Abstract
The main objective of this study is to evaluate the uncertainties due to sample size associated with the estimation of the standardized precipitation index (SPI) and their impact on the level of confidence in drought monitoring in Africa using high-spatial-resolution data from short time series. To do this, two different rainfall datasets, each available on a monthly basis, were analyzed over four river basins in Africa—Oum er-Rbia, Limpopo, Niger, and eastern Nile—as well as at the continental level. The two precipitation datasets used were the Tropical Rainfall Measuring Mission (TRMM) satellite monthly rainfall product 3B43 and the Global Precipitation Climatology Centre full-reanalysis gridded precipitation dataset. A nonparametric resampling bootstrap approach was used to compute the confidence bands associated with the SPI estimation, which are essential for making a qualified assessment of drought events. The comparative analysis of different datasets suggests that for reliable drought monitoring over Africa it is feasible to use short time series of remote sensing precipitation data, such as those from TRMM, that have a higher spatial resolution than other gridded precipitation data. The proposed approach for drought monitoring has the potential to be used in support of decision making at both continental and subcontinental scales over Africa or over other regions that have a sparse distribution of rainfall measurement instruments.
