Meteorological drought monitoring in the Upper Olifants sub-basin, South Africa

Abstract

Measured climate data are often scarce and expensive to collect through a network of measurement points. There are various reanalysis products that can be used in certain regions in studies concerned with drought monitoring. This study aimed at assessment of the adequacy of using the National Centers for Environmental Prediction Climate Forecast System Reanalysis (NCEP-CFSR) rainfall data and the ERA5 rainfall data from the European Centre for Medium-Range Weather Forecasts (ECMWF) in monitoring meteorological drought over the Upper Olifants sub-basin, South Africa. A comparison was made between the observed rainfall data and the two reanalysis products based on the Mean Error (ME), the Nash-Sutcliffe Efficiency (NSE) and percentage of bias (PBIAS). The data were also used in computing and comparing the results of the Standardised Precipitation Index (SPI). The CFSR were found to underestimate rainfall amounts, while the ERA5 overestimated the rainfall amounts except during winter. The ERA5 showed a better correlation which was R2 = 0.78 when compared to the one for the CFSR, R2 = 0.57. Also, the NSE, ME and PBIAS for the ERA5 products were generally better in almost all-time scales. The three products detected one of the major droughts but the CFSR were unable to detect the second major drought in the study area during the chosen study period. The study concludes that the ERA5 is a better product for meteorological drought monitoring in the study area and can be useful in predicting meteorological drought using artificial intelligence due to the spatiotemporal resolution of the product.