【作者】 郭靖；

【导师】 郭生练；

【作者基本信息】 武汉大学 ， 水文学及水资源， 2010， 博士

【摘要】 气候变化是当今国际社会普遍关注的全球性问题。水循环作为气候系统的重要组成部分,首当其冲受到气候变化的影响。研究气候变化对流域水循环及水资源的影响,对维护河流健康生命,保障水资源可持续利用,保证流域社会经济可持续发展具有重要的现实意义。论文在综述国内外研究现状和进展的基础上,结合国家自然科学基金等课题,开展气候变化对流域水循环及水资源的影响研究。论文的主要工作及成果如下：(1)比较分析了ANN、SSVM、SDSM、ASD和LARS-WG五种统计降尺度方法以及全球气候模式CGCM2对降水和气温的模拟结果。结果表明统计降尺度法能够明显改善全球气候模式降水和气温输出结果；SSVM法对降水量的模拟最好,SDSM和ASD法对降水的变化过程模拟较好；五种统计降尺度法对气温模拟的结果要优于对降水的模拟,并且它们的模拟值基本一致,其中SDSM和ASD法的模拟效果较好。(2)应用SSVM、SDSM和ASD统计降尺度方法预测汉江流域未来降水、气温变化。在CGCM2和HadCM3模式和未来A2情景下,三种降尺度方法模拟汉江流域降水结果存在一定差异,但是可以看出它们预测未来3个时期(2020s(2011～2040),2050s(2041～2070),2080s(2071～2100))汉江流域降水长期将存在增加的趋势。气温预测的一致性较好于降水,预测结果表明：2020s时期气温将比基准年(1961～2000)降低,2050s时期气温和基准年基本一样,2080s时期将比基准年升高,并且未来3个时期汉江流域气温存在上升趋势。(3)通过Bay-LSSVM统计降尺度方法对GCM的输出进行尺度降解,以及分别采用两参数月水量平衡模型和VIC分布式水文模型与GCM进行耦合,预测未来丹江口水库及整个汉江流域径流量情况。结果显示,虽然模拟情况不尽相同,但未来3个时期汉江流域的径流量有增加趋势。(4)预测汉江流域社会经济发展、人口增长及水资源的需求。采用GCM输出和SSVM法模拟在A2情景下,汉江流域未来径流量的变化情况,分析了汉江流域水资源系统对气候变化的脆弱性反应。结果表明汉江流域水资源系统的调节能力较强,不会出现水资源系统恶化的“阈值”。(5)把经过ASD尺度降解的CGCM2输出作为HBV水文模型的输入,分析气候变化情景下,汉江上游径流极值事件的变化情况。在A2情景下,未来汉江上游同一量级的洪水重现期将缩短,同一重现期的枯水历时将增加,这可能增加丹江口水库的防洪风险,同时给中线调水工程及整个流域水资源利用带来不利影响；在B2情景下,未来汉江上游洪水和枯水极值事件都趋于平缓,比较有利于汉江流域的水资源综合利用。(6)应用ASD降尺度方法降解全球气候模式HadCM3的输出,得到未来长江流域的降水情况；利用Mann-Kendall检验分析未来长江流域降水极值事件的时空变化情况。结果显示：未来在A2和B2情景下,长江流域年降水有上升的趋势,降水增加显著的区域主要位于长江流域上游东部和中游中南部；未来90年(2010-2099)长江流域的年／夏季降水极值事件明显增加的站点,主要位于长江上游西北部和中游区域；长江流域上游西北部和中游区域未来遭受洪涝灾害的可能性将有所增加；未来长江流域上、中、下游降水极值事件可能在2050年后有所增加。更多还原

