中文摘要：

以雅鲁藏布江奴下水文站以上流域为研究对象,针对缺资料流域的水文计算和预测问题,采用流域水文模型THREW,用地面气象观测、遥感植被覆盖和积雪面积等资料,基于断面水文监测数据对模型进行率定,应用CMIP5数据对径流演变进行预估。结果表明：对于雅鲁藏布江奴下水文站以上流域,THREW模型对1991—1995年率定期月径流模拟的纳什效率系数为0.75,对1996—2000年验证期月径流模拟的纳什效率系数为0.76;IPCC AR5所设置在CO2排放量最大的情况下,径流明显增加。

英文摘要：

Considering the hydrological calculation and prediction of ungauged basins,a distributed hydrological model,the THREW model,was adopted to study the upper basin of the Brahmaputra River from the Nuxia Hydrological Station. Surface meteorological observation,remote sensing of vegetation cover,snow cover area,and section hydrological monitoring data were used to calibrate the model,and the latest CMIP5 data were used to estimate the runoff evolution. The results showed that,in the upper basin of the Brahmaputra River from the Nuxia Hydrological Station,the Nash-Sutcliffe efficiency coefficient of monthly runoff simulation of the calibration period from 1991 to 1995 was 0. 75,the coefficient of monthly runoff simulation of the validation period from 1996 to 2000 was 0. 76,and the runoff markedly increased when the CO2 discharge in the AR5 of IPCC was maximal.