Runoff Variation And Its Attribution Analysis In The Central Loess Plateau Under The Time-varying Budyko Framework

Affected by climate change and high-intensity human activities,water cycle processes have undergone significant changes,and the annual runoff series of many basins show unstable characteristics.The Budyko framework has always been a hotspot in eco-hydrological research due to its solid theoretical foundation and clear physical meaning.However,traditional analysis based on the Budyko framework is mostly carried out under the assumption that the coupled hydrothermal parameter remains unchanged.The analysis based on time-varying parameter is limited,which affects the rationality for analyzing the response mechanism of annual runoff.This is not conducive to the scientific decision-making for water resources management.Based on this condition,this dissertation takes 36 basins in the middle Loess Plateau as the research area,and builds a time-varying Budyko framework that considers the influence of multiple factors on the basis of in-depth analysis of spatio-temporal evolution of meteorological,hydrological,vegetation,land use/cover and socio-economic factors.By using this time-varying Budyko framework,this dissertation quantitatively evaluates the impact of climate change and human activities on runoff at different time scales,and scientifically explores the response mechanism of runoff to different types of human activities.Based on the CMIP5 multi-model data,the climate change under the scenarios of 1.5? and 2? global warming in the future is comprehensively predicted,and then the response of runoff to climate change is quantified by using the Budyko framework.Meanwhile,the Copula function is used to construct the joint distribution of climate factors and runoff to explore the trend of runoff changes with probability under different climatic conditions in the future.The results are of great significance for comprehensively understanding the runoff generation mechanisms in the middle Loess Plateau and its response mechanism to changing environment,and can provide decision basis for water resources management and water ecological protection in this region.The work and conclusions of this dissertation are as follows:(1)Spatio-temporal evolution of climatic,vegetation,land use and socio-economic factors in the research area.The precipitation presents a distribution pattern of high southwest and low northeast.The potential evapotranspiration presents a distribution pattern of northeast and low southwest.In most basins,the precipitation and potential evapotranspiration show decreasing and increasing trends during the study period,respectively.The runoff is high in the southwest and low in the northeast,and shows a significant decreasing trend during the study period.The proportion of cultivated land in most basins shows a decreasing trend,while the proportion of forest land,grassland,shrub land and the leaf area index(LAI)show significant increasing trend.The socio-economic factors present an overall distribution pattern of high south and low north,among which the population,GDP and effective irrigation area show increasing trend in all the basins,while the water consumption for irrigation and industry show increasing trend in most basins,but decreasing trend in some southern basins.(2)Construct a time-varying Budyko framework considering the influence of multifactors.The impacts of multi-factors(precipitation,potential evapotranspiration,LAI,proportion of cultivated land,proportion of construction land,effective irrigation area,population,GDP,irrigation water consumption and industrial water consumption)on the integrated Budyko hydrothermal parameter n are considered,and the relationships between these factors and parameter n are described by optimizing function expression forms(linear,exponential or logarithmic).Then,a time-varying Budyko framework is constructed with the consideration of multicollinearity.The results show that the proportion of cultivated land,GDP,industrial water consumption and irrigation water consumption have significant impacts on hydrothermal conditions of basins.Compared to the constant-parameter model,the time-varying Budyko framework considering multifactor influence significantly improves the accuracy of annual runoff simulation.(3)Attribution analysis of runoff change based on the time-varying Budyko framework.The impacts of climate change and different forms of human activities on runoff changes under different time scales are quantitatively assessed by using three methods(decomposition method,elasticity coefficient method and complementary method).The results show that compared to climate change,the contribution of human activity on runoff decrease is greater,among which the impacts of industrial and irrigation water consumption,population and GDP increment are most significant,followed by the decreasing cultivated land and increasing vegetation cover.The attribution results of annual runoff changes in typical basins show that the impact of human activities is slightly lower than or equal to the impact of climate change in the early period,while significantly higher in the later period.(4)Based on the CMIP5 multi-model data,the impact with probability of climatic factors on runoff is estimated under the 1.5? and 2? global warming scenarios.The time-varying Budyko framework is used to quantitatively assess the impact of future climate change on runoff,and the Copula function is used to explore the probability of runoff changes under changing climate conditions in the future.The results show that under the scenario of global warming of 1.5?,most of the basins show a warming and drying trend,while under the scenario of global warming of 2?,all basins show a warming and wetting trend,but the magnitude of the trend is insignificant.Under the scenarios of global warming of 1.5? and 2?,the combined effects of climate change(precipitation and potential evapotranspiration)will reduce runoff by 10.6%?37.7%and 0.8%?30.7%,respectively.The results of probability analysis show that the risk of runoff reduction may increase due to the impact of climate change under the scenario of global warming of 1.5? compared to that of 2?.