Drought And Flood Outlook For The Yangtze River Basin

The Yangtze River Basin?YRB?is the largest river basin in China,which covers19%of China's territory,accommodates 33%of China's population,stores 36%national water resources and contributes 40%of China's total GDP.However,under the intense influence of extreme climate events,water resources in the YRB have become more fragile,and extreme climate events became more frequent in recent years.The extreme turbulence of water in the YRB could affect the ecological balance,economic development,public stability of the YRB,and even worse,affect the national prosperity.This study focuses on water resources assessment and outlook for the YRB.A land-atmosphere coupling modeling and prediction system was constructed to help evaluate and predict how the drought and flood would unfold in both space and time in the YRB during the 21^st century.The main efforts and conclusions of this study are as follows:?1?This study detected the historical trends of both the meteorological and hydrological variables during the period 1956-2013.First,the precipitation and temperature for the period 1956-2013 from 148 meteorological stations were detected.The results show that the daily minimum temperature and daily maximum temperature were increased with 0.021°C and 0.013°C each year,respectively.While the precipitation decreased with the rate of 0.36 mm each year.Second,the alteration trend of monthly streamflow during 1956-2013 from 18 hydrological stations was detected,the results show that only the streamflow from three stations,namely,Gaochang,Wusheng,and Chenglingji,showed a downward trend with-5.969 m³/s,-5.005 m³/s and-45.267 m³/s,respectively.The remaining 15 hydrological stations did not show a statistically significant trend.Third,the daily streamflows during 1990-2010 from 18hydrological stations were also detected.The results showed that Stations Panzhihua,Huatan,Pingshan,Wusheng,Beibei,Cuntan,Huanglingmiao and Shashi were increased in trend,Stations Zhutuo and Wulong showed no significant trend,while Stations Shigu,Gaochang,Wanxian,Yichang,Chenglingji,Luoshan,Hankou,and Datong showed a downward trend.Fourth,the detection results for abrupt change of streamflow during the period 1990-2010 from 18 hydrological stations showed that only two stations,Huatan and Chenglingji,showed statistically significant abrupt change,which indicated that the influence of human activities did not cause significant pattern change to the water resources in the YRB.These results offer background knowledge for further understanding and properly interpreting the future water resources alterations.?2?This study evaluated the performance of 10 CMIP5 models in modeling the historical precipitation and temperature for the YRB,and meanwhile,a new method for statistical downscaling was proposed.First,the comparison between the modeling results of CMIP5 models and historical observations showed that the ability of CMIP5in representing the historical climate was weak,and the bias was substantial.Taking the YRB as a whole,the simulated precipitation from the CMIP5 models of interest was higher than the observations,while the simulated temperature was lower than the observations.Second,the evaluation result for the proposed downscaling method was satisfying,which improved the mean,standard deviation,error range and precipitation statistics,and it can also preserve the long-term trend and enhance the representation of the extreme values.Third,the predicted climate variables for the YRB during 2006-2100 based on the downscaled CMIP5 showed that precipitation,maximum temperature and minimum temperature in the YRB were all projected to increase during2006-2100.The increasing rate was higher under the RCP 8.5 scenario than under the RCP 4.5 scenario.During the 95 years from 2006-2100,daily precipitations were anticipated to increase 0.35 mm and 0.17 mm under the RCP 4.5 and RCP 8.5 scenarios,respectively.Similarly,the maximum temperatures were anticipated to increase 2.89°C and 5.70°C under the RCP 4.5 and RCP 8.5 scenarios,respectively.The minimum temperatures were anticipated to increase 2.54°C and 5.20°C under the RCP 4.5 and RCP 8.5 scenarios,respectively.?3?We built the Soil and Water Assessment Tool?SWAT?for the YRB to simulate the rainfall-runoff process in the YRB and to predict the future water resources both in time and space.First,the calibration and validation results based on daily streamflow from 18 hydrological stations showed that the mean?variance?of R² for the calibration and validation period were 0.76?0.023?and 0.67?0.031?respectively,while the Nash–Sutcliffe model efficiency coefficient?NS?for the calibration and validation period were 0.67?0.031?and 0.68?0.017?,respectively.These results indicated that the SWAT model was capable of simulating the water cycling characteristics of the study area,which satisfied our further research needs.Second,we compared the simulation results between the CMIP5 simulated climate and observations,and the results turned out to be similar.We further carried out the prediction for runoff,soil water content and streamflow in the station Datong for the period 2006-2100,and all the three variables were projected to increase in the future period.Setting up this hydrological model for the YRB provided a reliable technical approach for understanding how the water resources response to the climate change in the long-term.?4?We carried out the prediction of meteorological drought,hydrological drought,and agricultural drought during the late 80 years of the 21^st century for the YRB.First,the evaluation result for drought intensity,severity,frequency,and duration showed that the future drought intensity exhibited great transition compared to the historical period,low-intensity drought was anticipated to decrease,while extreme and exceptional droughts were anticipated to increase.In the future period,the severity of meteorological drought and hydrological drought were projected with a downward trend,on the contrary,agricultural drought was projected with an upward trend,which was more significant under the RCP 8.5 scenario than the RCP 4.5 scenario.The headwater area and the area surrounded the intersection of Sichuan,Guizhou and Chongqing showed an obvious upward trend.During the coming decades till the end of this century,drought frequency and duration showed no obvious trend alteration.But the future drought duration would prolong compared to the historical period.Second,the capability of precipitation,runoff and soil moisture contents in detecting drought is different.Precipitation is more sensitive in detecting mild drought than runoff and soil moisture content;besides it can detect drought early on.In comparison,runoff and soil water content are more reliable in detecting drought duration than precipitation.In addition,soil water content takes the evapotranspiration into consideration,which can be more effective in describing the warming climate.These findings improve our understanding of the relationships of different drought descriptors in describing drought and offer practical guidance for the formulation and implementation of drought mitigation measures.Third,we argued that incorporating CMIP5 models to detect drought are model dependent,so the prediction results based on a The remaining 15hydrological stations did not show single model may not reveal the great uncertainties of future climate change.Multiple scenarios and multimodel ensemble provide different directions for future drought development,which is more reliable.?5?Utilizing the non-stationary flood frequency analysis framework,we assessed the flood peak,flood frequency,return periods and return levels for the 125 major channels in the YRB.The result indicated that flood peaks and flood frequency all showed an increasing or non-significant trend under the RCP 4.5 and RCP 8.5 scenarios,which meant the future flood peaks and frequency would not drop lower than the historical level,and most of the channels would rise.The enhanced flood peaks would pose a great threat to the safety of existing engineering projects.Second,the results of return level with the return periods of 20-,50-and 100-year showed that the return level of headwater area was relatively lower than the middle and eastern area,with the main channels the highest.These results offer theoretical grounds and decision supports for scientific,realistic and effective flood mitigation measures.