Abstract
Abstract. In this paper we propose a tri-stage cluster identification model that is a combination of a simple single iteration distance algorithm and an iterative K-means algorithm. In this study of earthquake seismicity, the model considers event location, time and magnitude information from earthquake catalog data to efficiently classify events as either background or mainshock and aftershock sequences. Tests on a synthetic seismicity catalog demonstrate the efficiency of the proposed model in terms of accuracy percentage (94.81% for background and 89.46% for aftershocks). The close agreement between lambda and cumulative plots for the ideal synthetic catalog and that generated by the proposed model also supports the accuracy of the proposed technique. There is flexibility in the model design to allow for proper selection of location and magnitude ranges, depending upon the nature of the mainshocks present in the catalog. The effectiveness of the proposed model also is evaluated by the classification of events in three historic catalogs: California, Japan and Indonesia. As expected, for both synthetic and historic catalog analysis it is observed that the density of events classified as background is almost uniform throughout the region, whereas the density of aftershock events are higher near the mainshocks.
Highlights
IntroductionC5. Regular Magnitude Zone C6. Danger Magnitude Zone Aftershock3.2 StagewIh:etreemMpo=r5alisctlhuestneurmibdeernotfifimcaaintsiohoncks of catalog. 245 From fthroism, ththeesycnluthsetetirccceantatleorgs events are classified into 5 clusters (represented by red, yellow, cyan, magenta and greenThe temp2.orFaolrceluacshteproiidnetnptiifi∈cPatNio×nDis(CcaatarrloiegdhoavuitnugsNingevteenmts-, poral clusteDrindgimaenndsitoinm)ecazlcounleatbeathseedEucclluidsteearnindgistaasncfeol(l2o-wnosr:m) colorsd) s(hpoiw, ξnji)n Fig. =D(botptoim,d l−eftξ).j,d (4)from the cluster centers– Time Zone Basedd=C1lustering3.2.1 Temporal clusteDring d(pi,ξj) = (pi,d − ξj,d)2
We present a different method to decluster seismicity, based on a tri-stage cluster identification model that only considers event location, time and magnitude of earthquake catalog data in an objective, mathematical formulation designed to efficiently classify events as either background or mainshock and aftershock sequences
For this work we considered a synthetic seismic catalog generated by a combination of a non-stationary Poisson process and the epidemic type aftershock sequence model (ETAS) model detailed above
Summary
C5. Regular Magnitude Zone C6. Danger Magnitude Zone Aftershock3.2 StagewIh:etreemMpo=r5alisctlhuestneurmibdeernotfifimcaaintsiohoncks of catalog. 245 From fthroism, ththeesycnluthsetetirccceantatleorgs events are classified into 5 clusters (represented by red, yellow, cyan, magenta and greenThe temp2.orFaolrceluacshteproiidnetnptiifi∈cPatNio×nDis(CcaatarrloiegdhoavuitnugsNingevteenmts-, poral clusteDrindgimaenndsitoinm)ecazlcounleatbeathseedEucclluidsteearnindgistaasncfeol(l2o-wnosr:m) colorsd) s(hpoiw, ξnji)n Fig. =D(botptoim,d l−eftξ).j,d (4)from the cluster centers– Time Zone Basedd=C1lustering3.2.1 Temporal clusteDring d(pi,ξj) = (pi,d − ξj,d)2
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
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
