Locating Pyroclastic Flows on Soufriere Hills Volcano, Montserrat, West Indies, Using Amplitude Signals from High Dynamic Range Instruments

Document Type

Article

Publication Date

11-2002

Keywords

pyroclastic flow, amplitude-based location, attenuation-corrected reduced displacement, Montserrat

Digital Object Identifier (DOI)

https://doi.org/10.1016/S0377-0273(02)00299-8

Abstract

Pyroclastic flows are located using amplitude signals from a seven-station high dynamic range seismograph array located 1.9–6.1 km from Soufriere Hills Volcano in Montserrat, West Indies. Locations are determined by measuring the seismograph signal amplitude for an event recorded at several stations in a moving time window analysis. For a given window, the measured amplitudes are corrected to a trial source location by removing the effect of the surface wave geometric spreading, instrument gain, and the attenuation at calculated travel-times. The trial source location is then compared to other trial locations via an iterative localised grid search where the root-mean-squared amplitude residual (δA) is minimised. The process is repeated for subsequent time steps resulting in a best-fit event location and size through time. The method has been tested on four small events occurring on April 8, 1999, August 12, 1999, February 25, 2001, and July 4, 2001, when visual observations of pyroclastic flows coincided with good seismograph station coverage (number of stations ≧5, azimuthal gap <160°). Based on the location results the four events propagated ∼0.5, 1.4, 1.3 and 1.0 km from the dome, and had maximum attenuation-corrected reduced displacements (DRQ) of 9.0, 2.8, 6.9 and 2.3 cm2 and maximum pyroclastic flow velocities of 7, 30, 20 and 8 ms−1, respectively. A time-lapse video of the event of August 12, 1999, shows that amplitude-based location through time closely matches the observed run-out distance and velocity. In contrast, amplitude-based locations for the events of April 8, 1999, and July 4, 2001, underestimated the actual flow run-out by ∼1.5 km. Underestimation of the true run-out distance is probably due to both the increased distribution of sources as coherent dome material disaggregates into many blocks, and signal contamination from other sources. Results indicate that pyroclastic flows and rockfalls can be located using amplitude signals from high dynamic range seismograph stations yielding estimates of size and trajectory, regardless of visibility conditions on the volcano. This new method is being tested as a hazard mitigation and research tool on Montserrat.

Was this content written or created while at USF?

No

Citation / Publisher Attribution

Journal of Volcanology and Geothermal Research, v. 118, issues 3-4, p. 299-317

Share

COinS