Search

In this paper, we construct complex weighted networks for seismicity based on space–time influence domain for three different regions and study the properties and correlations present in the networks. It is shown that links with high weight take a certain proportion in the networks. The results show that degree distribution, weight distribution, and nodal strength distribution of the weighted networks all obey power-law distribution for the three regions. Both the unweighted and weighted networks are found to be assortative mixing and hierarchical. On observation of the evolution of the average nodal strength and the strength of node that contains main shock, we find that they show an increasing trend before main shock from the data of California. The weighted form of earthquake network may be helpful to study the interactions between seismic activities and to model seismicity by complex networks. • We construct complex weighted networks for seismicity based on space–time influence domain. • Degrees, link weights, and nodal strength follow power-law distribution. • The unweighted and weighted networks are both assortative mixing and hierarchical. • Nodal strength shows an increasing trend before main shocks for California. [ABSTRACT FROM AUTHOR]

Seismic activities show a space-time symmetry in some research. They have been recently studied using complex network theory. Earthquake network similarity is studied by us from seismic catalogs in the same region for a given period of time. In this paper, we first calculate the distance between feature vectors which represent the topological properties of different networks. A hierarchical clustering of earthquake networks in the same region is shown by using this method. It is found that similar networks are not the networks of adjacent years but those with decades time difference. To study the period of similar earthquake networks in the same region, we use wavelet analysis to obtain the possible periods at different time scales of the regions of the world, California and Japan. It is found that some of the possible periods are consistent with the results which have been already found by seismologists. The study of similar seismic activities from the perspective of the complex network will help seismologists to study the law of earthquake occurrence in a new way, which may provide possible research thinking for earthquake prediction. [ABSTRACT FROM AUTHOR]