【作者】 张升堂；

【导师】 康绍忠；

【作者基本信息】 西北农林科技大学 ， 农业水土工程， 2004， 博士

【摘要】 黄土高原丘陵沟壑区是一个干旱缺水且水土流失严重的地区,生态环境极为脆弱,人类活动会引起本地区水分转化与循环的强烈响应,研究本地区各种人类活动的水文效应已成为改善本地区生态环境的前提,尤其是对近年来大规模进行的各类水土治理活动所引发的降水径流过程的变化研究更是指导流域科学治理,缓解地区水资源危机,防治洪涝、水土流失灾害,实现地区水资源可持续发展的理论依据。　由于水土治理活动所引起的流域面地理因子或水文要素的变化具有高度空间变异性,而分布式水文模型可以对研究域进行离散,对各要素的空间变异性进行定量分析,所以本论文用建立分布式水文模型的方法为水土治理影响下的流域水文过程研究提供了一种工具。　本文所建模型对流域进行二级离散,即按自然水系将流域离散为若干个子流域,然后将每一子流域栅格化,在栅格单元范围内如坡长、糙率、植被类型、覆盖度等因子无剧烈变化认为其是均一不变的,利用集总式水文模型方式模拟栅格内大气降水的植被层截留、地表入渗等水分转化与循环过程。然后,在栅格单元之间和子流域之间模拟径流汇流过程。随水土治理活动的开展与时间持续,模型可以根据下垫面变化状况通过更新栅格单元内坡度、坡长、糙率、植被类型、覆盖度等因素,考虑水土治理对流域水分转化和循环的影响作用。　论文在总结以往分布式水文模型模拟降水分布所用的空间插值方法的基础上,认为空间插值法不能模拟降水中心及降水中心降水量,而且空间插值法在模拟降水分布时会形成人为间断,不符合降水连续分布的物理现实。因此,本文建立了一种能模拟降水分布中心和中心降水量的降水分布数学模型。模型符合降水量在流域面上连续分布的特点,勿需进行繁琐的距离计算,与其它已有降水分布模型相比更简单,更易于实际操作与应用。　在栅格单元范围内的植被截留和入渗模拟中,论文采用的模型参数均具有明确的物理意义,使模型参数易于确定。模型研究域经二级离散后,具有较高分辨率,因而对降水的植被截留和入渗模拟具有较高精度。模拟过程中如果植被类型、覆盖度、坡度等对植被截留和入渗有显著影响的因子在栅格单元内具有明显空间变异时,本论文首次提出对这种情况明确的处理方法。　栅格范围内净雨形成后在向下游栅格单元汇集过程中,径流将在栅格单元之间分配,以往分布式水文模型对栅格单元之间的径流分配采用单流向或多流向等多种方法进更多还原

