以黄河利津站54年(1950~2003年)的月径流序列为基础资料,以1970年为分界点,对比分析黄河下游1950~1969年和1970~2003年的各月径流序列和年径流序列的变化特征;并把1950~1969年和1970—2003年的各月径流序列的特征值和主周期作为自组织映射网络(SOM)的输入向量,分别建立小波分析-自组织映射网络(WA—SOM)耦合模型,根据其输出结果对比分析1970年前后黄河下游各月径流序列的变化情况。研究表明:1970年以后,黄河下游各月径流量明显减少,各月径流序列(除1月和12月)的CV和CS显著增大,整体来看,主周期变得更加复杂;并且在1970年之后,3月、4月、5月和6月的径流序列变得相似,汛期8月、9月和10月的径流序列特征差异很大。
In recent years, the impact is great of human activities and climate change on runoff especially in Yellow River, because the water resources system of Yellow River is fragile. The Yellow River's zero flow has been becoming one of the most serious environment problems in our country. Taking the 54 year( 1950 -2003 )month runoff series and annual runoff series obtained from Lijin hydrologic station as a basic data and taking 1970 as a demarcation point, a WA-SOM coupling mode was established by comparing the characteristics of month runoff series and annual runoff series of 1950 - 1969 with those of 1970 ~ 2003 to analyze characteristics of the lower Yellow River's runoff. The wavelet analysis (WA)method and the self-organized map (SOM)were used in establishing the model. The results show that for the lower Yellow River after 1970, the month runoff and the annual runoff series decreased significantly, while CV( coefficient of variation) and CS( coefficient of skewness) of the month runoff series (except January and December)and the annual runoff series increased remarkably. As a whole,the main periods of month runoff after 1970 became more complex than that of before 1970. After 1970, the change of runoff series of January and February was not obvious, but the change of runoff series of the other months was significant. The runoff series of March, April, May and June became similar, while the runoff series of August, September and October became different. This might be because of the increased water demand by the development of national economy,the overexploitation of groundwater, and changed water cycle by the hydraulic engineering construction. Of course, climate change is also an important factor.