Simulation Of Runoff Variation In Upper Reaches Of Manas River Based On SWAT And SDSM

As the impact of climate change continues to deepen,it is of great significance to study the response of runoff to future meteorological factors,which can provide theoretical basis for the scientific planning of water resources allocation,reservoir scheduling,reservoir construction,and ecological maintenance of related departments.This paper takes the upper Manas River as the research object,and conducts a simulation study of runoff changes based on mathematical statistics and numerical simulation methods.The SWAT distributed hydrological model was established to simulate the runoff in the historical stage;the statistical downscaling SDSM model was established,using the three scenarios under the CanESM2 model RCP2.6,RCP4.5 and RCP8.5 future temperature and precipitation Carry out a simulation and input the obtained future scenario results into the calibrated SWAT model to obtain the response of the Manas River under future climate change.The conclusions obtained in this study are as follows:(1)Analyze the characteristics of precipitation and runoff during the historical period(1956-2008)of the two main hydrological stations in the study area using mathematical analysis methods.The results show that the runoff shows a significant increase trend,with a sudden change in the runoff sequence in 1995,a sudden change in precipitation and evaporation in 1997,and a 20-year and 10-year cycle change in runoff.(2)The SWAT model of the upper Manas River in the 1992-2008 time series was established.The simulated values of the hydrological station are close to the observed values,and the flow process is well fitted.The regular rate and the verification period R2 are greater than 0.75.The annual average runoff of the river is 1.381 billion m~3,the annual average evaporation is 1.5 billion m~3,and the average annual precipitation is 2.8 billion m~3.The evaporation accounts for more than half of the precipitation.(3)The SDSM statistical downscaling model establishes the statistical relationship between the predictors and predictors in the study area.The simulation period is 1961-2005.The simulation results show that the temperature simulation effect is better than the precipitation simulation effect.Input CanESM2 model RCP2.6,RCP4.5 and RCP8.5 three scenarios,through the downscaling model,get the future climate scenario of the river basin.The results show that under the three modes,the overall performance of the study area in the next two periods will be characterized by increasing temperature and humidification.(4)Use the meteorological data in the future climate scenarios to drive the SWAT model to estimate the runoff process of the Manas River.The results of the study show that the average annual runoff from 2030 to 2048 is 1.52 billion m3,1.93 billion m3,and 2.01 billion m3under the three climate scenarios,respectively,an increase of7%,35%,and 42%compared with the baseline period.The average annual runoff from 2080 to 2098 is 1.2 billion m3,1.54 billion m3,and 1.76 billion m3under the three climate scenarios,which are a decrease of 16.7%,an increase of 8.7%,and an increase of 30%compared with the baseline period.The monthly runoff shows a trend of decreasing in summer and increasing in spring and winter.The average annual runoff in 2080-2098 shows a decreasing trend compared with the average annual runoff in 2030-2048.