Team Deform: P. Jackson (SRU), J.B. Shepherd (University of Lancaster), R. Robertson (SRU), G. Skerritt (Dept. Lands & Surveys), S. Sol (SRU), R. Herd (BGS), G.E. Norton (BGS), P.D. Cole University of Luton), J. Devine (Brown University), D. Silcott (MVO), S.R. Young (BGS)
Montserrat Volcano Observatory, Old Towne, Montserrat
Ground deformation surveys using a Leica TC1100 Total Station (combined theodolite and Electronic Distance Meter) were begun in September 1995, shortly after the beginning of phreatic activity at the Soufriere Hills Volcano. A network of single radial lines around three quarters of the volcano were repeatedly measured using either single or triple reflectors mounted either on tripods or permanently fixed into bedrock.
Rationalisation of the network was performed in December 1995 and January 1996 in the form of the addition of several new lines to the pre-existing target sites. Each target site was made the apex of a triangle. All three sides of the triangle were measured in a single occupation providing an internal check on the quality of the data and allowing a certain amount of redundancy. Furthermore these data permit the location of the target site using a simple geometrical construction. Four such triangles were established covering three-quarters of the volcano. The fourth quarter is difficult to access and is an unfortunate gap in the network. Atmospheric correction using the pressure and temperature taken at the instrument site was also included.
The data recorded by the Total Station are in the form of slope distance, horizontal distance, height difference between target and instrument site and vertical angle. The slope data are reduced to point-to-point distance on return to the Observatory. In order to perform a rapid check whilst recording the data the instrumental set-up (instrument and target tripod heights) is kept constant for each line allowing a first order estimation of any change.
The data collected over the past year shows that there is very little or no deformation taking place at a volcano-wide scale. Target sites on the northern and western flanks of the volcano have not shown any consistency in their changes. The absence of a trend in a data set is assumed to be variation within the precision limits of the instrument and technique. The only consistent trends of ground deformation have been observed on the eastern and southern triangles measuring to reflectors at Castle Peak and Chances Peak, respectively. These targets are the closest of all the targets to the active area within English's Crater. By far the most active of the targets in the whole network has been the Castle Peak reflector. Lines to this target have recorded in excess of 1 metre of outward movement of the target since October 1995.
The concentration of measurable ground deformation to areas very close to the volcanic activity suggests a local source. Evidence of inflation or deflation of the volcano by intrusion at depth has not been observed. Three sources of the deformation are considered likely to have influence on the deformation: thermal expansion, loading by the growing lava dome complex and pressurisation of the conduit. It is considered likely that the observed deformation is a combination of these and possibly other factors.
The movement history of the Castle Peak reflector can be divided into three distinct periods on the basis of the rate of deformation. Rapid deformation with shortening rates in excess of 10 mm per day characterises the period from mid-October 1995 to mid-November 1995. This corresponds to the period up to the appearance of new material in the crater. This was replaced by a much reduced rate of approximately 1 mm per day shortening which lasted until mid-July 1996. During this period the dome gradually grew in size and produced pyroclastic flows which reached the sea at the mouth of the Tar River valley. From mid-July the rate increased to approximately 6 mm per day shortening, similar to the rate seen at the beginning of the campaign. This increase is very well correlated with the re-appearance of volcano-tectonic earthquakes at that time.
The step-like motion of the Castle Peak reflector is believed to be the result of three processes: mechanical pushing of the Castle Peak rocks by intrusion of magma, thermal expansion due to heating of the country rocks by the sub-surface magma, circulating super-heated fluids and the growing lava dome and, thirdly, pressurisation of the conduit. These process are most likely to have acted together for the majority of the crisis but the distinct periods with specific rate of deformation are believed to be related to one process exceeding the other two in the contribution to the deformation of Castle Peak.