CASE 2: DO INFLATION CYCLES OF THE ACTIVE SOUFRIERE HILLS VOLCANO, MONTSERRAT CORRELATE WITH DOME GROWTH?

 

SCIENTIFIC BACKGROUND:

In 1958, the British Ordnance Survey (BOS) as part of its mapping of the West Indian Crown Colonies of the United Kingdom, established a tight trigonometric network on Montserrat consisting of over 30 fiduical marks. Two field campaigns were conducted in 1966 and 1973. An estimate of the average RMSE is ±10.0 cm. In late August 1995, approximately one month after the start of a phreatic eruption on Montserrat, we redetermined the absolute positions and baselines between some BOS trig. stations azimuthally distributed around the Soufriere Hills volcano using single-frequency carrier-phase, differential GPS geodesy (Figure 1). The baseline repeatability is ±6 cm. Our results demonstrated that the positions of the trig stations had not changed within error. The maximum surface deformation that may have preceded the prhreatic eruption and associated crustal magma injection, therefore, was about 10 cm. Subseqent to L1-only occupations, dual-frequency, geodetic quality GPS measurements were made on the same sites and the network was densified. Measurements began in October 1995 and are on-going. (Additional CGPS sites will be installed in late 2002 along with borehole geophysical equipment as part of CALIPSO, a collaborative project among the University of Arkansas, the Pennsylvania State University, Duke University, the Montserrat Volcano Observatory, Bristol University (UK) and Leeds University (UK) funded by NSF-Continental Dynamics and NERC). In summer 1996, two continuous GPS sites were installed. During the course of the eruption, the CGPS sites were damaged or destroyed completely. Observed deformation showed subsidence at all sites, which was a strong function of radial distance from the active vent. The horizontal deformation field was not, however, axisymmetric (Figure 2). The observations were modeled to derive two subsurface pressure sources: a shallow subvertical dike and a deeper mid-crustal (~ 6km) magma storage area. Since 1998, a six station CGPS network established jointly by the British Geological Survey-Montserrat Volcano Observatory and us has operated around the Soufriere Hills. Continuous data show significant departures from simple steady-state motion and appear to correlate with eruptive events and dome growth (Figure 3).

 

 
OBJECTIVES:      
 
The objectives of this case study are to have the students quantify the limitations on the L1 data to estimate inflation/deflation relative to 1958, to identify trends of different temporal scales in the CGPS data and to determine which trends correspond to geophysical processes and which result from “noise”; and describe the effect of equipment type on precision and accuracy. In addition, they will assess the predictive capabilities of GPS.
 
SELECTED ACTIVITIES      
 
a) Compare L1 measurements with trigonometric network observations: Students should calculate that the errors on the L1 and trigonometric network measurements are of similar size (~10 cm) and that no deformation occurred within error (Figure 1).
b) Examine trends in CGPS data: Students will break data into trends depending upon temporal scale, e.g. days, weeks, months, years, and compare trends with eruptive events (e.g. and ). Data will be reprocessed in GIPSY using different and migrating time windows. Most daily variations arise from noise, whereas those of weeks to months correspond to events (Figure 3.
c) Download CGPS data from the Montserrat network, process, and add to time-series: Students will watch for developing trends and dramatic changes in data that may indicate imminent changes in volcano behavior. The purpose is to involve the students directly in the process of assessing the eruptive potential of the volcano.
d) Prepare a statement for the Governor of Montserrat concerning risk from the volcano for specified time periods: This forces students to trust their data analysis, to understand data limitations for prediction, and to realize the implications when applying scientific results to societal issues.