【作者】 李倩；

【导师】 都金康；

【作者基本信息】 南京大学 ， 地图学与地理信息系统， 2012， 硕士

【摘要】 本文以秦淮河流域为研究对象,利用遥感与GIS技术提取研究区城市化信息。分析秦淮河流域1986-2009年的城市化空间格局变化。运用CA-Markov模型预测秦淮河流域未来土地利用情况并构建城市化空间格局情景。通过建立HEC-HMS分布式水文模型,探讨研究区城市化及其空间格局变化带来的长期径流和短期洪水水文效应,揭示城市化进程与径流过程之间的内在联系。为今后秦淮河流域的土地利用规划、水资源规划、城市建设和防洪减灾等提供了决策的参考依据。本文由国家自然科学基金重点项目(40730635),水利部公益项目(200701024)“长江三角洲地区城市化对河流系统与水文过程的影响研究”——“秦淮河流域城市化对水文过程影响的研究”支撑。本文的主要研究工作如下：1利用多时相、多源遥感影像来提取不透水面等土地利用信息。采用Landsat、中巴资源卫星两种遥感影像数据,每个像元均可视为由高反照度地物、低反照度地物、植被和裸土组成,运用线性光谱混合模型提取不透水面信息。分析秦淮河流域多年的城市化空间格局变化。根据研究区2003-2006年土地利用的变化趋势,利用CA-Markov模型预测2009年、2012年和未来2018年的土地利用。在此基础上,构建秦淮河流域城市化情景,使城市面积逐渐扩大而保持其它土地利用类型不变；构建城市化空间格局情景,在流域不透水率不变的前提下,运用ArcGIS构建四种空间格局情景,即将不透水面分布在流域的上游、下游、外围子流域以及平均分布在流域内,为后面研究做准备。2建立适用于秦淮河流域的HEC-HMS分布式水文模型。根据HEC-HMS系统的原理和特点,选择适合研究区特点的降雨、蒸散发、产流、汇流模型,构建该流域的分布式水文模型。结合GIS与遥感技术获取模型所需的驱动数据和初始参数,进行参数率定及模型验证,为分析流域内城市化的长期水文效应和洪水效应提供科学的分析方法。3通过情景分析法,模拟城市化的水文效应,揭示秦淮河流域城市化及其空间格局变化的长期水文效应和洪水效应。(1)分析城市化对长期径流过程的影响：主要采用固定气象条件而仅仅改变土地利用情景的方法,模拟日径流过程,统计并分析年径流量变化与城市化之间的联系,从而得出秦淮河流域城市化的长期水文效应。当不透水率由1988年的3.4%增加到2018年的31.4%时,年均径流深增加了6.8%。在相同城市化情景下对不同降水条件进行模拟,发现城市化对3个典型水文年的径流深影响幅度为：枯水年>平水年>丰水年,这表明年径流对城市化的响应,以枯水年最为明显,丰水年最弱。统计长时间序列和典型年的干季、湿季径流,发现城市化对干季径流的影响更为明显。(2)分析城市化对暴雨洪水过程的影响：通过模拟不同城市化情景下的短期暴雨洪水过程,分析城市化对洪水的影响。模拟结果表明,在同样降雨条件下,城市化对流域洪水过程的影响表现为：次洪水洪峰流量增加、洪量也增加。同时发现规模不同的洪水对城市化的响应程度不同,对于规模较小的洪水,洪峰流量、洪量变化更为明显；在同一场洪水中,洪量的变化比洪峰明显。(3)模拟空间格局情景的长期径流和洪水径流过程,分析城市化空间格局的水文效应。利用建立的HEC-HMS模型,模拟4种城市化空间格局情景下的长期径流和短期洪水过程,分析空间格局对流域的水文过程影响如何。研究结果显示秦淮河流域不透水率增量一定时,不透水面仅分布在流域上游时对流域的长期径流、短期洪水的影响最大；不透水面分布在流域下游时,对流域的长期径流和短期洪水的影响最小；当城市化平均分布在流域内或者只分布于流域外围时,流域的长期径流和短期洪水的水文效应居中。当流域内城市化的空间格局相对一致时,径流变化率与不透水率之间的线性拟合程度很好。更多还原

