EGI OpenIR  > 研究系统  > 荒漠环境研究室
高海拔山区流域水文过程及气候变化影响研究—以叶尔羌河为例
刘蛟
学位类型博士
导师刘铁 ; Philippe De Maeyer
2017-05-01
学位授予单位中国科学院大学
学位授予地点新疆乌鲁木齐
学位专业理学博士
关键词水文过程 模型联合 遥感数据 方差分析模型 气候变化 叶尔羌河
摘要在干旱半干旱区,山区水资源对下游绿洲区的发展具有支配作用,而准确理解山区水循环过程对下游可持续地水资源管理具有十分积极的作用。然而,由于极端地形变化、复杂水文过程以及实测站点稀缺等因素,使得山区流域水文过程模拟的难度加大。本文选取了发源于喀喇昆仑山北坡的叶尔羌河流域为例,通过多水文模型的联合应用、遥感技术以及统计分析模型,对高海拔山区流域水文过程进行了分析研究。进一步通过与 GCM 的耦合,对不同的未来气候变化情景下水循环过程的转变,水文要素的再分布等进行了探讨。本文研究主要包括了以下几个方面的工作:(1)通过 SWAT 和 MIKE SHE 模型的联合应用,从径流、积融雪以及蒸散发三个水文要素方面研究了模型结构和模块算法对水文过程的影响。结果表明,在缺少多率定目标的情况下,单一模型只能在某些水文要素上取得可行的模型结果,而不能对整体水文过程进行准确描述。整合多模型联合应用的模拟结果,可以减少模型结构和算法的不确定影响,从而提高对流域水文过程的认识。结合遥感数据以及相关研究的交叉验证,整体上明确了叶尔羌河流域的空间水循环过程。(2)借助遥感数据和统计模型,明确了输入数据的不确定性对水文过程中不同水文要素的影响。首先以站点实测数据为基础,对遥感数据进行了精度验证和偏差校正。校正后的 TRMM,MODIS 以及潜在蒸散发数据在 MIKE SHE 的模拟中并没有提高对日径流过程的模拟精度。但是,进一步利用多因子方差分析模型,确立了在复杂水文循坏系统中“输入”与“输出”之间的显著性影响关系。借助这种显著性影响关系,分析得到了在 TRMM 驱动下,模型得到的积雪分布空间差异更加明显;在 MODIS 温度数据驱动下的模型在中低山区存在更多的永久性薄层积雪,而高山区的储雪量减少;在模型的水量平衡控制下,遥感蒸散发驱动的模型在自动率定中通过参数调整,减小了实际的植被蒸腾作用。(3)通过 GCM 与水文模型的耦合,分析了未来气候变化对研究区水文过程的影响。提取 21 个 GCM 中气候因子的变化信号,通过改进的百分位数波动法和 Delta 法,将变化信号与历史实测数据叠加后,得到未来的降水和温度数据,并用于驱动 MIKE SHE 模型,从流域水平衡角度分析了未来水文过程的变化情况。结果表明,因冬季极端降水的增强,未来叶尔羌河流域的降水呈现小幅度的增长趋势。未来情景下,山区的增温小于平原区,而在 RCP4.5 和 RCP8.5 两种情况下,增温差异很大。到本世纪末,RCP4.5 的增温速度稳定在 0.31°C/10 年,而 RCP8.5 将增加到 0.59°C/10 年。未来气候变化情景下,叶尔羌河流域的水文过程将发生明显的变化,不同形式的水资源之间的转化改变了水文要素的分布。蒸发耗散的不断增加,使得山区的固体储水量和下游的可用水资源量减少。流域不同高程带的积雪量变化差异很大,5600m 以下的永久性积雪区将在 2060~2079年消失,5600~6400m 区域积雪显著减少,6400m 以上基本没有变化。同时,增温导致的前期融雪增加,后期融雪量减少,在出山口卡群站也形成了前期径流增加,后期减少的年内径流过程。同时,极端洪峰流量出现的频率增加,时间提前。未来叶尔羌河流域将面临更加不稳定的水资源情况以及更严峻的洪水安全问题。本研究关注于流域尺度上空间水循环过程。针对高海拔山区流域水文过程模拟中的主要问题,开展的联合应用,为缺资料地区水文模型的模拟,提供了有效的多要素评价方法;借助于统计分析模型,得到了复杂系统中“输入”与“输出”之间的显著性影响关系,以此明晰了输入数据不确定性对不同水文要素的影响;在提取 GCM 降水变化中,通过明确不同分位数段的降水频率变化对百分位数波动法提出了有效地改进,减小了未来降水数据的不确定性。本研究在高海拔山区流域水文过程及其气候变化影响研究方面取得了进展,为山区水文过程模拟提供了参考。
其他摘要In arid regions, the water resources in mountain regions dominate the development of downstream oases. An accurate understanding of the mountainous hydrologic cycle plays a very positive role in sustainable water resources management. However, because of the steep topography, a more complex hydrologic cycle, and sparse gauging with low representativeness for mountain regions, hydrological modeling becomes more difficult in high-altitude mountainous catchment. The Yarkant River basin originating from the North Slope of Karakoram was taken as a case study in this dissertation. Based on joint application of multiple hydrological models, remote sensing data and statistical models, the hydrological processes were analyzed. Additionally, the alterations of the hydrologic cycle and the redistribution of hydrologic components were investigated under climate change. The main objectives of this dissertation included the following: (1) By the joint application of the SWAT and MIKE SHE models, the effects of model structure and algorithms on the hydrological processes were analyzed in terms of runoff, snow and evapotranspiration. The results showed that without multiple calibrations, the using of a single model can successfully model only some parts of the natural hydrologic cycle, but not the entire process. The complementarity of the SWAT and MIKE SHE models could overcome most model structural and algorithmic uncertainties and assist in understanding integrated hydrological processes. (2) Combining with remote sensing data and statistical model, the responses of different hydrological components to input data were analyzed. Taking observations as a benchmark, precision checking and bias corrections were implemented firstly. The corrected TRMM, land surface temperature (LST) of and global potential evapotranspiration (PET) failed to improve