【Abstract】 The Loess Plateau is a place where water is deficient, but soil and water loss is severe. The human activities can change the hydrological conditions and ecological systems easily in this area. Specially, the soil and water conservation activities that take place almost on every hill slope of the Loess Plateau nowadays can lead to the responses of the water conversion and hydrological cycle strongly. For many years hydrologists have attempted to relate the hydrologic response of watersheds to human actives and soil and water conservation measures. In this case, studying the hydrological effects of the soil and water conservation activities will premise the improving of the ecological environment in this area. For example, studying on the changes of the precipitation-runoff process owing to the soil and water conservation activities in recently years can guidance the basin soil and water management, relaxed the crisis of the water resource, prevention to the heavy flood and soil erosion. Moreover, it shows the theory on which the water resource sustainable development can be realized in the Loess Plateau. The hydrological and geological elements changes owing to the soil and water conservation have the characters of high spatial variation in the basin. The distributed hydrological model can disperse the basin, and quantitative analysis the spatial variation of the elements. On this theory, this dissertation established a distributed hydrological model and studied the basin hydrological process impacted by the soil and water conservation using the model. The model dispersed the basin two times. The basin is dispersed into subbranch basins according to the natural water systems in the model, and the subbranch basins are dispersed into grid cells. In the grid cells scale, the hydrological processes such as the vegetation interception, the soil water infiltration and so on are simulated by the lumped hydrological model. Moreover, the flow concentration process is simulated in the subbranch basins scale. The soil and water conservation activities changed the elements of the grid cells such as the gradient, the length of slope, the Manning roughness coefficient, the types of vegetation, the ratio of vegetation coverage and so on. The effects of soil and water conservation on hydrological cycle and water conversion can be thought of through renewed the factors value in the grid cells scale. The paper analyses space plane interpolation method that is used by the former distributed hydrological models when simulated the distributing of the precipitation. And the paper present that this interpolation method cannot indicate the position of the precipitation center either the precipitation amount of the center. To make the matter worse, this method make man-made intercepted of the basin precipitation distribution. On the base of the analysis, a new mathematical model that able to indicate the precipitation center and the precipitation amount of the center is presented in the dissertation. Moreover, the model can simulate the continuous distributing character of natural precipitation. The model needn’t the cockamamie distance calculation at all. Therefore, it is not only simpler but can be used more easily as well. On the simulating the vegetation interception and infiltration in the grid cells, the model parameters in the model have definite physical properties. This make the value of model parameters can be estimated. Additionally, the model has high resolution through discreteness. Because of all of these, the distributed hydrological model can simulate the hydrological process accurately. At the same time, this dissertation presented a way to dealing with the problems in detail when the geological factors such as the vegetation types, the vegetation coverage ratio, the gradient and so on have high spatial variation in one grid cell. The surface runoff formed in one cell will flow into the lower cells around it. The problem of water quantity distribution among the lower cells comes into being. The former distributed hydrological models used to deal with this problem by the single direction method or multi-direction method. Both methods determine the flow direction(s) by the slope factor alone. They neglected the effect of the Manning roughness coefficient on the water quantity distribution among the cells. However, there have different values of Manning roughness coefficient in the different directions of a grid cell or in different grid cells. The different Manning roughness values affect the surface runoff confluence. On this theory, the concept of vector roughness was first introduced in the paper. And a new multi-directions water quantity distribution model, which think of the Manning roughness effects, is deduced from one-dimension Saint-Venant equation and Manning equation. This new model improves the runoff confluence pattern of the distributed hydrological models in the grid cells. When the isochrone of the surface runoff is calculated, the paper presumes that the concentration time of a grid cell is the sum of the surface runoff flowing across the grid cell time and the lower cell concentration time. Accordingly, a physical based grid cellconcentration time equation developed in this paper. Because there are a few of lower grid cells around a grid cell, the lower cell concentration time of the grid cell can be regarded as the average concentration time of all of the lower grid cells. Therefore, after the precipitation begins, the surface runoff of a cell will reach the basin outlet in its concentration time later. And a cell’s runoff hydrograph is formed. The runoff hydrograph of the whole basin will be obtained by linear adding the all grid cells’runoff hydrograph according to the concentration time order. After the distributed hydrological model constructed, the author program turbo C++ computer programs. The programs can carry out the distributed hydrological model simulation about the basin precipitation-runoff. The grid cells’data about the geography position, the gradient, the Manning roughness, the vegetation types, the vegetation coverage and so on are stored in data files. The main program read those data when there have precipitation process input. And a runoff process at the basin outlet, which responds the precipitation process input, will be outputted at high speed by computer. Meanwhile, the data can be refreshed on the conditions of the soil and water conservation. Therefore, the hydrological effects of the soil and water conservation in the basin can be real-time simulated. Finally, the distributed hydrological model is tested in the Nanyaogou small experimentation watershed of the Wudinghe middle reaches. Despite the limitations in the data availability, the results of river discharges predictions seem to be reasonable in the scale concerned. The result indicated that this distributed model can simulate the effects of the soil and water conservation activities on the hydrological cycle in the Loess Plateau, and has higher precision than the Shanbei lumped hydrological model. Moreover, the model not only has a sensitivity response to the increase the area of afforestation and terrace, but also to the precipitation process as well.更多还原