Statistical Downscaling Methods In Upper Reaches Of Yangtze River Basin Of China

The influence of climate change on the water cycle and water resources becomesa global problem for both the governments and experts. But there is large mismatch ofscale between GCMs and the scale of regional models. Different downscalingmethods were brought forward to deal with this problem. Statistical downscalingmethods have superior ability and simple computation, it has become the focus ofclimatologists and hydrologists, and been paid more and more attention.The upper Yangtze River basin was divide into five subareas including JinshaRiver, Min tuo River, Jialing River, Wujiang river and main stream of the upperYangtze River.Based on the analysis on the spatial and temporal changes ofprecipition and temperature，the multiple regression and statistical downscalingmethods were used to simulate and predict the precipitation and temperature changesduring three periods(from2010to2039, from2040to2069and from2070to2099) inthe upper Yangtze river basin. Finally, the two parameters water balance model wasused to analyze the runoff changse. The main work and results were summarized asfollows:(a) The changing characteristics of mean surface precipitation and temperaturewere analyzed. The characteristics including the complexity, trend and period wereanalyzed through calculating the Hurst coefficient, Mann Kendall rank andcumulative anomaly methods. It was found that the precipitation in the Jinsha RiverBasin increased, unchanging in the Wujiang River, decreased in the other river. Thetemperature in five subbasins increased.(b) The precipitation and temperature data of NCEP from1961to2000wereanalyzed in the view of seasonal changes. It was found that the precipitation in theupper Yangtze River Basin only increased in spring, decreased in the other seasons.The temperature in the four seasons wre both in warming trend, especially in winter.(c) The multiple regression and statistical downscaling methods were used tosimulate and predict the precipitation and temperature in the upper reaches of YangtzeRiver. The simulation results wre good, and those of temperature were better thanthose of the precipitation. The future precipitation in the Jinsha River will increasemostly, followed by that in the Wujiang River and Min tuo River. The temperature inthe5subbasins will increase in different degrees. The two parameters water balancemodel was used and applied in the upper Yangtze River. The simulation accuracy was high in the Min tuo River, Jialing River and Wujiang River. The two parameters waterbalance model was used to preidicted the future runoff changes and it was found therunoff in the Jialing River will decrease mostly.(d) The future runoff change and climate change were compared. Although theprecipitation from2070to2099will increase about10%, the changes ofcorresponding runoff is less than those of precipitation, or even will decrease. Thatmaybe cause by rising temperatures and increasing evaporation activities. There aresome uncertainties about estimates of future runoff. But possible changes in trends inwater resources can be obtained and useful for watershed planning and water resourcemanagement.