Dialy rainfall runoff modeling using artificial neural network

The application of Artificial Neural Network (ANN) methodology for modeling daily flows during monsoon flood events for a large size catchment of river Krishna in Karnataka State of India is presented. A linear multiple input single output (MISO) model is employed for determination of the response function of the catchment. The spatial variation of rainfall is accounted for by subdividing the catchment into two parts and treating each of the sub catchment as separate lumped input to the model. The ANN is applied for each input scenario above with different input combinations. The results of this analysis show that performance of ANN with previous day's runoff along with current and antecedent rainfall as input is best in term of Nash-Sutcliffe efficiency (R-2). The conclusion can be drawn that the ANN model can be successfully used for modeling of daily flows in a large size catchment, which has a highly non-linear rainfall runoff relationship.