程晓光,张静,宫辉力.半干旱半湿润地区HSPF模型水文模拟及参数不确定性研究[J].环境科学学报,2014,34(12):3179-3187
半干旱半湿润地区HSPF模型水文模拟及参数不确定性研究
- HSPF hydrologic simulation and parameter uncertainty in a semi-arid and semi-humid area
- 基金项目：国家自然科学基金(No.41271004,40901026);北京市科技新星项目(No. 2010B046)
- 程晓光
- 1. 城市环境过程与数字模拟国家重点实验室培育基地, 首都师范大学, 北京 100048;2. 三维信息获取与应用教育部重点实验室, 首都师范大学, 北京 100048
- 张静
- 1. 城市环境过程与数字模拟国家重点实验室培育基地, 首都师范大学, 北京 100048;2. 三维信息获取与应用教育部重点实验室, 首都师范大学, 北京 100048
- 宫辉力
- 1. 城市环境过程与数字模拟国家重点实验室培育基地, 首都师范大学, 北京 100048;2. 三维信息获取与应用教育部重点实验室, 首都师范大学, 北京 100048
- 摘要：本研究选取半干旱半湿润地区北京妫水河流域2005—2007年和2008年月径流数据为率定期和验证期,建立HSPF水文模型进行径流模拟, 结合人工率定和PEST自动率定程序进行参数优选,并通过GLUE方法分析模型参数不确定性.通过Monte-Carlo随机采样得到30000组参数组合,分析参数与似然值散点图,把参数分为敏感参数(LZSN、AGWRC)、区间敏感参数(BASETP)和不敏感参数(AGWETP、INFILT、CEPSC、DEEPFR、UZSN、INTFW、IRC).针对比较敏感的参数LZSN、AGWRC和BASETP分析其相关性,发现LZSN和AGWRC相关性较强.模型存在大量"异参同效"现象,表明影响结果的是参数组合而不是单一参数.进一步计算90%置信度下的不确定性范围,发现不确定性范围与径流大小密切相关,径流愈大其不确定性范围愈大,反之亦然.本文对参数不确定的分析研究可为HSPF模型在区域尺度水文预测等提供参考和依据.
- Abstract：HSPF model was built for Guishui River Basin, Beijing, a semi-arid and semi-humid area. The years of 2005—2007 were selected as the calibration year and 2008 as the validation year in this study. Model parameters were calibrated both manually and automatically, and the results were reasonable. The Generalized likelihood uncertainty estimation (GLUE) approach was then applied to analyze parameter uncertainty. 30000 parameter samplings were generated with Monte-Carlo method. By using scatter distribution of likelihood of parameters, ten parameters were classified as sensitive parameters (LZSN and AGWRC), regionally sensitive parameters (BASETP), and non-sensitive parameters (AGWETP, INFILT, CEPSC, DEEPFR, UZSN, INTFW and IRC). Strong correlation was found between LZSN and AGWRC. Different parameter sets could gain the same likelihood, and the results were influenced by parameter sets instead of a single parameter. Besides, uncertainty ranges at 90% confidence level was selected to test the probability prediction for discharge, and the results showed that uncertainty ranges had a significantly positive relationship with flows. This study provides great significant references and suggestions for future prediction of regional-scale hydrology with HSPF model.
摘要点击次数： 1366 全文下载次数： 3733
