Research On Drivers And VLP Prediction Of Runoff Under Changing Environment

Affected by human activities and climate change,the dynamics of hydrological and water resources system in the Yellow River Basin is strengthened.How to analyze runoff distribution characteristics and runoff response under the changing environment,and how to make runoff prediction is of great significance to the comprehensive development and utilization of water resources.In this study,aiming at the variation caused by the change of physical conditions in the watershed under the changing environment,the typical hydrological stations Toudaoguai,Longmen,Sanmenxia,Huayuankou,Gaocun,Lijin of the Yellow River,and Kuye River in the middle reaches of the Yellow River,are selected to study the power law distribution of runoff and the response of runoff in the Kuye River,and to make runoff prediction.The main contents and contributions are shown below:(1)Based on the annual streamflow series of Toudaoguai,Longmen,Sanmenxia,Huayuankou,Gaocun,and Lijin stations located in the Yellow River,the M-K and MWP methods were used to identify change-points.Thereafter,the power law distributions of the daily streamflow series before and after abrupt changes were studied based on the two-parameter power law distribution method.The results reveal that,runoff abrupt changes are 1968,1986 at Toudaoguai,1968,1985 at Longmen,1985,1985,1994 at Sanmenxia,Huayuankou and Gaocun,1968,1985 at Lijin,and the main reasons for the abrupt changes are the operation of large reservoirs and the implementation of large-scale water-soil conservation measures.Except for Longmen and Sanmenxia stations,the daily streamflow series of other stations have power law relationship.The power-law characteristics of the series before abrupt changes are basically same with the original series,and the power law characteristics of the series after abrupt changes disappear gradually.(2)In order to explore the main driving factors of significant runoff change of the Yellow River,Kuye River,a typical watershed in the middle reaches of the Yellow River,was selected as the research object,and the SWAT distributed hydrological model was established to analyze the contribution of climate change and human activities to runoff change and study suggestions for water and soil conservation measures.The results reveal that,the contribution rate of climate change and human activities to runoff change was 32% and 68%,respectively.For water-soil conservation measures,For afforestation,grassing,terrace 5 °-15 °,terrace 15 °-25 ° and check dam,the runoff are decreased by 1.17%,0.63%,2.93%,0.1% and 11.4%,respectively.For terrace 15 °-25 °,there is almost no peak cutting and water reducing benefit.(3)In order to improve the accuracy of runoff prediction under the condition of variation,a variational mode decomposion-long and short-term memory network hybrid model VLP(VMD-LSTM-PSO)was proposed to predict 1-5 days ahead daily streamflow series of four hydrological stations on the Yellow River.Furthermore,the model was compared with LSTM model,complementary empirical mode decomposition(CEEMD)based LSTM model and wavelet decomposition(WT)based LSTM model.The results reveal that VLP model has high runoff prediction accuracy,with NSE up to 0.9 ? 0.99.It has strong adaptability,good stability,and ability to accurately describe the non-linear and non-stationary of hydrological series.