【Abstract】 Climate change is a global issue that is greatly concerned by the international community. As one of the important components of the climatic system, hydrological cycle has to bear the brunt of climate change. Therefore, the study of climate change impacts on hydrological cycle and water resources can play an important role in maintaining river health, protecting water resources sustainable utilization and ensuring sustainable socio-economic development of the watershed. The present situation and advancement of related research were concisely reviewed. Supported by the National Natural Science Foundation of China, this thesis was mainly focused on coupling the general circulation model (GCM) and hydrological model by statistical downscaling to investigate climate change impact on spatial-temporal distribution of water resources, discuss and analyze the vulnerability of hydrological cycle and the extreme hydrological evens to climate change. The main works and innovations were summarized as follows:(1) The simulated results of precipitation and temperature by five statistical downscaling methods, including Artificial Neural Network (ANN), Smooth Support Vector Machine (SSVM), Statistical Downscaling Model (SDSM), Automated Statistical Downscaling model (ASD) and Long Ashton Research Station Weather Generator (LARS-WG), were compared with the output of CGCM2. The results show that statistical downscaling can significantly improve the precipitation and temperature output of general circulation model. It is found that SSVM is the best at simulating amount of precipitation in the five methods, while SDSM and ASD are better at simulating variability of precipitation and temperature. Furthermore, the comparison results also show that the simulation of temperature by the five downscaling methods are basically the same, and better than the simulation of precipitation.(2) Under A2 climatic scenario, the change of precipitation and temperature in the Hangjiang Basin during 2020s (2011～2040),2050s (2041～2070) and 2080s (2071～2100) were predicted by SSVM, SDSM and ASD, respectively. In spite of the predicted trends of precipitation corresponding to CGCM2 and HadCM3 by the three downscaling methods are rather different, it is shown that the precipitation will increase during these three periods compared with baseline value. For predicting the trend of temperature, the results are more consistent than precipitation, and it is shown that the temperature will decrease during 2020s, no significantly change during 2050s, and, increase during 2080s. Throughout these three periods, there is increase temperature trend in the Hanjiang Basin.(3) The future runoffs of the Danjiangkou Reservoir and whole Hanjiang Basin were predicted by Bay-LSSVM to downscale the output of GCM, and then coupled with two parameter monthly water balance model and the distributed VIC model, respectively. The results show that the runoff will increase in these three periods in future.(4) The socio-economic development, population growth and water demand in the Hanjiang Basin were predicted. By taking account of the prediction runoff under A2 climatic scenario downscaled by SSVM, the vulnerability response of water resource system in the basin to the climate change was analyzed. The results indicate that the regulation ability of water resources system in the Hanjiang Basin is strong, and it is not found the threshold of deteriorating the system.(5) Under A2 and B2 climatic scenarios, the runoff of the upstream Hanjiang Basin were predicted by coupling CGCM2 with HBV hydrologic model by means of ASD downscaling method. The study results show that the intensity and frequency of flood and drought events, measured by peak discharge and low flow duration, respectively, will increase significantly in the period of 2011 to 2100 compared to the period of 1961 to 2000 under A2 climatic scenarios, which give the adverse effects to flood control in the Danjiangkou Reservoir, implement of the middle route of South-to-North Water Diversion Project and utilization of water resources in the whole basin. However, under B2 climatic scenarios, both flood and drought events will moderate, which is beneficial to comprehensive utilization of water resources in the basin.(6) An automated statistical downscaling method (ASD) was first presented to downscale the output of HadCM3 and predict the precipitation of the Yangtze Basin under A2 and B2 climatic scenarios. The Mann-Kendall test was applied to investigate spatial and temporal changes of precipitation extremes over the Yangtze Basin at 95% confidence level under the climate change conditions. It was revealed that the annual precipitation has an increasing tendency during in the period of 2010 to 2099, and the significant increasing trend would mostly dominate in the eastern part of the upper region and the southern and central parts of the middle region. The results also show that the significant increasing trend of precipitation maxima are mainly occurred in the northwest upper part and the middle part of the Yangtze Basin for the whole year and summer under the both climate change scenarios, and the starting point for pattern change of precipitation maxima will appear about 2050. Therefore, it can be concluded that the northwest upper and middle Yangtze Basin might encounter higher risk of flood hazards after about 2050.更多还原