Characteristics And Trend Forecast Weihe River Runoff Change Under Climate Change

Studies showed that runoff changes is mainly determined by the atmospheric precipitation, however, precipitation needs certain conditions. So the effect of climate change on runoff is of great importance. The effects of climate change on runoff was studied and the amount of runoff was predicted by using the neural network model in Weihe River basin of Shaanxi section.The study of the effects of climate change on runoff is beneficial to the water resources planning and management of Weihe River basin in Shaanxi section, at the same time, it has important reference function to the regional economic development.Mann-Kendall method, the wavelet analysis and other research methods were used to analysis the correlation between the air temperature and runoff, precipitation and runoff of Weihe River basin in Shaanxi section according to the research between1959and2009in the area of Weihe River and meteorological data and data from main tributaries. Finally the developing trend of runoff was predicted according to the building of BP neural network forecasting model and the meteorological data as input conditions.The main conclusions are as follows:(1) The change of time series characteristics as well as their mutation point of air temperature, precipitation and runoff was revealed in research areas based on M-K analysis method.(2) The period of the air temperature, precipitation and runoff change were studied between1959and2009based on the method of wavelet analysis. The results showed that the temperature cycle is about35years obviously, the precipitation is about32years and runoff cycle of each site is different. But basically change cycle is between32to42years.(3) The run off changes were affected by the factors such as climate change, hydrothermal conditions and other factors. In this paper, the impact of climatic conditions on runoff was reflected by using the correlation of the temperature and runoff, precipitation and runoff. The runoff and rainfall are of positive correlation of all sites in the area, the temperature is negative correlation, and the correlation between rainfall and runoff is greater than the temperature and runoff.4) BP neural network prediction model was established with the temperature and precipitation as input variable. The results showed that the predicted values and the measured values of correlation coefficient R2is0.8078and the prediction accuracy is higher by the forecast of runoff. It shows that this model can preliminary forecast climate change of the Shaanxi section of Weihe River basin runoff under the condition of recent trends.