Future projections of streamflow in the Songhuajiang River basin based on the artificial neural networks

Abstract

Based on observed climate data and climate projection in the Songhuajiang River basin, streamflows at Jiamusi hydrological station under three emission scenarios during 2011–2050 are projected by applying artificial neural networks. The results show that annual streamflow will not change significantly and its decadal variations are also small, but there are significant seasonal variations under three emission scenarios.