Surface runoff modeling using artificial neural network in alpine watersheds of Tianshan Mountains, Northwestern China

This paper attempts to investigate the impacts of precipitation and temperature on surface runoff in poorly gauged alpine basins originating from alpine Tianshan Mountains. The three-layered feed forward artificial neural networks (ANNs) with a back-propagation algorithm have been proposed to model runoff generation trained using four structures of (5, n, 1)1, (5, n, 1)2, (6, n, 1) and (8, n, 1). The results indicate that ANN models trained using the four structures performed equally well in both Kaidu and Manas River Basins.