The role of the cell shapes on the characteristics of a complex earthquake network

We studied the behavior of the earthquake network using the HEALPix spherical pixelization and square cells methods. In the first method, the geographical region is divided into isolatitude rhombic-spherical cells of the same areas using the HEALPix method. In the second method, we divided the geographical region into isolatitude equal areas of the square cells. To construct a network, if an earthquake happens in a cell, that cell will become a node, and two nodes will be connected with an edge for two successive events. The earthquake network is built from Iran’s seismic data from 1900 June 12 to 2015 December 12. We determined the Hurst exponent (H = 0.6) due to the rescaled range (R/S) analysis. This value reveals a long temporal correlation in earthquake time-series; therefore, the earthquake system is suggested to be self-organized. We showed that among the five major seismotectonic provinces of Iran (Alborz-Azarbayejan, Kope Dagh, Central-East Iran, Zagros, and Makran), the earthquake network hubs are located in the Zagros region, which is a seismically very active region. According to this result, the Zagros earthquakes affect the surrounding earthquakes. The probability distribution function’s power-law behavior with a network built in the pixelization rhombic-spherical cells shows scale free behavior’s properties than a network constructed based on the square cells. The mean clustering coefficient’s power-law nature with networks built using two methods shows that the earthquake network is scale-free and non-random. We concluded that the rhombic-spherical cell pixelization is a more reliable method for building the large geographical region’s earthquake network.