双参数月水量平衡模型在汉江流域上游的应用  

A Two-parameter Monthly Water Balance Model and Its Application in Hanjiang River Basin

作　　者：乐通潮[1] 张万昌[1] 

机构地区：[1]南京大学国际地球系统科学研究所,南京210093

出　　处：《资源科学》2004年第6期97-103,共7页Resources Science

基　　金：国家重点基础研究发展规划项目(973)"中国水圈环境时空信息的应用系统与典型示范"(编号:2001CB309404)。

摘　　要：该文介绍双参数(C和SC)月水量平衡模型的原理,依据水量平衡方程和经验公式构建流域径流深度的模拟方程,利用计算机程序自动率定模型参数。利用1980年1月到1985年12月共6年的水文站点数据对汉江流域上游地区15个水文站点控制的子流域进行了模拟和校验,模型效率指数大都超过80%,其中模拟效果最好的是铁锁关,率定期Nash指数R2达到0 9923,校验期达到0 9377,而大竹河模拟效果最差,率定期Nash指数R2为0 5004,校验期达到0 7382,达到了较好的模拟效果。以典型的两子流域铁锁关和大竹河为例,对模型参数C、SC对模型效率指数的敏感性进行了分析,发现当C变化时,模型效率变化明显,而SC变化时,模型效率指数变化不太明显,但R2极大值附近总能找到相应的C和SC使得RE,REm的值也比较接近0,满足模型的要求,从而验证了模型参数C、SC是能够通过计算机程序自动率定。最后提出结合遥感与地理信息技术改进模型使之能在分布式模型中得到应用的思路。Brief principles on the two-parameter (C & SC) water balance model have been introduced in this paper. Furthermore, the function to simulate the depth of runoff using water balance function and empirical formula has been constructed and a program has been used to auto-calibrate the two unfixed model parameters (C & SC). Simulations and validations are conducted with the basic hydrological observation data collected from 15 stations at upper stream of Hangjiang River Basin. Simulations are attempted in monthly time steps and the time span for modeling is 6-year period from 1980 to 1985. Good simulation results have been gotten on Hangjiang catchments with most of the simulation efficiencies of the 15 stations are above 0.80. The best simulation result is on Tiesuoguan Station and Nash is 0.992 3 in simulation period and 0.937 7 in validation period. While the worst simulation result is on Dazhuhe Station and Nash is 0.500 4 in simulation period and 0.738 2 in validation period. The sensitivity of the two parameters (C & SC) of the model to simulation efficiencies is analyzed on the two typical sub catchments, Tiesuoguan and Dazhuhe. The result is that simulation efficiencies have changed obviously when C changed while almost unchanged when SC changed. However, C and SC could always be found close to the maximum of R^2 so that RE and RE_m are close to zero, and the model could be satisfied. As a result, it is proved that C and SC can be auto-verification by computer program. Finally, a scheme　that takes advantages of remote sensing techniques and GIS powerful management capability in a distributed manner to predict the water resource response to the possible climate and land use/cover variations under the global change scenarios are discussed and prospected.

关 键 词：双参数水平衡模型 模型效率 自动率定 汉江流域 

分 类 号：P343.1[天文地球—水文科学]