ORDER CLUSTERING IN THE SEQUENCES OF VOLCANIC ERUPTIONS AND EARTHQUAKES OF THE WORLD
Any time-ordered list of events of variable size or «weight » can manifest order clustering, or an inherent tendency of the largest events of such a list to occur «too often » as close neighbors in the list; i.e. significantly more often than should be in a similar randomly-shuffled event list. Accu...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Text |
| Language: | English |
| Subjects: | |
| Online Access: | http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.519.2051 http://kiska.giseis.alaska.edu/kasp/kasp04/abstracts/gusev.pdf |
| Summary: | Any time-ordered list of events of variable size or «weight » can manifest order clustering, or an inherent tendency of the largest events of such a list to occur «too often » as close neighbors in the list; i.e. significantly more often than should be in a similar randomly-shuffled event list. Accurate timing is irrelevant for order clustering, in contrast to «common» clustering when events (whose size is inessential) occur as tight, compact temporal groups or bursts. Recently, order clustering was revealed for the sequence of Holocene volcanic eruptions on Kamchatka. (Gusev a. o. 2002). In this study I look for order clustering behavior in global volcanic data sets. I analyzed the global catalogue of volcanic eruptions kindly provided by Lee Siebert (Siebert&Simkin 2002, labeled SMI). The most reliable and uniformly reported part of this catalogue is that with VEI≥3 in 1960-2002. Order clustering is present here, with the statistical significance level Q=5%. Unexpectedly, common clustering (i.e., the formation of tight temporal groups) is absent for this data subset. However, common clustering was found for |
|---|