摘要
应用SWAT(Soil and Water Assessment Tool)模型,研制了黄河万家寨水库流域分布式降雨径流模拟系统,并对万家寨水库流域进行了10年(2000-2009年)径流模拟。其中,以2000-2006年为系统参数的敏感性分析期和模型率定期,以2007-2009年为模型的验证期,模拟结果显示,万家寨站点上模拟径流与实测径流过程线的相关系数达到0.97以上。在此基础上还将数值预报模式与模拟系统相耦合,对万家寨水库入库径流进行了预报和检验分析,结果较好。因此,该模拟系统能够较好地反映万家寨水库入库径流量的变化过程,具有一定的适用性,为进一步进行降雨径流预测提供了技术支撑。
In this paper, using 1 : 1 000 000 soil data set, the soil attribution database in the Wanjiazhai basin of the Yellow River Was built through the soil texture conversion and calculating a part of soil parameters by SPAW proce- dure. Using 1:100 000 land use/cover data,the vegetation type in the study area was divided into nine,which were mixed forest land, pasture, water, residential, residential-low density, industrial, bare land, combining wetland and agriculture. Basing on that, the vegetation type database was builL Using meteorological observed data of 7 meteoro- logical stations from 1977 to 2006 in study area, weather generator were generated by statistical method. Using meas- ured daily runoff data on the hydrological stations of Toudaoguai and Wanjiazhai and meteorological observed data of 7 meteorological stations from 2000 to 2009 in study area, hydrological and meteorological data sets were built, Also Using DEM (digital elevation model) and GIS/RS technology, the data of water cycle space distribution was ob- tained ,522 HRU (Hydrologic Response Units) were divided at last. On the basis of all these, the hydrological mod- el based on the SWAT (Soil and Water Assessment Tool) model in the Wanjiazhai basin of the Yellow River was built. Also month and day runoff was simulated by SCS runoff method with the actual measurement data of day runoff on the hydrological stations of Toudaoguai and Wanjiazhai from 2000 to 2009 ,in which 2000 to 2006 was taken as the calibration time and 2007 to 2009 was taken as the validation time. In the calibration time ,47th sub-basin was chose, namely incoming flow of Wanjiazhai, the systematic sensitive analysis and calibration were taken by LH-OAT algorithm which based on LH (Latin Hypercube) sampling and OAT (One-factor-At-a-Time) sensitivity analysis. Through sensitive analysis and calibration of month scale, 10 system parameters were chose, which were Alpha_Bf, Ch_K2, Ch_N, Cn2, Esco, Gw_Delay, Gw_Revap, Gwqmn, Sol_Awc and Sol_K. And through which of day scale, 6 system parameters were chose, which were Alpha_Bf, Blai, Ch_K2, Cn2, Esco and Sol_Awc. Also, the simulation sys- tem was verified according to measured daily runoff data from 2007 to 2009. Through analysis of simulation, calibra- tion and verification to month/day runoff process in the Wanjiazhai basin of the Yellow River from 2000 to 2009, at the same time, uncertainty of the incoming flow which was induced by the man-made factors such as dam, poling,levees, irrigation, etc was considered, the results showed that this system on the structure was reasonable, the correla- tion coefficient reaches more than 0.97and Nash - Sutcliffe model efficiency coefficient reached more than 0.96 not only in the calibration time but also in the validation time. The result was good in the simulation process of 2000 to 2009, which could reflect the change of runoff in the study area. And the effect of month runoff simulation is better than that of day runoff simulation. In addition, putting the numerical forecast model and the simulation system cou- pied, runoff forecast and analysis were done through taking elements forecast of T639 numerical forecast model as inputs of meteorological elements of this simulation system, the result was still good. In general, the simulation sys- tem could reflect the inflow runoff process perfectly. Therefore, the simulation result could provide a positive refer- ence of comprehensive utilization of water resources and proper reservoir operation for the Wanjiazhai basin of the Yellow River.
出处
《干旱区地理》
CSCD
北大核心
2012年第5期717-723,共7页
Arid Land Geography
基金
内蒙古气象局科技创新项目(nmgxkicx201012)资助
关键词
降雨径流
敏感性分析
率定
数值预报模式
气象因子
rainfall runoff
sensitivity analysis
calibration
numerical forecast model
meteorological elements