Initial Data

Figure 1
Figure 2
Figure 3
[7]  We analyzed seismic records of the vertical component with discretization frequency of 20 Hz of wide-band stations IRIS before four strong earthquakes: Neftegorskoe earthquake of 27 May 1995 with coordinates [52.55oN, 142.75oE], M = 7.0; Kronotskoe earthquake of 5 December 1997 [54.64oN, 162.55oE], M = 7.8; Hokkaido earthquake of 25 September 2003 [41.81oN, 143.91oE], M = 8.3; Sumatra earthquake of 26 December 2004 [3.32oN, 95.85oE], M = 9.2. The data were kindly given by the Geophysical Service RAS. Horisontal component records were analyzed incidentally. In the studies of Kronotskoe and Neftegorskoe earthquakes we used records of stations PET, MAG, YSS, YAK, ARU, and OBN; their location is shown in Figure 1. For Hokkaido earthquake the system comprised stations ERM, MAJ, INC, MDJ, BJT, PET, YSS shown in Figure 2 and OBN station (Figure 1). For Sumatra earthquake, the system of stations CHTO, KMI, XAN, COCO, PALK, MBWA, DGAR, DAV, QIZ is shown in Figure 3. Stations selected for the studies were located at distances ranging from 70 km to 7160 km from the above-mentioned earthquakes and in different seismogeological conditions. The total volume of analyzed data was more than 24 Gb.

Figure 4
[8]  Amplitude-frequency characteristics of IRIS channels provide permanent sensitivity of recording the velocity of bases displacement in the period range of 0.3-357 seconds [Starovoit and Mishatkin, 2001]. Oscillations up to peaks of 12 and 25 hours caused by earth tides are reliably registered in spite of the sensitivity drop in the range of longer periods. A typical spectrum is shown in Figure 4. We note three intervals of period values in the spectrum plot. In the short-period range up to 6 minutes, a drop in oscillation strength is observed that is a result of gradual decrease in the influence of microseisms of oceanic origin and weak earthquakes. In the period range of hundreds of minutes, the influence of earth tides is noted which is corroborated by peaks of 1440 and 720 minutes corresponding to diurnal and semidiurnal oscillations. We analyzed microseismic signals in the relatively low-noise minute range marked with an arrow in Figure 4.


Powered by TeXWeb (Win32, v.2.0).