摘要
运用Mann-Kendall检验法与Morlet小波变换分析法对长江源区直门达水文站1953-2003年径流资料进行趋势和周期分析。根据径流与气候要素的关系建立了年径流预测的混合回归模型，构建方便、灵活、具有一定精度，能反映年径流量随降水、气温变化的规律。该方法比从径流序列本身趋势分析更具优势。
Based on the data of annual runoff of Zhimenda gauge records, the Mann-Kendall method and Molet wavelet transform were applied to the tendency and period analysis of runoff in the source region of the Yangtze River during 1953 to 2003. According to the relationship between runoff and meteorological elements, a mixed regression model was proposed and posterior error is 0. 4883, which shows the model reaches good grade. Therefore, the model is better than the simply tendency analysis to forecast the future runoff.
出处
《水电能源科学》
2008年第3期14-16,共3页
Water Resources and Power
基金
国家科技支撑计划基金资助重大项目(2006BAB04A08)
科技部科研院所社会公益研究专项基金资助项目(2005DIB3J057)
关键词
长江源区
径流
混合回归模型
source region of Yangtze River
runoff
mixed regression model
