ANN-SCS-based hybrid model in conjunction with GCM to evaluate the impact of climate change on the flow scenario of the River Subansiri

Abstract

Abstract The River Subansiri, one of the largest tributaries of the Brahmaputra, makes a significant contribution towards the discharge at its confluence with the Brahmaputra. This study aims to investigate an appropriate model to predict the future flow scenario of the river Subansiri. Two models have been developed. The first model is an artificial neural network (ANN)-based rainfall-runoff model where rainfall has been considered as the input. The future rainfall of the basin is calculated using a multiple non-linear regression-based statistical downscaling technique. The proposed second model is a hybrid model developed using ANN and the Soil Conservation Service (SCS) curve number (CN) method. In this model, both rainfall and land use/land cover have been incorporated as the inputs. The ANN models were run using time series analysis and the method selected is the non-linear autoregressive model with exogenous inputs. Using Sen's slope values, the future trend of rainfall and runoff over the basin have been analyzed. The results showed that the hybrid model outperformed the simple ANN model. The ANN-SCS-based hybrid model has been run for different land use/land cover scenarios to study the future flow scenario of the River Subansiri.