Groundwater Modeling in Kadiridimba Catchment

Sammanfattning

Rapid population growth, climate variability, and limited hydrogeological data create uncertainty for groundwater decisions in Rwanda. Kadiridimba catchment was selected as a case study to develop a groundwater model, test what can be concluded under these constraints, and identify the measurements that most strengthen confidence in the results and assumptions. The aim was to develop and evaluate a steady state groundwater model as a practical step toward elevating the knowledge in the region, not to deliver final predictions. A clear GIS to FEFLOW workflow was applied. National datasets were combined with a small set of borehole logs and targeted field checks. Eleven steady-state simulations explored model sensitivity by varying hydraulic conductivity, recharge, river and seepage conductance, boundary heads, and deep lateral in and outflow. These trials showed that the catchment water balance can be brought close to equilibrium, but simulated groundwater levels still differ substantially from observations because the present model structure does not yet capture the real system. This indicates that uncertainty in geometry and layer continuity is larger than uncertainty in local parameter values, and that calibration is at this stage more preferable in a focused and well-monitored subcatchment than for the entire catchment. The study contributes a transparent and repeatable workflow that makes data gaps visible and sets priorities for improvement. Recommended actions are to centralise and quality control existing datasets, expand continuous groundwater level monitoring and borehole logging, better constrain recharge and boundary conditions including possible cross basin exchange.