【作者】 穆振侠；

【导师】 姜卉芳； 董新光；

【作者基本信息】 新疆农业大学 ， 水文学及水资源， 2007， 硕士

【摘要】 水文现象在时间尺度、时间方向上的不可逆和空间上的变异性及水文资料信息的不足,在一定程度上限制了水文学理论的发展。随着计算机、地理信息系统(GIS)、遥感(RS)以及全球定位系统等科学技术发展的日新月异,为水文科学注入了新的血液。分布式流域水文模型可帮助人们更加深入的研究和了解自然界水文循环在不同时间和空间尺度上的演变过程和规律。流域信息是进行水文预报的必要信息,提取流域信息是构建分布式水文模型、进行水文预报及其他相关研究的前提。因此本文借助ArcGIS软件以新疆1:25万DEM数据研究了天山西部山区流域特征的提取,包括水系的提取、子流域的划分、流域边界线的提取等。传统的水文预报模型主要依赖雨量站实测降水和水文站实测流量,把整个流域水文过程作为集总过程来处理,基本上不考虑流域特性和降水量的空间分布。而影响降雨量空间分布因素中,除了有高程外,还有经纬度、坡度、坡向、风速等因素,不同的坡度坡向降雨量是不同的。本文通过选取了地学统计方法中插值效果较好的反距离加权法(IDW)、克里金方法(Kriging)、引进高程、坡度、坡向影响因素的协克里金方法(Co-kriging)探讨了天山西部山区降水空间分布规律。通过分析天山西部山区降水和融雪的特性,采用EOS/MODIS八日合成积雪数据,通过改进新安江三水源模型,在融雪型新安江三水源概念性水文模型的基础上,建立了反映天山西部山区流域特点的分布式水文模型。为了验证模型的可行性,本文选用了研究流域2005年5月各遥测及水文站的日降水过程以及日平均气温过程及月蒸发量资料,使用本文所建立的分布式水文模型进行模拟流域的逐日径流过程。将模拟结果与实测径流过程进行比较,经过分析表明,使用本文所建立的分布式水文模型能够取得较好的效果。月水量的合格率为100%,日水量的合格率为76.5%,预报方案的确定性系数为81.2,预报方案的有效性为乙等。更多还原

【Abstract】 Hydrology phenomenon limits the development of Hydrology theory in certain degree, such as hydrology phenomenon irreversible in time scale and direction of time, variation of space and hydrology data deficient. With the quick development of GIS, RS, GPS etc, Hydrology science has new power. Distributed hydrological model could help people go into further research and know the evolutionary process and disciplinarian of hydrological cycle in different time and space scale.Watershed information is necessary of hydrological forecast;extracting watershed information is Premise Conditions of building distributed hydrological model, making hydrological forecast and other relatively researching. So the method of extraction of watershed information in the west of Tianshan Mountain area is discussed in the paper by assuming DEM with 1:250000 scale of Xinjiang and 100m resolution and with ArcGIS software, including extracting water system, dividing subcatchments, extracting basin boundary etc.Traditional hydrological forecast model is main dependent on actual rainfall data of Precipitation Stations and observation discharge data of hydrology stations.It thought wholly Hydrological Processes is a integer and basically didn’t consider catchment’s character and spatial distribution of rainfall.But the factors which influence spatial distribution of rainfall have geographic coordinate, slope, aspect and so on besides height. so in this paper choose the better interpolation effect methods in genomic statistical means. The spatial distribution of rainfall in the west of Tianshan Mountain was discussed, based on IDW, Kriging and Co-kriging which considered graphic location, slope, aspect and height. By analyzing the rainfall and snow character in the west of Tianshan Mountain area and eight day artificial snow data of EOS/MODIS, SanShui Yuan Xi’anjiang Model is improved. Distributed hydrological model is built, which could reflect basin character of the west of Tianshan Mountain area based on the conceptual hydrological model of snowmelt type of Xi’anjiang Watershed Model.In order to verifying the feasibility of the model, using the Distributed hydrological model simulate the daily Runoff process based on remote measuring and hydrological station data which daily rainfall data, daily average temperature data and daily evaporation data of the area of May, 2005.By comparison the simulate result and the observed runoff founding that the Distributed hydrological mode is feasible and obtains better effect. The qualified rate of month water quantity is 100%, the qualified rate of daily water quantity is 76.5%, the deterministic coefficient of forecasting scheme is 81.2 and the efficiency of forecasting scheme is class 2.更多还原