【作者】 王帅；

【导师】 王浩；

【作者基本信息】 天津大学 ， 水文学及水资源， 2013， 博士

【摘要】 水是生命之源，也是经济和社会发展的命脉。近些年来受全球气候环境变化及人类活动影响，许多国家地区出现河流流量减少、水环境恶化以及水质污染等问题，严重制约了社会经济的可持续发展。渭河流域是资源性缺水地区，气候变化使气温、降水、径流等水循环要素在季节和空间上发生改变，增加了流域旱涝灾害的发生机率，导致水资源安全情势将更为严峻，未来流域供需矛盾将更加突出。随着气候变暖和用水需求日益增大，充分认识并分析气候变化下流域水循环演变规律显得尤为迫切和重要。本文首先利用分布式水文模型对渭河流域水文循环系统进行模拟，并在此基础上运用现代统计学手段对流域水循环演变特征进行分析，同时结合统计降尺度模型开展未来气候变化情景下水循环要素模拟及预测研究，以期为流域水资源综合开发利用及水量调度管理提供科学参考和依据。主要研究成果与结论如下：（1）根据渭河流域地理条件及气象水文特征，基于EasyDHM水文模拟平台建立渭河流域分布式水文模型，完成模型参数率定及验证工作，取得了较好的模拟效果。结果表明，EasyDHM在渭河流域具有较好的适用性，构建的模型能够客观描述该流域水循环过程，可以应用于实际问题研究。（2）在构建渭河流域分布式水文模型的基础上，采用多种统计检验手段，从年、季、汛期和枯水期等不同时段角度，对渭河流域近55年（1954~2008年）气温、降水及径流要素的时空分布特点及演变特征规律进行分析，结果表明：①从空间分布看，多年平均气温及降水量在空间上存在不均衡性，气温以西安站为中心向外呈辐射递减趋势，降水呈现由东部和南部向西北部逐渐减少的分布格局。②从均一性分析看，气温、降水及径流时间序列在95%置信水平下都存在均一性。③从趋势变化上看，流域年平均气温呈现显著上升趋势，降水量变化趋势不显著，径流量呈现显著减少趋势；除冬季的降水量呈现显著上升趋势以外，其余季节及汛、枯期变化趋势与年际大体相同，且上述变化趋势未来仍可能持续。④从突变分析上看，流域年平均气温及径流量存在明显突变情况，而降水量无显著突变现象发生；除冬季的降水量存在显著突变现象以外，其余研究时段突变分析结果与年际结果一致；⑤从周期分析上看，不同时段的气温、降水和径流要素均存在多个特征时间尺度，具有不同的周期成分，径流的周期变化特征与气温和降水的周期变化特征表现出很好的相关性。（3）基于统计降尺度模型SDSM建立渭河流域气温及降水要素的降尺度模型，并完成模型标定和验证期模拟能力评估，结果表明模型的模拟结果较为可靠，可以应用于渭河流域气温、降水等气候要素未来变化趋势的预估。选取A2和B2气候变化情景进行分析可知，两种情景下未来90年（2010~2099年）渭河流域的平均气温呈现较为显著的上升趋势，而降水量总体变化趋势不显著。将统计降尺度模型与已构建的分布式水文模型相耦合，对渭河流域未来径流量总体变化进行预测并对干支流上5个主要控制性水文站径流变化展开研究，结果显示：渭河流域未来径流量相对基准期（1961~2009年）多年平均水平总体上有所减少，其中A2情景下减少趋势十分明显，而B2情景变化趋势不显著。更多还原

【Abstract】 Water is the source of life, also is the lifeblood of economic and socialdevelopment. In recent years, influenced by global climate change and humanactivities, many countries and regions have the problems of river flow reduction,water environment deterioration and water pollution, which restrict the sustainabledevelopment of social economy. Weihe River basin is a water-lacking area, andclimate change brings an important impact to water cycle factors such as temperature,precipitation and runoff in season and space, which will increase the occurrenceprobability of flood and drought in the river basin, lead to water security situationmore serious, and result in the imbalance between supply and demand more poignantin the future. With climate warming and water demand increasing, it is especiallyurgent and important to understand and analyze the evolution of dynamic water cycleunder climate change in Weihe River basin.This paper firstly established the distributed hydrological model to carry out thesimulation of water cycle process in Weihe River basin, and then used a variety ofmodern statistical methods to full-scale analyze the variation characteristics ofhydrologic and meteorological elements, and in addition, combining with statisticaldownscaling model, this paper carried out the research on simulation and predictionof water cycle elements under future climate change scenarios, so as to providescientific reference for reasonable development and utilization of basin waterresources and management of water regulation. The main results and conclusions areas follows:(1) According to the geographical conditions and meteorological hydrologicalcharacteristics of weihe river basin, and based on EasyDHM hydrological simulationplatform, this paper established the distributed hydrological model in Weihe Riverbasin, completed calibration and validation of parameters for14parameter subareas,and obtained a good simulation result. The results showed that EasyDHM had goodapplicability in Weihe River basin, and the established model described the innermechanism of hydrologic processes objectively, therefore it could be applied toresearch and solve practical problems.(2) Based on the distributed hydrological model in Weihe River basin, the paperused many kinds of statistical test methods to analyze the temporal-spatial distributionand evolution characteristics of temperature, precipitation and runoff during recent55 years (1954-2008) in the different period of time, such as year, season, flood seasonand dry season. The results showed,①There was an irregular spatial distribution ofannual mean temperature and precipitation. The temperature showed a decreasingtrend from the centre near the Xi’an station to all directions, and precipitationappeared a decreasing trend from the east and south to the northwest as a whole.②Inthe95%confidence interval, the temperature, precipitation and runoff time series ofWeihe River basin existed homogeneity.③As for trend change, mean temperaturedemonstrated a remarkable significant increasing trend, there was no significant trendin precipitation, and runoff showed a significant decreasing trend in annual. Thetrends are still likely to continue in future. Except in winter, when there were asignificant increasing trend in precipitation, the variation trends in the rest of theseasons, flood season and dry season were approximately the same as those in annual.④From sudden change analysis, there were significant mutation phenomenon inmean temperature and runoff time series, and precipitation had no significant suddenchange in annual. Except in winter, when the mutation phenomenon was very obviousin precipitation, the conclusions between in other study periods and in annual wereconsistent.⑤As far as period variation was concerned, temperature, precipitation,and runoff existed multi-time scale change characteristics with different periodcompositions, and the periodical variation characteristics in runoff appeared goodcorrelation with those in temperature and precipitation.(3) Based on statistical downscaling model SDSM, this paper established thedownscaling model for temperature and precipitation respectively in Weihe Riverbasin, and carried out the calibration and validation of the model. It showed that thesimulation results were reliable and the model could be applied to predict thevariation trend of meteorological elements such as temperature and precipitation. Thepaper chose A2and B2climate change scenarios for research, and the resultsindicated that in the next90years (2010-2099), mean temperature demonstrated aremarkable significant increasing trend, and variation trend of precipitation was notobvious as a whole under the two scenarios in Weihe River basin. The statistical downscaling model and the distributed hydrological model were coupled to predict thefuture runoff changes of the whole basin and5control hydrological stations onmainstream and tributary. The results showed that the total runoff would be less thanthe average value of reference period (1961-2009) and the decreasing trends in thefuture were very obvious under A2scenario and not significant under B2scenario.更多还原