Font Size: a A A

Projected Changes In Hydrological Drought Based On Multi-model Ensemble In The Yuanjiang River Basin

Posted on:2022-07-02Degree:MasterType:Thesis
Country:ChinaCandidate:C Y LiuFull Text:PDF
GTID:2480306335958109Subject:Geophysics
Abstract/Summary:PDF Full Text Request
The Yuanjiang-Honghe River is an important international river in southwest of China and Southeast Asia,and its hydrological condition is related to the economic and cultural cooperation,agricultural irrigation,shipping,fishery,and residents' production and life in the basin countries.The change characteristics of hydrological drought under future climate change scenarios in the Yuanjiang River Basin are studied in this paper,which can provide a scientific reference for water resources management and prevention of hydrological drought disasters in the basin,and provide a case for indepth research on hydrological process changes in the mountainous basin of the Southwest Plateau.This paper takes the Yuanjiang River Basin as the research object,based on the five climate model data in ISI-MIP 2a,the observed meteorological data and the runoff of the hydrological station,and then uses the Bayesian model averaging method and the Delta method to carry out multi-model ensemble and bias correction.Then use the ensembles and GR4 J hydrological model to simulate future runoff,and use the standardized runoff index(SRI)to extract hydrological drought information and analyze its changes.The main conclusions obtained are as follows:(1)The frequency of hydrological droughts in the 1960 s,1980s and 2010 s was higher than that of other years.The frequency of light drought,moderate drought,severe drought and extreme drought were 33.14%,9.05%,4.53% and 3.07%respectively,and the overall was 49.78%.The drought duration was 6,5,7 and 13 months,the drought severity was 3.95,2.90,2.74 and 6.19,and the drought intensity was 0.59,0.55,0.55 and 0.39 at different time scales,respectively.The onset month of drought was generally from October to December,and the end month was mainly in May.(2)The accuracy of the climate model was significantly improved after ensemble and bias correction.For precipitation,the coefficient of determination(CC)increased from 0.74-0.78 to 0.84(monthly scale),and from 0.24-0.26 to 0.38(daily scale);the root mean square error(RMSE)decreased from 54.65-66.23 to 47.34(monthly scale),and from 7.05-7.40 to 5.59 in daily scale.For temperature,CC increased from 0.91-0.94 to 0.95(monthly scale),from 0.82-0.85 to 0.89(daily scale);RMSE decreased from 2.10-2.50 to 1.98(monthly scale)and 3.06-3.43 to 2.66(daily scale).(3)The analysis of future climate interannual changes shows that,compared to the baseline period,the temperature will increase respectively by 1.11 ?,2.01 ? and 2.62 ?in the near-period,mid-period,and long-period of the RCP4.5 scenario,and will increase respectively by 1.20 ?,2.73 ? and 4.60 ? under the RCP4.5 scenario.The increase in RCP8.5 scenario was greater than that in RCP4.5 scenario,and the longperiod > mid-period > near-period.The precipitation in the RCP4.5 scenario changes-2.16%,1.58%,5.24%,respectively,and changes-5.23%,1.96%,4.58% under the RCP8.5 scenario,respectively.The range of change of RCP8.5 is greater than RCP4.5scenario,and all of them are decline in the near-period,a slight increase in the midperiod,and a significant increase in the long-period.The monthly average temperature will increase in three period,but there will be differences in the increase of different months under different scenarios in different period,with the largest increase in March,April,and May.The monthly precipitation is declining mainly in the near-period and mid-period,with the largest decrease in January and February,and are mainly increasing in the long-period,with the largest increase in September and October.(4)The hydrological simulation results show that the NSE are 0.81 and 0.75 in the calibration and validataion period,respectively,the CC is 0.93 and 0.94,the PBIAS is-0.82% in the calibration period and-8.56% in the validation period,indicating that the simulated runoff has well consistency with the observed runoff.The annual runoff is decrease dominantly.Under the RCP4.5 scenario,the runoff changes in the near-period,mid-period,and long-period are-8.71%,-2.63%,5.90% respectively,and are-18.1%,-4.33%,-6.71 under the RCP8.5 scenario,respectively.The dry season runoff increased by 14.79%,14.98%,27.83%(RCP4.5 scenario)and 2.90%,13.52%,8.74%(RCP8.5scenario).The rainy season runoff decreases by-16.55%,-8.50%,-1.42%(RCP4.5scenario)and-25.10%,-10.28%,-11.86%(RCP8.5 scenario),and its change characteristics are related to annual runoff.(5)Under the future RCP4.5 and RCP8.5 scenarios,take the SRI-6 for example,the frequency of light droughts and extreme droughts will decrease by about 3% and2%,respectively,while the frequency of moderate and severe droughts will increase,with an increase of approximately 2%,1%.The drought duration will prolong,ranging from 7 to 9.5 months,the drought severity will increase,ranging from 3.96 to 6.93,and the drought intensity also will increase,ranging from 0.57 to 0.66.The onset of drought is delayed by 1 to 2 months,and the end of drought is delayed by about 1 month.
Keywords/Search Tags:Hydrological drought, Climate model ensemble, Bias correction, Runoff simulation, Yuanjiang River Basin
PDF Full Text Request
Related items