【Abstract】 Taking Qinhuai River Basin for study, RS and GIS technology was used to extract impervious surface and land use information in the study area. Variation characteristic of spatial pattern of urbanization since1988was studied in Qinhuai River Basin. Hydrological response of long-term runoff and short-term flood to impervious surface in Qinhuai River Basin was studied quantitatively and qualitatively here, using the HEC-HMS distributed hydrological model. Besides, for a given impervious rate, hydrological effects of different spatial pattern of impervious surface were discussed, revealing the intrinsic link between urbanization and runoff process. Based on the prediction of future landuse scenarios for Qinhuai River Basin, the long-term hydrological effects in the future was analyzed. All above could support decision for city construction, allocation of land use and water resource, flood control and disaster reduction in this area. This paper was supported by the National Natural Science Foundation of China (40730635), Public Projects of Ministry of Water Resources (200701024),"Study of impact of urbanization in Yangtze River Delta on river systems and hydrological processes"-"Study of impact of urbanization in Qinhuai River Basin on the hydrological processes".Research in this paper focused on the following three aspects:1Extraction of impervious surface for study area with multi-temporal, multi-source remote sensing images was implemented in this paper. Using Landsat and CBERS images as two data sources, each pixel was dealt as a mixture of vegetation, high-and low-albedo and bare soil, impervious surface was extracted with linear spectral mixture model, landuse information was determined with supervised classification method. It was analyzed that change of spatial pattern of urbanization in the Qinhuai River Basin since1988. Based on the trends of landuse in study area from2003to2006, the future landuse information in2009,2012and2018was predicted with the CA-Markov model. On this basis, urbanization scenarios in Qinhuai River Basin were built with gradually expanding impervious surface, while other landuse types remaining unchanged; spatial pattern scenarios of urbanization were built for the following study.2A distributed hydrological model named HEC-HMS was built for Qinhuai River Basin. According to the theory and characteristics of HEC-HMS system, appropriate modules were selected for the study area including precipitation, evapotransporation, runoff, transforming and baseflow. The required driving data and initial parameters were obtained in help of combination of GIS and RS technologies. And then HEC-HMS model should be calibrated and validated, giving a scientific method for the analysis of long-term and flooding hydrological effects of urbanization in the watershed.3The hydrological effects of urbanization were simulated through scenario analysis method which revealed the relationship of urbanization and long-term/flood hydrological effects in Qinhuai River Basin.(1) Analysis of the long-term hydrological impact of urbanization:keeping meteorological conditions fixed, the daily runoff process was simulated under different urbanization scenarios, and statistics of annual runoff for long-term was analyzed which showed the relationship between change of annual runoff and landuse type, especially the impervious surface. Impervious rate of3.4%in1988increased to31.4%in2018, the annual average runoff depth of the basin has increased by6.8%. Meanwhile annual runoff was simulated under the same urbanization scenarios but different conditions of precipitation, it could be concluded that the hydrological impact of urbanization on annual runoff depth of typical hydrological years was as following order: dry years> average year> wet year. This indicated that among the hydrological responses of annual runoff to urbanization, dry year was the most obvious, and wet year the weakest. Statistics of dry season, wet season runoff of long term and the typical years, the hydrological response of dry season to the urbanization was more obvious.(2) Analysis of the short-term hydrological impact of urbanization:the impact of urbanization on storm flood was studied through simulation of flood process under different urbanizatioin scenarios. The simulation results showed that under the same precipitation condition, both peak flow and flood volume increased, and flood with smaller scale increased more obviously; for a given flood, the change rate of flood volume were more obvious than the change rate of peak flow.(3) Hydrological effects of spatial pattern of urban space in Qinhuai River Basin. Four scenarios of different spatial patterns were built with ArcGIS by locating urban area in the upper reaches, the lower reaches, the periphery, and each sub basins of the whole basin, while keeping the impervious rate of the whole Qinhuai River Basin unchanged. Long term runoff and some flood events were simulated under the same meteorological condition but different scenarios of urban area. And the results of simulation showed that both long term runoff and flood events had hydrological response to different spatial patterns of urban area. The spatial pattern of urban area bringing the largest change on runoff was when the impervious surface was located in the upper reaches, then in the periphery, next in each sub basin, and the last in lower reaches. When the spatial pattern of urban area was relatively consistent, the degree of linear fit in changing rate of runoff depth and impervious rate was much better than that of real case of urban distribution.更多还原