Toward Development of Multi-model Subseasonal Probabilistic Forecasts of Precipitation and Temperature for West Africa

Abstract:

This study presents an assessment of multimodel subseasonal hindcasts and forecasts of precipitation and temperature over West Africa for the summer monsoon period, based on models from the S2S and SubX databases. Several regression-based calibration methodologies are compared using both deterministic and probabilistic skill metrics, from 1 to 4 weeks ahead. While skill is found to be largely limited to the first two weeks, aspects of predictability at longer subseasonal lead times are investigated in the context of ENSO and MJO sources of predictability.