Abstract
We study the spatial distribution of clusters associated to the aftershocks of the megathrust Maule earthquake MW 8.8 of 27 February 2010. We used a recent clustering method which hinges on a nonparametric estimation of the underlying probability density function to detect subsets of points forming clusters associated with high density areas. In addition, we estimate the probability density function using a nonparametric kernel method for each of these clusters. This allows us to identify a set of regions where there is an association between frequency of events and coseismic slip. Our results suggest that high coseismic slip is spatially related to high aftershock frequency.
Highlights
Recent applications of clustering techniques range over an enormous set of disciplines, both in the natural sciences and in the social sciences
We have considered 6,714 aftershocks in a map [32–40◦S]×[69–75.5◦E], for a period between 27 February 2010 and 13 July 2011 and for local magnitudes Ml ≥ 2.0 (ContrerasReyes and Arellano-Valle, 2012). All of these observations have been pre-processed with SEISAN 8.3 software starting from the information provided by 22 stations located in a map with coordinates [−33.32, −39.80] latitude and [−70.29, −73.24] longitude
The most recent major Chilean earthquake occurred on February 27th, 2010 (Mw = 8.8) filled a seismic gap (Ruegg et al, 2009) that has experienced little seismic activity since 1835, when it broke with an estimated magnitude of Mw ∼8.5 (Darwin, 1851, p. 768)
Summary
Recent applications of clustering techniques range over an enormous set of disciplines, both in the natural sciences and in the social sciences. Relatively more recent approach is the model-based clustering formulation which regards the observed data as generated by a probability distribution of finite mixture of multivariate random variables having distribution belonging to some parametric family. In the implementation of this approach, the most common option is to adopt the multivariate Gaussian assumption for each of the density components, and estimate their parameters using an EMtype algorithm. An useful account to this approach is provided by Dasgupta and Raftery (1998). By applying the model-based clustering approach to earthquake data in the coastal area of central California, Dasgupta and Raftery (1998) have obtained six clusters, some of which are clearly linked to active faults. One or two of their clusters do not correspond to some already identified area
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
