Analysis of Drought Using Meteorological and Microwave Remote Sensed Data: A Case of Wami Watershed, Tanzania

Abstract

Agricultural sector is important for the economy of Tanzania, but in recent years there is decline in its growth and performance because of persistent droughts. An in-depth study of droughts was conducted on Wami watershed through rainfall and satellite microwave remote sensing data leading for estimates of meteorological droughts and soil moisture based droughts, respectively. Rainfall data during 1973-2008 was used to obtain Drought Severity Index (DSI) and active imaging microwave radar data during 1997-2009 from ESA’s SAR missions of ENVISAT and ERS was used to obtain soil moisture anomalies (SMA). Soil map was used to explain discrepancies in droughts from SMA to DSI maps at intervals of time. Seasonality analysis and DSI results showed that the main sub-seasons contributing to rainy season are October through December, January-February and March through May, and drought years were 1984, 1991, 1994, 2004 and 2006. Results showed that the last decade (2000s) had severe droughts that covered 35-39% of the Wami watershed and could have affected 1128000 people. The soil moisture based drought maps showed the same drought conditions as DSI maps in January, March, May and October. This indicated that in most areas the meteorological droughts can be used to infer to droughts conditions in the soil during the rainy season. The obtained drought events and impacts were confirmed in the field through interviews. However, in July SMA map showed normal and wet conditions whereas it was a dry season for DSI map. This showed that when rainy season ends, the soil still holds some moisture, which can be available for simple crops like vegetables. Therefore, it can be concluded that the SMA was able to provide a better alternative to DSI especially for increased spatial coverage and accuracy of drought monitoring for agricultural production. The SMA enables to map droughts conditions at any point spatially rather than point based DSI maps, which may be prone to rainfall data gaps and spatial interpolation errors. The SMA approach for drought monitoring may be useful to rainfall data scarcity areas of Tanzania and for agricultural droughts risk management.