利用复杂网络研究中国温度序列的拓扑性质

依据粗粒化方法，将中国1961—2002年逐日平均温度序列转化为由5个特征字符{R,r,e,d,D}构成的温度符号序列.以符号序列中的125种3字串组成的温度波动模态为网络的节点(即连续4d的温度波动组合)，并按照时间顺序连边，构建有向加权的温度波动网络，进而将温度波动模态间的相互作用等综合信息蕴含于网络的拓扑结构之中.对随机序列和Lorenz系统的混沌序列分别构建随机和混沌波动网络.计算三种网络的度与度分布、聚类系数、最短路径长度等动关键词：气候变化气候复杂网络拓扑结构

An approach to research the topology of Chinese temperature sequence based on complex network

To analyze the dynamics of the temperature data, using homogenous partition of coarse graining process, the series of Chinese daily main temperature from 1961 to 2002 is transformed into symbolic sequences consisting of 5 characters {R,r,e,d,D}. The vertices of the temperature fluctuation network is 125 3-symbol strings (i.e., 125 fluctuation patterns in durations of 4 days), linked in the network's topology by time sequence. It contains Integrated information about interconnections and interactions between fluctuation patterns of temperature in network topology. Random fluctuant network and chaos fluctuant network by using random sequence and chaos sequence of Lorenz system were built. We calculated the dynamical statistics of the degrees and the distribution of degrees, clustering coefficient and the shortest path length, and compared the difference of these three sequences from the view point of network. The result is that. The main vertices of temperature fluctuant network generally contain the 3 characters R, r and e, and corresponding the background of global warming, the feature of temperature fluctuation is mainly ascending. On all accounts, the temperature fluctuant network is similar with chaos network in its dynamical statistics, but is markedly different from random network, which reflects the complexity of temperature fluctuation. And besides, some vertices in temperature fluctuant network have high betweenness centrality (BC), 4% of vertices bear 71.9% of betweenness centrality of networks, these vertices of importance in topological statistics are helpful to understanding the inherent law and information transmission. The vertices' BC in chaos network is similar with that of the temperature network, but vertices in random network almost have identical betweenness centrality. As a result, the patterns of temperature fluctuation corresponding to the process of weather change have similar property with chaos rather than the random fluctuation process.