Assessment of agricultural drought over Africa and its relation to El Niño-Southern Oscillation using remote sensing-based time series

Kurzfassung

Ranked amongst the most destructive natural disasters of the world, droughts may have severe impacts on ecosystems and society. Particularly in Africa, where water is a limiting factor and countries strongly rely on rain-fed agriculture, droughts have constantly led to widespread crop failure, food shortages and even humanitarian crises. In regions over eastern and southern Africa, such dry conditions have been attributed to the effect of El Niño-Southern Oscillation (ENSO). Given the recent El Niño episode of 2015/16 and the associated severe droughts that occurred in many parts of Africa, this interconnection has once again become an issue of importance. In this regard, remote sensing data and image analysis provide new opportunities for generating substantial information on the evolution of droughts at large spatial and temporal scales. This thesis focusses on monitoring agricultural droughts over Africa during 2000-2016 and their relation to ENSO by means of remote sensing time series. The used continental-scale approach is based on drought indices. In particular, TRMM-based Standardized Precipitation Index (SPI) and MODIS-derived Vegetation Condition Index (VCI) were used for analysing the spatio-temporal patterns of agricultural droughts. All in all, a comprehensive insight into the evolution of agricultural droughts in Africa was gained. The applicability of SPI and VCI as indices for continental-scale drought monitoring was proven. Observed discrepancies were linked with variabilities in sensitivity of vegetation to rainfall over Africa, which in turn merits further research. Moreover, the relation between droughts and ENSO was examined by applying a correlation analysis between time series of drought indices and Multivariate ENSO Index (MEI). This complex relationship could be described in its fundamentals. Based on revealed correlation patterns, droughts tend to occur during El Ni~no over large parts of southern Africa. In contrast, a divided pattern was observed in eastern Africa, where areas with bimodal annual rainfall cycles tend to be affected by droughts during La Niña and, in zones of unimodal rainfall regimes, droughts tend to arise during the onset of El Niño. However, no universal El Niño- or La Niña-related response pattern of droughts could be deduced. In this regard, multi-year atmospheric fluctuations and characteristics of ENSO variants were discussed as possible influencing factors. Regional impacts of the drought episodes during El Niño 2002/03 and La Niña 2010/11 were illuminated by comparing observed regional drought patterns with statistics on national crop production. Focus is laid on each southern and eastern Africa, where decreases in production numbers were observed for major drought-affected countries. In order to achieve improvements in quality and reliability of the output, the incorporation of more accurate cropland information, adaptions of the correlation analysis as well as an uncertainty assessment and a validation are proposed. Using remote sensing data as a toolset for drought monitoring, this thesis represents an attempt to contribute to a better understanding of spatio-temporal patterns of agricultural droughts in Africa and their dependencies. Such knowledge is essential as it forms the basis for implementing strategies of drought hazard mitigation in the affected regions.