基于变异诊断分析的大通河流量预报模型研究

Annual Average Flow Prediction Model Research of Datong River Basin Based on Hydrological Variation of Diagnostic Analysis

采用水文变异诊断方法对大通河享堂水文站1940年以来年平均流量序列进行了诊断,结果表明年平均流量变异显著,变异程度为中等,突变点为1990年,说明大通河年平均流量自20世纪90年代以来受人类活动影响发生了显著的突变。同时,在变异诊断分析的基础上,分别建立了基于跳跃变异和趋势变异的年平均流量预报模型,比较模型预报值与实测值可知,相对误差为±10.0%的年份,跳跃预报模型占89%,趋势预报模型占97%。采用所建模型分别预报2011—2020年平均流量,并用2011—2013年实测值进行验证,预报误差符合《水文情报预报规范》的规定,取得了较好的预报效果。

This paper carried on the preliminary diagnosis of the annual average flow series data in 62 years of the Datong River basin and detailed diagnostic by using the method of hydrological variation in the diagnosis. The results show that the sequence variation is significantly and jumps mutate, variation of medium and the most likely change point is 1990, which indicates that is influenced by human activities. The annual average flow of Datong River basin happened fact significant mutations since 1990s. Meanwhile, it established annual average flow forecast model based on jump variation and trend variation based on the variation of diagnostic analysis respectively. Comparing model predicted value and measured value, the relative error is less than ±10.0% of the year with the hopping model accounted for 89% and the trend model accounted for 97%. By using the model to forecast the annual average flow in 2011?2020, and to test it with measured values in 2011?2013, the prediction error conforms to"Standard of Hydrology Information Forecast" rules, which shows the forecast precision is higher and achieved good prediction effect.