岷江都江堰降雨与径流变化趋势分析及预测

Analysis and Prediction of Variation Tendency of Precipitation and Runoff at Dujiangyan Section,Min River Catchment

依据岷江上游紫坪铺水文站监测1955—2008年降雨量与年径流量资料为基础数据,用线性倾向法、滑动平均法、Mann-Kendall法进行系列趋势检验,采用小波分析进行序列周期性分析,用ARIMA模型法构建降雨径流预测模型。结果表明：近54 a来,岷江流域都江堰段降雨量与径流量均有下降趋势,其中降雨下降较快,降雨的突变点比径流少且二者存在6～21 a的周期,得到的ARIMA模型预测精度较高,结合序列趋势变化与周期性,可得到该区未来降雨径流更合理的预测。

According to data on the precipitation and runoff（ 1933-2008 ） from Zlpmgpu rlydrologic 3tatlon at the upstream ot the Min Rive as well as by application of the linear tendency method, moving average method and Mann - Kendall method, a series of tendency tests are performed. Morlet wavelet is applied to analyze its periodicity and ARIMA model method is utilized to build the prediction models of precipitation and runoff. The results indicate that, in past 54 a, both precipitation and runoff at Dujiang-yan section of the Min River decrease in tendency. Among them, precipitation decreases rapidly than runoff. And it has less breakpoints and is with a pe- riod of 6 - 21 a. Accuracy of the AR1MA model is high in predication. In combination with series tendency variation and periodicity, more reasonable prediction of future precipitation and runoff at this section can be available.