STRESS TRANSFER TO LARGE DISTANCES ALONG SEISMIC ZONES

Within less than a minute to a few hours seismic activity began in several areas up to 90 km to the west and to the north of the June 17 main shock. (Figure 1). 3 earthquakes of magnitude 4.5-5 occurred on the Reykjanes peninsula, one at 60 km distance, 27 seconds after the mainshock and two at 80-85 km distance, 30 seconds and 5 minutes respectively after the main shock (Kristín S.Vogfjörð 2002, personal communication).

The IMO runs a network of continuous GPS (CGPS) stations in the southern part of Iceland (Árnadóttir et al. 2000). At the time of the earthquakes 3 CGPS were in operation 70-100 km to SE and 4 in operation 30-60 km to W of the earthquake epicenters (Figure 4). Most of the stations show coseismic signals (Geirsson 2003; Figure 4) which are larger than expected from earthquakes as modelled above, assuming an elastic halfspace environment, which seems to be justifiable in modelling strain fields as can be expected from the individual large events. The observed CGPS signals can be explained by general left-lateral motion of a few centimeters across the SISZ and RP plate boundary and a significant component of NS expansion across this boundary (Figure 4). The relative displacement recorded on June 17 between VOGS, which is located south of the westward prolongation of SISZ, and REYK, which is located to the north of this plate boundary, is of special interest (Figure 1 and Figure 4). VOGS moved 2 cm east and 1 cm south within a few hours of the main ahock (Figure 4). An alternative explanation for this large signal movement of VOGS could be that the large signal was caused locally by combination of seismic and aseismic motions. There was intense seismic activity along a NS fault close to VOGS immediately following the 17 June earthquake. The other GPS signals may of course have been modified by comparable local strain release events (mostly aseismic) triggered by the main shocks. In spite of such possible modifications of the signal the most likely explanation is an aseismic/seismic strain episode in a huge area following the June 17 earthquake.

By comparing InSAR images before and after the June 2000 earthquakes deformation was detected 80-85 km west of the June 17 main earthquake comparable to what would be expected from a N-S right-lateral magnitude (moment) 6 earthquake in that location, i.e. one meter slip on a 5 km long and 7 km wide fault (Clifton et al. 2003; Pagli et al. 2003). As the largest earthquake at this location was magnitude 5 (Mb) this was largely an aseismic event.

 

Figure 4: Above: Observed horizontal coseismic displacements related to the June 17 (black arrows) and the June 21 (blue arrows) earthquakes, assuming that REYK is fixed. Large black star notes the location of the June 17 main event and smaller black stars denote subsequent earthquakes in the magnitude range 4.5-5. The large blue star denotes the epicenter of the June 21 earthquake. Red lines indicate the faults of the two large earthquakes. Below: Time series of east and north components of displacement of VOGS relative to REYK covering the time of occurrence of the two large earthquakes, denoted by vertical lines. (Geirsson 2003).

=

The seismic activity, the displacements of CGPS stations and the deformation detected by InSAR probably indicate fast redistributions of strains along the EW elongated SISZ and RP zones, the plate boundary, up to 100 km to the west of the June 17 main shock, reflected in displacements on faults and rifts close to fracture criticality. In some cases the displacements are directly caused by dynamic triggering. Similar effects are observed seismically to the north of the June 17 earthquake (Figure 1) where postseismic stresses also triggered earthquakes at sites close to fracture criticality.