Makoanye/Climate-Change-Streamflow-South-Phuthiatsana

Climate change impacts on streamflow and hydrological extremes in the South Phuthiatsana Catchment, Lesotho, using SWAT+, XGBoost and CMIP6 climate projections. # Climate Change Impacts on Streamflow and Hydrological Extremes ## South Phuthiatsana Catchment, Lesotho **MSc Research Project | National University of Lesotho | 2025** This study assessed the impacts of climate change on streamflow dynamics and hydrological extremes in the South Phuthiatsana Catchment, Lesotho, using a hybrid modelling framework combining **SWAT+ hydrological modelling** and **Extreme Gradient Boosting (XGBoost)** with **CMIP6 climate projections**. The research evaluated historical hydro-climatic variability and projected future changes in precipitation, temperature, streamflow, and hydrological extremes under the **SSP2-4.5** and **SSP5-8.5** scenarios. **Keywords:** Hydrological Modelling · Climate Change · XGBoost · SWAT+ · CMIP6 · Streamflow · Hydrological Extremes · Lesotho --- ## Research Context The South Phuthiatsana Catchment is an important water-resource system in Lesotho, with the South Phuthiatsana River serving as a perennial tributary of the Mohokare (Caledon) River within the Orange-Senqu River Basin. Climate variability and long-term climate change can alter precipitation, temperature, streamflow regimes, and the occurrence of hydrological extremes. Understanding these changes is important for supporting long-term water-resources planning and climate adaptation in the catchment. This research therefore investigated historical hydro-climatic variability and assessed potential future changes in streamflow and hydrological extremes under different climate-change scenarios. --- ## Research Objectives The study aimed to: 1. Assess historical trends and variability in precipitation, temperature, and streamflow in the South Phuthiatsana Catchment. 2. Evaluate the performance of the SWAT+ hydrological model in simulating historical streamflow. 3. Develop and evaluate an XGBoost machine-learning model for streamflow simulation. 4. Assess projected changes in precipitation and temperature under the SSP2-4.5 and SSP5-8.5 …