中文摘要：

为了实现对新疆玛纳斯河流域的雪盖率预测,为该区融雪径流预报提供数据支撑,本文选取反向传播（BP）神经网络模型时间序列预测的方法对研究区雪盖率的预测问题进行了研究.首先基于2001-2012年研究区中分辨率成像光谱仪（Moderate Resolution Imaging Spectroradiometer,MODIS）积雪数据产品MOD10A2提取研究区雪盖并计算雪盖率,分析研究区雪盖率变化在时间上的自相关性,据此选取对预测值影响较大的相邻若干个历史时刻的雪盖率以及往年同期雪盖率的均值作为模型输入变量,建立神经网络模型,并选取相关系数与确定系数作为模型预测效果的评价指标.结果表明：以BP神经网络时间序列预测方法预测2012年积雪覆盖率,输入变量为与待预测雪盖率相邻的前期雪盖率值与往年同期雪盖率均值,预测值与实际值的相关系数为0.80,确定系数为0.61,预测结果基本反映了2012年雪盖率的年内变化趋势,可以为融雪径流模型（SRM）径流预报提供一定的数据支撑.

英文摘要：

Snow is an important factor of earth＇s surface,it has great influence on the cold and arid area,especially western China.The purpose of the study is to predict the snow cover fraction in Manasi River Basin of Xinjiang Province,where snow melt water is one of the most important freshwater resources.In order to achieve good prediction accuracy,this paper presented a prediction method based on back propagation（BP）neural network.The snow cover data used in this study was extracted from MODIS snow products MOD10A2 from 2001to 2012.To predict the snow cover fraction,several adjacent snow cover fractions of historic moment was selected by autocorrelation a-nalysis.Besides the adjacent snow cover fractions,the input variables of the BP network also included the average snow cover fraction of precious years.Meanwhile,the coefficient of correlation and the coefficient of determination between predicted value and the actual value were chosen to evaluate the prediction effect of the model.The results showed that the BP network could reflect the general trend of 2012 snow cover fraction,and the coefficient of correlation could reach 0.80,and the coefficient of determination was 0.62.This study may provide some support for the snowmelt runoff forecast.