MVO/VSG - Open Scientific Meeting
27 November 1996
Three-dimensional P-wave velocity model for
Soufriere Hills Volcano, Montserrat, W.I.
Antonio Villasenor1, H.M. Benz1, J.A. Power2
Antonio Villasenor1, H.M. Benz1, J.A. Power2
1USGS, PO BOX 25046, MS 966, Lakewood, CO 80255, USA
2Alaska Volcano Observatory, USGS, 4200 University Dr, Anchorage, AK 99508, USA
A preliminary three-dimensional P-wave velocity model for the Soufriere Hills Volcano has been developed using data collected from the Montserrat Volcano Observatory (MVO) seismic network during August and September of 1995. First arrival times are calculated using a finite difference technique, which allows for a flexible parameterization of the slowness model. The 3D P-wave velocity structure and refined earthquake hypocentres are determined simultaneously. The method is well suited for volcanic environments because of its robustness in handling large topographic variations and large local velocity variations. For a detailed discussion of this technique see Benz et al., (Three-dimensional P and S wave velocity structure of Redoubt Volcano, Alaska, JGR, v.101, n. B4, p.811-8128, April 10, 1996).
A comparison of 1D (HYPO71PC) and relocated 3D hypocentres is shown in Figure 1 (A and B). NW-SE cross sections of initial and relocated hypocentres are shown in Figure 2 (A and B). These plots provide a refined view of events associated with the August 5-6, 1995 intrusion, as well as the Georges Hill swarm which occurred on August 12-15, 1995. The 3D locations reduce scatter of the August 5-6 hypocentres, suggesting the intrusion covered less horizontal distance and is perhaps more steeply inclined than original locations suggested. The 3D locations also suggest the Georges Hill swarm was more tightly constrained in depth, which is consistent with the interpretation of a stress response for this activity.
Figure 1A and 1B
Figure 2A and 2B
Figure 3 shows P-wave velocities in a NW-SE profile across Montserrat. These results suggest that the area seismically active during the August 5-6 swarm is faster than surrounding areas. This is consistent with what we have observed in other volcanic areas where intrusions are present.
