若尔盖湿地黑河径流分析及预测  

Analysis and Prediction of Heihe River Runoff Characteristics in Ruo'ergai Wetland

作　　者：周泽江[1] 覃光华[1] 于春平[1] 张立 

机构地区：[1]四川大学水利水电学院,四川成都610065 [2]佛罗里达海湾大学大沼泽湿地研究园,佛罗里达州那不勒斯34112

出　　处：《水电与新能源》2013年第3期18-22,共5页Hydropower and New Energy

基　　金：国家自然科学基金项目(41071203);中国科学院成都山地所基金项目

摘　　要：根据若尔盖湿地若尔盖水文站逐日平均流量资料(1988～2008年),应用滑动平均法、不均匀度、变化幅度以及自相关系数分析了若尔盖湿地黑河径流年际变化以及年内分配规律,并采用门限回归模型以及最近邻抽样回归模型对日径流进行了拟合和预测。研究表明,若尔盖黑河径流量年内分配十分不均,年内变化曲线呈现双峰型;日径流短期内具有良好的相依性;且年径流量自1988年以来具有明显的减小趋势;门限回归模型、基于时间序列分析的最近邻抽样回归模型用于逐日平均流量预测,效果较好。According to data of daily average runoff （ 1988 - 2008 ） of Ruoergal hyclrologlcat stauon m the rtuoergal wet-land, the runoff variation between years and distribution law within the year of Ruoergai wetland in Heihe River are ana-lyzed by applying moving average method, non-uniformity method, change amplitude and autocorrelation analysis. Daily runoff is fitted and forecasted by using threshold regression model and nearest neighbor bootstrapping regressive model. The research shows that Ruoergai Heihe runoff annual distribution is very uneven and double-peak pattern is shown in variation curve of the year; daily runoff has high dependence in the short term; the annual runoff has obviously de-creased trend since 1988 ; average daily runoff can be effectively forecasted by application of threshold regression model, the nearest neighbor bootstrapping regressive model based on time series analysis.

关 键 词：若尔盖湿地黑河 时间序列分析 径流特征 最近邻抽样回归模型 门限自回归模型 

分 类 号：TV121.4[水利工程—水文学及水资源]