Higher order Gibbs point process modeling of 2005-Kashmir earthquakes

Data on the locations of earthquake epicentres serve as a realization of a spatial point process. Since earthquakes cluster in space, their locations are thought to exhibit spatial interaction. A suitable form of the conditional intensity function of the point process is, therefore, required for modeling the clustered behaviour of the earthquakes epicentres. Due to limitation of the general pairwise Gibbs point processes in modeling the clustered patterns, the higher order Gibbs point processes, namely area-interaction model and Geyer’s saturation model, are fitted to describe the spatial distribution of the 2005-Kashmir earthquake and its aftershocks in Northern Pakistan. For an improved assessment of the seismicity of the study area, available information on geological and geophysical attributes of the area is incorporated in the models. The fitted Gibbs models prove their effectiveness in modeling the interaction structure and describing the spatial characteristics of the 2005-Kashmir earthquake and its aftershocks distribution.