玛纳斯河中长期径流预报研究  

Research on the Manas River's Mid-and Long-term Runoff Hydrological Forecasting

作　　者：靳晟[1] 雷晓云[1] 李慧[1] 

机构地区：[1]新疆农业大学水利与土木工程学院,乌鲁木齐830052

出　　处：《南水北调与水利科技》2010年第4期87-90,151,共5页South-to-North Water Transfers and Water Science & Technology

基　　金："十一五"国家科技支撑计划项目"水资源与水灾害空间信息服务系统研发"(2007BAH12B03);新疆水利水电工程重点学科基金项目(XJSG2007SW01)

摘　　要：随着统计学的不断完善,利用时间序列和回归分析理论来预报各种水文变量的未来变化已得到越来越广泛的应用。本文基于1955年-2008年玛纳斯河实测年径流量、蒸发量、平均气温资料分别建立了时间序列模型和多元回归模型,并进行了预报拟合及精度检验。结果表明玛纳斯河时间序列模型虽然预报精度较低,但有较好的适应性;而多元回归模型精度较高,完全可以达到实用要求,可为新疆北疆各地区抗旱、年用水计划、水库的兴利调度等提供较好的参考。Along with the gradual development of statistics, both the time series and the regression analysis theory have been more and more widely used to forecast the future change tendency of all kinds of hydrology variables. In this article, based on the measured data such as annual runoff, evaporation, average temperature of Manas River from 1955 to 2008, the time sequence model and the multi-dimensional stepwise regression model have been established respectively which are applied to forecast. The accuracy of the two models have been tested, which indicates that although the Manas River time sequence model＇s forecast precision is low, it has good compatibility; however, the multi-dimensional stepwise regression model＇s precision is good and definitely meet the practical requirements, and may be referred by anti-drought, the annual water use plan and reser voir operation for the Xinjiang northern border areas.

关 键 词：时间序列 多元回归 模型 径流 预报 

分 类 号：TV121[水利工程—水文学及水资源]