EARTHQUAKE SWARMS IN THE KAMCHATKA-COMMANDER REGION

Citation
Vm. Zobin et Ei. Ivanova, EARTHQUAKE SWARMS IN THE KAMCHATKA-COMMANDER REGION, Geophysical journal international, 117(1), 1994, pp. 33-47
Citations number
23
Categorie Soggetti
Geosciences, Interdisciplinary
ISSN journal
0956540X
Volume
117
Issue
1
Year of publication
1994
Pages
33 - 47
Database
ISI
SICI code
0956-540X(1994)117:1<33:ESITKR>2.0.ZU;2-A
Abstract
A swarm sequence is an earthquake sequence where no single event is do minant in size. In this paper we define a 'swarm' as an earthquake seq uence where the energy difference between the largest shocks of the se quence is Delta K-F less than or equal to 0.5 (K-F is an energy class of earthquake and represents an estimate of the energy of earthquakes in Log units of Joules for the frequency range 0.5-2.0 Hz. The relatio n between K-F and magnitude MLH is K-F = 0.85 MLH + 8.3. Magnitude MLH =1.3 M(s)-1.3 for the Kamchatkan earthquakes.) We analysed swarm seque nces if they contained five or more events of the energy class K-F gre ater than or equal to 8.5 (MLH greater than or equal to 1.0) occurring three days either side of the largest event (or, when more than one e vent happened with the same magnitude as the largest, the first of the se)-termed a 'clue' event-and were located not further than 50 km from this clue event. These criteria allowed us to select 32 swarms with c lue events of K-F greater than or equal to 11.5 (MLH greater than or e qual to 4.0) in the period 1962-1989 in the Kamchatka-Commander Region . The majority of swarms (86 per cent) were located within structures subject to horizontal tension (mainly on the slopes of the Aleutian an d Kurile-Kamchatka trenches and in zones of modern volcanism). At the same time the appearance of earthquake swarms did not depend on the fa ult nature of the clue event in that swarm. We studied the relationshi ps between the swarm characteristics (swarm area, duration and number of events) and the clue-event parameters (energy class, magnitude and stress drop). The regression equations were computed for parameters wi th significant correlation (at 95 per cent confidence level). All thre e swarm characteristics depend on the energy class of the clue event. Magnitudes MLH are correlated with swarm area and duration. The clue-e vent stress drop gives a negative correlation with the number of event s in a swarm.