A Quantitative Assessment of Drought Trends in Uganda Using Statistical Models for Hydrological Indicators
Genesis Magara
School of Ecology and Applied Meteorology, Nanjing University of Science and Technology, Nanjing, 210044, China.
Abraham Okrah *
School of Ecology and Applied Meteorology, Nanjing University of Science and Technology, Nanjing, 210044, China.
Yawlui Ignatius Senyo Yao
Department of Nanoscience, University of North Carolina, Greensboro,1400 Spring Garden Street, Greensboro, USA.
Emmanuel Yeboah
School of Remote Sensing and Geomatics Engineering, Nanjing University of Information Science and Technology, Nanjing, China.
George Darko
Research Institute of History of Science and Technology, Nanjing University of Information Science & Technology, Nanjing 210044, China.
Diane Akimana
School of Ecology and Applied Meteorology, Nanjing University of Science and Technology, Nanjing, 210044, China.
Vincent Awuku
School of Environment Science & Engineering, Nanjing University of Information Science & Technology, Nanjing 210044, China.
Ishmeal Quist
Research Institute of History of Science and Technology, Nanjing University of Information Science & Technology, Nanjing 210044, China.
Isaac Sarfo
College of Geography and Environmental Science, Henan University, Kaifeng, 475004, Henan, Province, China.
*Author to whom correspondence should be addressed.
Abstract
This study investigates drought patterns in Uganda's Eastern (E), Northeastern (NE), Northwestern (NW), and Southwestern (SW) regions, focusing on the frequency, intensity, and duration of drought events and their impacts on water resources, agriculture, and livelihoods. Using data from ERA5, GPCC, and CRU TS (1960–2020), the study applies the Standardized Precipitation Index (SPI), Standardized Precipitation-Evapotranspiration Index (SPEI), and Standardized Runoff Index (SRI) to analyze drought trends. Ordinary Least Squares (OLS) regression and random forest models identify key predictors of drought events. Results show that the E and NE regions experience more frequent but shorter droughts, while the NW and SW regions face prolonged and intense droughts. SPI and Potential Evapotranspiration (PET) are key predictors of SPEI, with PET consistently influencing drought severity. The random forest model showed moderate predictive accuracy (MSE = 0.29, R² = 0.35) but struggled to predict SPI in the NW and SW. The study highlights the need for region-specific interventions and improved drought prediction models, integrating remote sensing, land-use data, and machine learning. Policy recommendations include enhancing climate data integration for water resource management and strengthening regional drought response systems.
Keywords: Climate resilience, drought variability, hydrological modelling, SPI, SPEI