中文摘要：

气象数据是SWAT模型重要的输入数据之一,受当前大气观测技术的局限,实测气象要素数据存在一定的不确定性,这种不确定性会影响模型模拟结果的精度。以黑河上游莺落峡流域为例,通过蒙特卡洛采样获得服从正态分布的随机数,在实测数据上加入随机误差来表示数据的不确定性,进而探讨气象数据输入不确定性对SWAT模型径流模拟结果的可能影响。结果分析表明：总体上,输入的气象要素数据不确定性对模型模拟结果影响较大;其中降水数据的不确定性对模型模拟结果影响最大;同时发现在模拟过程中考虑气象数据本身的误差影响,可有效提高径流模拟的精度。

英文摘要：

Meteorological data is important input in SWAT models. Due to the limitations of meteorological observation technology,the uncertainty of measured meteorological data exists,influencing the model simulation accuracy. As for the Yingluoxia Basin in the upper Heihe River,the random number following normal distribution are obtained by Monte Carlo sampling and the measured data is considered with the random error to express the data uncertainty,thus the influence of meteorological data uncertainty on the runoff simulation accuracy by SWAT model is studied. The analysis results show that,in general,the uncertainty of meteorological data input has a large influence on the model simulation result,in which,the uncertainty of precipitation data has the largest influence; meanwhile,it is discovered that if the self- error of meteorological data is considered in the simulation process,the runoff simulation accuracy can be improved effectively.