Impact of Climate Change on Water Levels in the Poyang Lake

The water level change directly relates to scientific decision-making of flood control and drought relief, efficient utilization of water resources, healthy of water ecology and sustainable development of economy in the Poyang Lake basin. Back Propagation Neural Networks Model (BP-NNM) was established based on observed daily temperature, precipitation and water level data series in this study region. The statistical downscale method was coupled with BCC-CSM1-1 global climate model to obtain basin scale air temperature and precipitation data series, and then as inputs of BP-NNM to simulate future water levels of the Poyang Lake. The future average water levels of the Poyang Lake were analyzed and discussed. The results show that the future annual mean water level change will less than 2.5% compared to the baseline period, but the inter-annual variability of water levels among months is large under RCP4.5 scenarios.