Drought cascade lag time estimation across Africa based on remote sensing of hydrological cycle components

Abstract

Droughts in Africa have a disproportionate detrimental influence on food security, water supply, social equilibrium and the economy compared to other regions in the World. Crop growth is principally rain-fed with highly seasonal rainfall regimes over most of the Continent. Hydraulic infrastructure and harvest storage are scarce. Drought is primarily initiated by a lack of precipitation, and it spreads slowly to soil moisture, vegetation, streamflow, and groundwater, affecting the entirety hydrological processes across the landscape. This is called drought propagation and it is often characterized within a cascade framework linking the different types of droughts in sequence. In this study we perform continental-scale analysis of the drought cascade over the African Continent using remote sensing data from multiple sources. We examine the spatio-temporal variation of drought types. The drought cascade elements include precipitation, normalized difference vegetation index, total water storage, and surface soil moisture. Drought characteristics and the propagation time lag between different drought types are assessed over Africa. Time series analyses are conducted on the drought indices in order to estimate the behavior of each type of drought across twenty-three basins. Consistent propagation behavior is observed at continental and basin scales. The drought cascade begins with persistent precipitation deficit and translates instantaneously to soil moisture deficit, after one month to vegetation and after two months to total water storage of the hydrological system.