Development and Performance Evaluation of Artificial Neural Network Model for Runoff Estimation

Management of water resources in a proper manner is a big challenge nowadays. The scarcity of water in some areas and pollution causes the necessity of it. This can be done by estimating the accurate amount of water available in any watershed area. In this study, a soft computing tool (artificial neural network or ANN) was used to model the relation between rainfall and runoff with only two data sets namely, rainfall and runoff. The area considered for the study is Kelo macro-watershed of Mahanadi basin, Raigarh, Chhattisgarh. The daily rainfall and gauge – discharge data for past 12 years (2002 to 2013) were used. In this study two models on different combinations of data were made. From the results, we see that the model M1 performed better than the other model of ANN. It showed the highest coefficient of correlation (CC) and coefficient of efficiency (CE) values as 87.81% and 84.61% respectively during training and highest CC and CE values as 83.35% and 81.80% % respectively and minimum mean absolute deviation (MAD) and root mean square error (RMSE) values as 7.72 and 8.11% respectively during training and minimum MAD and RMSE values as 8.06 and 8.55% respectively during testing of the model. The results showed that the ANN is a good model for describing the rainfall-runoff relationship in the study area.