Geomorphology-based genetic programming approach for rainfall–runoff modeling

Nowadays, artificial intelligence approaches such as artificial neural network (ANN) as a self-learn non-linear simulator and genetic programming (GP) as a tool for function approximations are widely used for rainfall–runoff modeling. Both approaches are usually created based on temporal characteristics of the process. Hence, the motivation to present a comprehensive model which also employs the watershed geomorphological features as spatial data. In this paper, two different scenarios, separated and integrated geomorphological GP (GGP) modeling based on observed time series and spatially varying geomorphological parameters, were presented for rainfall–runoff modeling of the Eel River watershed. In the first scenario, the model could present a good insight into the watershed hydrologic operation via GGP formulation. In the second scenario, an integrated model was proposed to predict runoff in stations with lack of data or any point within the watershed due to employing the spatially variable geomorphic parameters and rainfall time series of the sub-basins as the inputs. This ability of the integrated model for the spatiotemporal modeling of the process was examined through the cross-validation technique. The results of this research demonstrate the efficiency of the proposed approaches due to taking advantage of geomorphological features of the watershed.