Effect Of Extreme Climate Change And Human Activity On Sediment Load In Songhua River Basin

Studying the effects of precipitation extremes and human activity on sediment load can not only reveal the mechanism of sediment load but also provide important scientific information for basin management and disaster warning.This thesis analyzed spatial and temporal variation characteristics of climate extremes and their association with atmospheric circulation patterns based on meteorological and hydrological observation collected from 60 meteorological stations in the Songhua River Basin(SRB).The variation characteristics of sediment load in the SRB and its response to extreme precipitation indices were analyzed.The impact of human activity on sediment load in the Second Songhua River basin(SSRB)was evaluated by the SWAT model.The trends and variability of extreme climate events under RCP4.5 and RCP8.5 emission scenarios during 2020-2099 were evaluated on the basis of a regional climate model RegCM4.0 nested by a global climate model BCC-CSM1.1.Runoff and sediment load at RCP4.5 and RCP8.5 emission scenarios were simulated although embedding the estimated climatic data into SWAT model.The main conclusions were as follows:(1)Warm indices,including the number of warm nights,warm days and summer days,showed significant increasing trends(P<0.01)with corresponding rates of change of 2.1,3.8 and 2.6d/10 a,respectively.Eextreme indices,including minimum TN and maximum TX,showed increasing trends with corresponding rates of change of 0.2 ? and 0.7 ? /10 a,respectively.On the other hand,cold indices,including the number of cold nights,cold days and frost days,showed decreasing trends with corresponding rates of change of-1.4,-4.1 and-3.7 d/10 a,respectively;Warm indices and maximum TX showed significant positive correlations with latitude(P<0.01);The Arctic Oscillation index(AO)displayed significant negative correlations with the cold indices(P<0.01)and positive correlations with the warm indices.The warm indices and extreme indices had positive correlations with the Northern Hemisphere Subtropical High area and intensity indices,while the reverse relationship between the cold indices and Northern Hemisphere Subtropical High.The Asia polar vortex area and intensity indices showed negative correlations with warm indices and extreme indices,while they were positive to cold indices.The multivariate ENSO Index(MEI)showed no linear correlation with any of the temperature extremes.(2)Precipitation extreme indices,including wet-day precipitation(PRCPTOT),flood season precipitation(FSPTOT),erosive precipitation(ERPTOT),very wet day precipitation(R95pTOT),maximum 1-day precipitation(RX1day)and heavy precipitation(HP),did not show significant change trends during 1960-2014.The western Pacific subtropical high index was significantly correlated with wet-day precipitation(PRCPTOT),erosive precipitation(ERPTOT)and very wet day precipitation(R95pTOT).(3)2.5 ? was the optimum threshold temperature for discriminating snowfall at the scale of the entire basin.Due to differences in the landscape conditions at the different stations,the optimum threshold varied by station.Compared with using a single suitable temperature threshold to discriminate snowfall throughout the basin,it was more accurate to use the optimum threshold at each station to estimate snowfall in the basin.In addition,snowfall was underestimated when the temperature threshold was the WBT and when the temperature threshold was below 2.5 ?,whereas snowfall was overestimated when the temperature threshold exceeded 4.0 ? at most stations.A suite of extreme snowfall indices including four comprehensive indices,four intensity indices,four grade indices and two date indices were identified to assess extreme snowfall events.The total snowfall(SNTOT),the number of snowfall days(SND),the ratio of snowfall to total precipitation(S/P),the snowfall intensity(SNI),and the amounts of extreme snowfall(SN95TOT)and very extreme snowfall(SN99TOT)display insignificant trends over the entire SRB.The changes in the ending snowfall date(ESD)exhibit a significant advancing trend(P<0.001,2.5 d/10a),while the changes in beginning snowfall date(BSD)display a significant delaying trend(P<0.05,1.3 d/10a),which have led to a reduced snowfall season length(SSL)(P<0.001,-4.1 d/10a).The indices related to snowfall duration(SND and SSL)and snowfall date(BSD and ESD)show significant correlations with atmospheric circulation patterns.(4)Sediment load occurs in a few extreme flood events in the SRB and SSRB,and 88.6% present of sediment load may concentrate on one flood event.Sediment load in SSRB showed the significantly decreasing trend(P<0.01),while sediment load in SRB extremes did not have.Precipitation extremes showed a significant correlation with sediment load(P<0.05),and the inter-annual fluctuation of sediment load was mainly influenced by precipitation extremes changes.(5)The abrupt change point between sediment load and each extreme precipitation index occurred in 1977 in the double mass curve method.Compared with the baseline period in 1960-1977,the effects of precipitation extreme indices on the sediment load ranged from-23.43% to 9.29%.Change in precipitation was not the main reason for the long sediment load variation.Reservoir construction and water diversion might be the main reason for the sediment load long-term variation.(6)The effects of human activities on sediment load in the SSRB were evaluated by SWAT model,and SWAT model can well simulate runoff and sediment load in the basin.Compared with 1960-1977,human activities decreased 211,000 t of sediment load during 1978-2014,accounting for 57.02% of the natural sediment load.Reservoir construction and water diversion were the main reason for the sediment load reduction in SSRB.(7)Based on the observed daily temperature and precipitation data collected from 60 stations in SRB,the simulated capability of a regional climate model RegCM4.0 one-way nested by the global climate model BCC_CSM1.1 was examined.RegCM4.0 can well simulate the trend of temperature extremes and precipitation extremes at a certain extent.The simulation of temperature extremes was better than that of precipitation extremes.Moreover,summer days(SU),frost days(FD)and minimum temperature(TNn)were underestimated,and wet-day precipitation(PRCPTOT)and flood season precipitation(FSPTOT)were overestimated(Re=82.74% and 61.2%).In RCP4.5 and RCP8.5 scenario,the extreme warm indices and extreme indices will be increased,cold indices will be decreased in the 2020-2099,on the other hand,precipitation extremes will be increased in the scenarios,and it was more significant in RCP8.5 scenario(P<0.05)than RCP4.5.(8)Based on the adopted SWAT model,runoff and sediment load were simulated in the climate scenarios.Runoff shows the insignificant increasing trend,while sediment transport increased significantly,which were more significant in RCP8.5 scenario(P<0.05)than RCP4.5 scenario.