Science Service System
You are here : Home : Proposals_Summary

Summary of Proposal GEO0436

TitleRadar detection of dome growth and mass wasting at Soufriere Hills Volcano
Investigator Wadge, Geoff - University of Reading, Environmental Systems Science Centre
Team Member
Mr Stewart, Roderick - University of the West Indies, Montserrat Volcano Observatory
SummaryWe propose to acquire TerraSAR-X high resolution spotlight mode images of the summit of Soufriere Hills Volcano, Montserrat. This volcano, on a UK Dependent Territory, has been erupting since 1995 and has made the southern half of the island uninhabitable. There is a modern, well-equipped observatory, but the scientists there can go for many weeks with no good visual observations of the main hazard posed by the volcano - the growth of a large volcanic dome (and its frequent collapses). We know from ERS,Envisat ,Radarsat and other data that changes to the dome morphology can be detected in amplitude images. The high resolution data provided by TerraSAR-X will be particularly valuable in this regard. Soufriere Hills is showing signs (July 2008) that it may be about to develop a new lava dome and we wish to be able to monitor this re-start and any new growth with X-band amplitude images. Such images will have an immediate, real-time benefit to the hazard assessment at the observatory. Also we will use surface change detection images and InSAR on the data series to assess mass wasting deposits and near-dome surface deformation. This effort will not be in real-time. Our project will run for 2 years during which we wish to acquire about 40 scenes. When the dome is growing we will require 11-day repeat images, during periods of no-growth, images every 44 days. We will report the findings of the work in international journals e.g. Journal of Volcanology and Geothermal Research and radar conferences. This work is funded through the National Centre for Earth Observation, Theme 6 programme.
Final ReportThe eruption of Soufriere Hills Volcano on Montserrat has been ongoing since July 1995, with variable periods of lava dome growth and quiescence. MVO had made use of radar data throughout the eruption, to try to understand the nature of dome growth, particularly when the volcano was covered by cloud, which was often. But prior to TSX, radar data was either too coarse a spatial resolution or too difficult to access to make the most of the technique. The project was triggered by the explosion of 28 July 2008, which was the first activity for 15 months. G.Wadge (Reading) and R.Stewart (MVO) responded to the crisis resulting from this (an evacuation was called) by invoking the Charter for Space and Major Disasters, via the UK government, to task satellites to help provide data. In addition, DLR (via M. Eineder) were also asked to provide TSX data. It was these data that were the most useful and enabled MVO to assess the nature of the explosion’s vent (hidden by cloud) and advise the civil authorities accordingly within 4 days of the explosion and well before the cloud lifted. This effort is described in full in the following report: Ager, G., Wadge, G., Tragheim, D. and Jordan, C.(2008). Volcanic eruption of 29th July 2008 on the island of Montserrat. Final operation report, Charter for Space and Major Disasters, Charter ID:213, British Geological Survey report CR/08/179, 48pp. MVO needs to know the shape of the lava dome and the volcano generally, where it is currently growing and where the pyroclastic flows, derived from collapses of the dome and explosions, have gone, in order to evaluate the hazard and risk from the volcano. In addition to its all-weather/day-night capability , TSX is of sufficiently high resolution (~ 2 m) and frequency (11 days) to capture the deposition of individual flows. However, during high levels of activity many such flows may be deposited in the same area each day, much too frequently to be resolved by TSX. Nevertheless, the project was intended to test the ability of TSX images to capture these morphological and depositional changes at the volcano. The volcanic change in four areas of interest were studied: i. Location of volcanic vents ii. Areas of new pyroclastic flow deposits iii. Infill of valleys by deposits iv. Topography. Three radar techniques were used to extract the volcanological information: Backscatter change differencing (i and ii) Radar shadowing (iii) InSAR (iv). These techniques showed real promise in enabling MVO scientists to extract useful operational information. They are evaluated in full in a paper currently under review: Wadge, G., Cole, P., Stinton, A., Komorowski, J-C., Stewart, R., Toombs, A., Legendre, Y. Rapid topographic change measured by high-resolution satellite radar at Soufriere Hills Volcano, Montserrat, 2008-2010. Journal of Volcanology and Geothermal Research. The conclusions are as follows: 1. TSX spotlight mode radar data have been shown to be of value in assessing the extent and thickness of newly emplaced pyroclastic flow deposits during periods of intense volcanic activity involving pyroclastic flows, vulcanian explosions and dome collapse at SHV. 2. The metric-scale spatial resolution of the radar allows decametric-scale features of the flow deposits to be mapped, provided there is sufficient radiometric contrast with the substrate. This is usually the case immediately following periods of quiescence when earlier, perhaps similar, deposits have been eroded, but much less so during the middle of a period of frequent and overlapping flow events. As in the case of the 28 July 2008 explosion vent, it is possible to sometimes define discrete features on the lava dome itself. 3. The temporal resolution of TSX (11-day repeat) is useful for extended periods of activity, but generally cannot separate individual flow events. However, we have shown that acquiring both east-and west-looking images of the volcano on the same day, 12-hours apart, allows the emplacement of individual pyroclastic flow deposits to be identified in some cases . A more frequent imaging still could be achieved with TSX using multiple look angles from more orbits. 4. A typical pattern of deposition at SHV is of a block-and-ash flow in the bottom of a valley with marginal surge deposits on the interfluves, and is expressed in TSX change difference images (before and after) as a central strand of higher amplitude signal from the rougher block-and-ash flow surface to flanking areas of lower signal due to the smoother ash cover. Subsequently, this pattern tends to reverse as fluvial erosion roughens the surge deposits and smooths the valley bottoms with water-borne sedimentation. The increase in the dielectric constant of the ash after rainfall may confuse this picture. Wet conditions before the flow event and dry after will tend to diminish the block-and-ash flow signal and enhance the surge ash signal in change difference images, and vice versa. We have not been able to test this in the field. 5. The infilling of radial valleys with many tens of metres of pyroclastic flow deposits and rockfall can be measured using radar shadowing. A combination of this technique and 11-day change difference imaging shows how the directional locus of deposition around the active lava dome changes over periods of weeks. The topography of the volcano’s surface denuded of vegetation by the eruptive activity can be measured by InSAR using 11-day interval TSX spotlight mode data from a period of little surface actvity. These data are subject to errors induced by atmospheric refraction and unwrapping, but locally they can capture the topography to an accuracy of a few metres and can be used for analysis of the potential flow paths of pyroclastic flows. These findings were recently presented at the following conference: Wadge, G., Cole, P., Stinton, A., Komorowski, J-C., Stewart, R. (2010) High resolution radar monitoring of Soufriere Hills Volcano, Montserrat with TerraSAR-X. Cities on Volcanoes 6 Conference, Tenerife, 31 May – 4 June 2010, abstract 1.4-O-02.

Back to list of proposals

© DLR 2004-2016