Spatiotemporal characterization of drought in Botswana: a regionalization study using long-term chirps satellite data

As the effects of climate change continue to intensify, drought trend analysis is becoming more common. Despite this, Botswana still lacks drought prediction and monitoring tools that are locally adopted. This study assesses the temporal and spatial aspects of drought in Botswana and their correlation to synoptic-scale climatic drivers using Climate Hazards Group Infrared Precipitation with Stations (CHIRPS) gridded satellite data from 1981 to 2024 and the Standardized Precipitation Index (SPI) at a 12-month timescale. Principal Component Analysis (PCA) and the K-means clustering technique were applied in the study to delineate Botswana into homogenous drought regions, while the Run Theory (RT) technique was applied to reveal drought characteristics within each distinct drought region. The Rescaled Range (R/S) test, together with the Mann-Kendall (MK) method, was used to examine drought variability and trends, while the Cross Correlation, Cross Wavelet, and Fast Fourier Transform (FFT) techniques were applied to explore the possible influences of synoptic-scale climatic processes as drought driving factors in Botswana. Regions 3 and 8 in the northern part of the country were determined to be more vulnerable to drought, characterized by extreme drought events that lasted more than 20 months with severities as high as 40. These regions are dependent on wildlife and tourism, and as such, extreme droughts pose a threat to their economic stability. This study further found that the Niño 3.4 and Southern Oscillation Index (SOI) indices are key drivers of drought variability in Botswana and therefore concludes that these climatic indices can serve as reliable drought forecasting and prediction tools when used in prediction models, thereby improving the capacity of early warning systems in Botswana.