the simulation performances at Kaqun station’s hydrography. However, on the premise of statistical hypothesis testing implemented by analysis of variables (ANOVA), their significant effects on different hydrological components were investigated Compared with model forcing by interpolated station data, the spatial deviations of snow storage in different elevation bands were more obvious in the TRMM driving model. In the LST driving model, there were larger snow coverage in the low-middle mountain region and less snow in the high mountain region. Additionally, remotely sensed PET driving model produced less transpiration because of the re-adjusted parameters in the auto-calibration. (3) Coupling with the GCMs and the well-calibrated hydrological model, the hydrological processes under future climate change scenarios were analyzed. The average ensemble change signals of 21 GCMs in precipitation and temperature were extracted, the modified quantile perturbation method (QPM) and the Delta method were applied to add these change signals into historical observations, then, the variations of hydrological processes under climate change were studied though coupling with the well-calibrated MIKE SHE model. The results suggested that the increased extreme precipitation in winter led to the moderate growth of annual precipitation volume. The warming in the mountain region was slower than in the plains region; moreover, large deviations were revealed between RCP4.5 and RCP8.5 scenarios after 2060. At the end of this century, warming values of 0.31°C/10a and 0.59°C/10a were predicted to occur under RCP4.5 and RCP8.5, respectively. Under the climate change, increasing evaporation dissipation would lead to decreasing snow storage in the higher altitude mountain region and likewise with regard to available water in the downstream region. The alterations of snow strong are quite different in elevation bands, the permanent snowpack area below 5600 m would completely vanished in 2060-2079, and the snow storage in 5600-6400 m would be reduced dramatically, however, there is little or no changes in the above 6400 m region. Warming could cause a stronger spring and early summer stream runoff and a reduced late summer flow due to the snowmelt change in temporal distribution. Furthermore, both the frequency and the intensity of the flood would be enhanced. In Karakoram region, more researches are needed about the transient water resources system and the worsening of flood threats in future. The entire hydrological processes on the catchment scale were focused in this dissertation, and some innovations were highlighted to overcome the main challenges. The joint application of multiple models provided a good suggestion to multi-factors evaluation of hydrological model in scarce gauging catchment. ANOVA successfully defined the significant relationships between “effects” and “caused” in hydrological processes, and help to understand the intervallic influences of input data on different hydrological components. The QPM was effectively optimized by considering the frequency changes in the different precipitation intensity ranges. This study makes a progress and supports a reference in high-altitude mountainous hydrological processes and climate change’s influences study.
学科领域地图学与地理信息系统
语种中文
文献类型学位论文
条目标识符http://ir.xjlas.org/handle/365004/14787
专题研究系统_荒漠环境研究室
作者单位1.中国科学院新疆生态与地理研究所
2.比利时根特大学
推荐引用方式
GB/T 7714
刘蛟. 高海拔山区流域水文过程及气候变化影响研究—以叶尔羌河为例[D]. 新疆乌鲁木齐. 中国科学院大学,2017.
条目包含的文件
条目无相关文件。
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[刘蛟]的文章
百度学术
百度学术中相似的文章
[刘蛟]的文章
必应学术
必应学术中相似的文章
[刘蛟]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。