Influence of Climate Model Selection on Long-Term Groundwater Recharge Forecasts across different African Regions

Abstract

The continuous availability and renewability of groundwater rely on the process of groundwater recharge. Therefore, recharge forecasting is essential for effective groundwater resource management, particularly in the context of climate change. In this study, we employ the Community Land Model version 5 to forecast recharge across Africa for the period 2071-2100 at 10km spatial resolution. The land module was forced with outputs from three regional climate model outputs (CCLM5, RegCM4 and REMO2015), each driven by two global climate models (MPI_ESM and NorESM) under two climate scenarios (RCP2.6 and RCP8.5), and recharge estimated using the water balance approach at both continental and regional scales. Based on the long-term recharge forecast, continental average recharge potential of 119 mm/year (with standard deviation of 68 mm/year) and 92 mm/year (with standard deviation of 59 mm/year) for RCP2.6 and RCP8.5 were recorded respectively. The standard deviation serves as an indicator of spatial variability across the models’ ensemble. Further analysis, including correlation, coefficient of variation and bias were used to assess regional recharge reliability and sensitivity, respectively. Model performance was found to be region specific, with significant differences in biases between CCLM5 and REMO2015, while REGCM4 demonstrated consistent pattern across most regions and both climate scenarios. The results indicate recharge projections are more influenced by model structures and the choice of driving GCM than emission scenarios. These structural differences and uncertainties highlight the complex interactions between global and regional climate processes that influence recharge projections. Such uncertainties present challenges for regional development and climate adaptation strategies. Therefore, this study recommends evaluating the performance of individual models within ensemble frameworks and highlights the importance of local and regional calibrations to enhance the reliability of groundwater recharge projections.