MMPP Models of Background Seismicity: A Comparative Analysis for Different Declustering Algorithms

Declustering a seismic catalog is a relevant preliminary step in many applications, such as earthquake forecasting and seismic hazard assessment. Declustering aims partitioning an earthquake catalog into background seismicity, which is supposed to reflect the steady tectonic loading, and clustered seismicity, which is formed by dependent events, possibly reflecting the transient changes in the stress field in the aftermath of moderate and large earthquakes occurrence. Accordingly, various methods have been proposed in the literature for declustering seismic catalogs; however, their application to a given catalog may discriminate differently between background and clustered events. Hence the need to compare the different declustered versions of a catalog, with the aim of identifying common features relevant for background seismicity modeling. In this study we compare the background time series obtained from two data-driven declustering algorithms: the nearest-neighbor (Zaliapin and Ben-Zion, J Geophys Res, 2013), which classifies the earthquakes on the basis of a nearest-neighbor distance between events in the space-timeenergy domain, and the stochastic declustering (Zhuang et al., J Geophys Res, 2004), which is based on the space-time ETAS point process model. The steps of our analysis are as follows (Benali et al., Stoch. Environ Res Risk Assess, 2020). Once the background sequences are obtained from the two declustering methods, we investigate if they meet the stationary Poissonian assumption by performing suitable statistical tests. In case the simple Poissonian hypothesis is rejected, then we resort to a model capable to capture the possible heterogeneity in the background time series. Specifically, we propose the Markov modulated Poisson process, which allows the Poisson seismicity rate to change over time, according to a finite number of states of the system. We exemplify the analysis for two different seismic regions: North-Eastern Italy and Central Italy.