Assessing the potential value of the regionalised input constraint indices for constraining hydrological model simulations in the Congo River Basin

The regionalised characteristics of a sub-basin's long-term hydrological behaviour are used as multiple constraint filters for constraining hydrological model simulations in the Congo Basin using the monthly time step Pitman model. The results suggest that the constraints are appropriate in many sub-basins (≥ 20 gauging stations), but not all. Detailed examination of these results suggested that the effects of high slopes (> 7%) could increase the runoff ratio and Q90/MMQ (low flow index) constraint values and that implementing an adjustment factor based on slope did improve some of the very poor results. The percentage points on the FDC falling within the simulated uncertainty band has increased from 0% to 59.6% and 2.4% to 39.9% for the Rift valley and Batéké plateaux regions, respectively. Future studies ideally need a range of different rainfall products to quantify the uncertainties related to the inappropriateness of the CRU rainfall data in some parts of the Congo Basin.