基于复杂网络的气象网络分析

Analysis of Meteorological Network Based on Complex Network

在复杂网络研究方法的基础上,针对气象要素间相关性关系,提出基于时间序列的温度气压相关性网络研究方法以及转换模式的概念。实验结果表明,大多数城市相关性网络中夏季的平均路径长度要大于其他季节,即相关性序列间的变化的复杂性越强而季节因素对不同城市的网络直径、网络平均聚类系数的影响不明显。虽然不同城市地理位置的不同,但网络的平均度、平均聚类系数、网络直径和平均路径长度都较好的保持一致,波动范围都比较小,表明中国气候网络的整体特性随地域变化具有整体性。

For the complex correlation between meteorological elements,a method about correlation between temperature and pressure and a conception of transformation patterns were put forward based on Complex Networks research methods in this paper.Experimental results show average path in summer is longer than other seasons in most correlation networks.In other world,this change of the correlation symbol sequence is more complex in summer than others.However,seasonal factor＇s effects on network diameter and clustering coefficient in different cities are not obvious.The network average degree,clustering coefficient and other topological properties are consistent in different city,which indicates that the overall characteristics of the China climate network have integrated with geographical variation.