中文摘要：

卫星遥感技术和水文模型的结合应用是当今水文学研究中的热点问题。但是,直接利用遥感信息提取水文模型相关参数来模拟径流过程存在较大的不确定性。为了分析其对径流模拟不确定性的影响,构建了利用遥感信息估计新安江模型3个产流计算参数的框架：基于土壤属性推求WM,基于流域几何特征推求B,基于土地覆被分类和Landsat遥感影像推求IMP;并采用GLUE方法在高关水库流域作了长系列径流模拟研究。传统的新安江模型参数需要根据实测资料进行率定,而在新的参数估计框架下,需要率定的参数减少了3个。模拟结果表明：新的参数估计框架减少了参数的不确定性,计算得到的模拟径流90%置信区间基本包含所有的实测数据,且区间范围明显小于传统方法,尤其是在低流量的模拟结果上差别更加明显,即基于遥感信息的参数估计框架显著降低了径流模拟的不确定性,这种方法可以应用到其他相似的流域中进行径流模拟。

英文摘要：

Combining satellite remote sensing technology with hydrological model is a hot issue of hydrology research at present, but uncertainty exists in basin runoff process simulations using model parameters extracted directly from the remote sensing information. To analyze the uncertainty, this study has constructed a framework for estimation of three runoff calculation parameters in the Xinanjiang model using remote sensing information. In this framework, WM is calculated indirectly from soil properties, B derived from basin geometry characteristics, and IMP directly extracted from land cover classification and Landsat remote sensing image. We also conducted a long-term runoff simulation for the Gaoguan reservoir basin （1982-1989） with the GLUE method. Traditionally all the parameters in the Xinanjiang model require calibration with measured data, while in the new framework three of them are free of calibration, thus lowering the mutual interference of parameters. Results show that compared to the traditional method, the uncertainty in parameter estimation can be reduced and the uncertainty in runoff simulations using the parameters estimated with the new method, is significantly lower, especially for low flow areas. The new method can be applied to stream flow simulation for other similar watersheds.