【作者】 叶爱中；

【导师】 夏军；

【作者基本信息】 武汉大学 ， 水文学及水资源， 2004， 硕士

【摘要】 在学习了经典的分布式模型(如SWAT)的基础上，本文对已有的分布式时变增益模型进行了改进，建立了一个大尺度的分布式时变增益模型。该模型中加入了土地利用、覆被变化、人类活动、水保工程以及水库调度等模块。该模型的主要作用是分析人类活动对流域水循环的影响。将本模型应用于全黄河79.5km~2流域，结果较好。 本文主要工作与研究的内容如下： 1、给出了一套通过DEM提取河网、划分子流域的方法。该法的优点的能够保证流域中的每一个点都是连通的，最终将整个流域建成一有向无环图。每个网格的流向借鉴了D8方法的思想，但计算的方法是应用了图论中图的遍历的方法。河网的提取根据水流累计矩阵与流向确定，子流域则由河网得到。 2、建立了一大尺度分布式模型。该模型在每个小子流域中产流计算，通过河道进行汇流。每个子流域中需要计算的量有：降雨、蒸发、地表径流、壤中流、土壤湿度。子流域的降水与气象资料通过数据同化技术得到。在每个子流域中通过遥测土地利用信息与水保工程资料来考虑了土地利用与覆被变化的影响。农业用水通过耕地的面积与作物的类型来确定，然后分配到其所属的分区中。水库的调蓄作用是改变了水资源的时空分布，该模型中加入了常规调度方案。 3、建立了黄河月分布式模型。黄河流域人类活动影响十分剧烈。本文分析了黄河流域多年的降雨、径流、水保工程、农业灌区等资料。从黄河源至入海口模拟了17个流量站。上游的天然流域模拟效果很好效率系数可达0.8以上，随着黄河中下游人类影响的剧烈，由于收集的资料有限，模拟的精度有所下降。 4、可视化系统的开发与集成方面。在做分布式模型时，用多种语言、多类型数据库开发了一可视化系统。本文第五、六章给出了一些系统集成的方案。更多还原

【Abstract】 On the basis of studying the classical distributed model (such as SWAT), this paper advances the Distributed Time Variant Gain Model (DTVGM) and set up a large scale Distributed Time Variant Gain Model. There are many new sub modules in this model, such as land use module, land cover module, human activity module, water and soil conservation project module, reservoir dispatching module. The main function of this model is to analyse the human activity impact on water circulation of a basin. This model is applied in the Yellow River 79.5km2 basin, the research results indicate that the application of the model is successful in some sense, which can satisfy the requirement of water resources management.The groundwork of this paper and content studied are as follows:1, Give a method, which can gain the river net and divide the sub catchments only passing the DEM of a basin. The advantage of that method is which can make each grid connected in a basin and gains a directed acycline graph. The method of computing the flow direction of each grid is traversing graph, which study the idea of D8 method. The river net can be gotten according to the flow accumulation matrix and the flow direction of each grid. The sub catchments are gained by river net.2, Establish a large-scale distributed model. The runoff is calculated in each sub catchments, and then river nets get river routing. Each of sub catchment must deal some hydro elements, which are precipitation, evaporation, runoff and soil humidity. The precipitation and meteorological data of each sub-catchment are gained by the technology of data assimilating. The influence of the land use and cover change is considered from remote sensing and water and soil conservation project data. The irrigation water is fixed with the type of the crop and the area of the cultivated land. The function of the reservoir is to change the space-time distribution of the water resource. There is a Routine dispatcher’s module about reservoir in the distributed model.3 Set up the month distributed model of the Yellow River. It is very violent that the human activity influenced the Yellow River Basin. The paper analyzes the rainfall, flow, the area of irrigation and water and soil conservation project etc. data in yellow river. From yellow river source to seaport, the model simulates 17 flow stations’ observer flow. The efficiency coefficient of natural catchment in the upriver is greater then 0.8. The efficiency coefficient is simply not good enough in middle and downstream catchment, because of thehuman activity and not enough observer data.4 Developing visual system and system integration. We developed a visual system about distribute model. There are several types database and several languages in this system. Chapter 5 and chapter 6 give some method of system integration.更多还原