Data assimilation of GPS-RO atmospheric profile data for improved rainfall forecasts over West Africa

Abstract: Forecasting rainfall is of great importance for the farmers in West Africa. However, due too lack of reliable weather observations, rainfall forecats in West Africa are difficult and primarly based on satellite observations. This thesis will study a satellite dataset that could possible work as a substitute for weather balloon soundings and thus improving the rainfall forecasts. A satellite dataset with atmospheric temperature and humidity profiles, obtained from GPS-RO, was compared with radiosondes available from Abidjan, Bamako and Niamey, to study the potential of improving rainfall forecasts over West Africa. Two case studies with simulated weather forecasts with and without assimilated GPS-RO data was also compared. Data assimilation is used to produce an estimate of the atmospheric properties. Temperature profiles obtained from GPS-RO data showed insignificant bias compared to the radiosondes. Probable humidity sensor failure resulted in problem analysing the dew point temperature. From simulations, it was shown that GPS-RO assimilation may have a large impact on the forecasts and could potentially be a substitute for radiosondes in West Africa.