluates the accuracy of medium-range precipitation forecasts derived from the Global Forecast System (GFS) for different transboundary river basins in Africa. The reference product used for evaluation is the research-quality, IMERG Final, product. The selected river basins are: Congo, Limpopo, Niger, Nile, Orange, Senegal, and Zambezi. The basins cover diverse climatic and topographic regimes. The results indicate that the performance of GFS varies from one basin to the other (and even within one basin) depending on the climate and topography. The forecasts are highly biased in the mountainous and wet climatic regimes, and relatively unbiased in the lowland and dry climatic regimes. The forecast accuracy decreases with increasing lead time, and decreases substantially for lead times exceeding 10 days. Averaging the forecasts at larger watershed sizes improves the accuracy of watershed-averaged GFS forecasts. Averaging the forecasts at longer timescales improves the accuracy of GFS forecasts. We recommend the use of post-processing calibration technique, through the use of near-real time products, such as, IMERG Early, to improve the performance of GFS.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFM.H15U1289Y