What Controls the Runoff Generation in River Basins?

Accurate streamflow estimation is essential in several water resources management applications. Therefore, hydrologists aim to develop a simple, robust and calibration-free rainfall–runoff model. In a typical rainfall–runoff model, soil is at the centre and is assumed to play an important role in runoff generation. However, the recently proposed Dynamic Budyko (DBv1) model predicts daily streamflow using only climatic inputs. Furthermore, the DBv1 model framework was recently modified for effective rainfall simulation. The study reports that meteorological inputs contain sufficient information for runoff prediction. However, the modified DBv1 (DBv2) model has not been tested for peninsular river basins in India, which is crucial for water resources management applications. Therefore, in this study, three basins from India were selected as the study area. The DBv2’s runoff generation module was used for effective rainfall estimation, and linear reservoirs (LR) connected in parallel were used to obtain streamflow at the basin outlet. Thus, this model (DBv2LR) has a calibration-free runoff generation module, while routing requires calibration. We compared the performance of DBv2LR with the well-known Probability Distributed Model (PDM), which is fully calibrated. In the case of PDM, the same routing module (reservoirs connected in parallel) was used to obtain streamflow at the basin outlet (PDMLR). The models, DBv2LR and PDMLR, were evaluated for Nash–Sutcliffe efficiency (NSE). The results indicate that the DBv2LR model shows performance very similar to the PDMLR. Overall, this study suggests that even in arid or semi-arid regions like India, climate information is sufficient for effective rainfall estimation.