Evolution Of Rainfall-runoff Relationship In Mountainous Areas Of The Daqing River Basin And The Analysis Of Driven Factors

The evolution of the rainfall-runoff relationship and the analysis of driven factors are explored from the perspective of time and space in the mountainous areas of the Daqing river basin,which laying the foundation for the research on rainfall-runoff relationship under a changing environment and the sustainable development of water resources.According to the relevant hydrometeorological data collected in the study areas,firstly,this paper analyses the evolution of rainfall-runoff relations in mountainous areas from the aspects of statistical laws,trend,sustainability,and periodicity,and the causes and changes of rainfall and runoff sequences in mountainous areas are analyzed in detail from the decomposition level.Secondly,constructing the encountering probability and a two-dimensional state transition probability matrix of rainfall-runoff in the main sub-watershed by the Archimedean Copula function and Markov correlation theory,the change of abundant and low state about rainfall-runoff relationship are explored.Then using the GAMLSS method and Flow Duration Curve related definitions,the inconsistency of rainfall-runoff relationship are explored.Meanwhile,due to the most intuitive manifestation of the rainfall-runoff relationship and driven factors is the change in the contribution to the runoff,the cumulative slope rate comparison method analyzes the influence of driving factors on runoff and related reasons.Finally,the Copula joint distribution probability is used to quantify the rainfall-runoff relationship,and the spatial variation of the rainfall-runoff relationship in the study areas is established by Geo-statistics.The main conclusions are as follows:(1)The statistics of rainfall and runoff sequences in the mountainous areas of the Daqing river basin are basically similar,and the evolution characteristics of the trend,sustainability,and periodicity of each runoff sequence are consistent.The mountainous areas mainly show small changes in rainfall and a sharp decrease in runoff.It can be clearly seen from the decomposition variable scale that the original sequence is reduced due to the combined effect of the partial distribution of the Res variables and the amplitude variation of each IMF variables,and the IMF variables has a clear period change.Meanwhile,the inflection point of runoff evolution in mountainous areas is 1980 and 2000 by the results of Mann-Kendall test,accumulative curve and related literature.According to the encountering probability and a two-dimensional state transition probability matrix of rainfall-runoff,the similarities and differences of abundant and low state evolution of the main sub-watershed are shown,and the synchronization probability of each sub-watershed is higher than the asynchronous probability,which can reach more than 50%,and the runoff state is more easily transferred to a worse state,which is less connected with rainfall.Based on the change of the synchronization probability and the combined state limit probability of long-term evolution,it can be judged that the sub-watershed is affected by other factors such as human activities from high to low in order: the basin above the Xidayang reservoir,the basin above the Zijingguan station and the basin above the Wangkuai reservoir.(2)According to the GAMLSS method,the main control Hydrological stations in each sub-watershed are taken as the research object to construct the inconsistency model that considers time-varying characteristics.The linear and parabolic linear change of runoff with time are more significant,which directly reflects the runoff trend and magnitude.However,the inconsistency model with rainfall as the covariate can effectively reflect the dynamic change process of runoff in the study areas,which obviously reflects the occurrence time of flood in the study areas.Taking Fuping Station as an example and combining FDC curve related definitions,it is found that the decreasing trend of runoff is more caused by the inconsistency changes of the rainstorm and flood,and the number and magnitude of high flows are significantly reduced,while the change of the rainfall-runoff relationship in the non-flood season is more consistent.(3)Pearson index is used to analyze the influence of meteorological factors.Evaporation factors combined high correlation with rainfall are taken as the main driving factors except rainfall.Three groups of periods are divided by 1980 and 2000,and the cumulative slope rate comparison method determines the influence degree of driving factors in each time period.After 2000,the decrease of rainfall on runoff changed from negative to positive,which was similar to the trend result of MannKendall test.And it is obvious that the comparison of the influence degree of human activities in each sub-watershed is consistent with the results in the rainfall-runoff evolution of the abundant and low state.(4)The Copula function is used to quantify the rainfall-runoff relationship,and the spatial correlation of the rainfall-runoff relationship in the study areas is established by Geo-statistics.The trend is more obvious in the middle,that is,the spatial variation of the basins above the Xidayang reservoir and the basin above the Zijingguan station is significant,and it has certain similarities with the time evolution of the rainfall-runoff relationship.Furthermore,the basin above Wangkuai Reservoir are used as the new study areas,and the sub-watershed have been verified.The method is consistent with the spatial variation of mountainous areas in relative position and Copula joint distribution probability and other aspects.