Lava domes are piles of viscous magma that, in essence, form bulbous plugs on top of volcanic vents. They grow slowly as partially solidified magma is squeezed up the volcanic conduit, and they can host high internal pressures as gases exsolving from the magma try to escape. Lava dome eruptions are notorious for they often suddenly transformation from being benignly effusive to violently explosive. Sudden removal of these plugs, either by vertically-driven explosions, caused once the internal gas pressure exceeds the tensile strength of rock, or sudden collapse and spontaneous disintegration as the piles grow and become unstable, can have devastating consequences. These collapse events spawn one of the major hazard in volcanology; devastating pyroclastic density currents, which move down the flanks of the volcano at speeds of up to 60 m/s. This project concerns the improved understanding and forecasting of lava dome collapses and has immediate and practical applications pertaining to the management of several on-going volcanic crisis. The work will be based on the Soufrière Hills volcano (SHV) eruption, Montserrat, combined with preliminary comparative studies from two other active lava dome eruptions, Mount St Helens, USA, and Santiaguito, Guatemala. Such eruptions occur with relative frequency, are potentially extremely destructive and can continue for years-to decades. The real merits therefore, will be reaped when relationships unearthed during this study are usefully and successfully applied to dome forming eruptions elsewhere.
The fundamental objective of this work is to stringently quantify aspects of activity associated with lava dome failure and collapse including information on the timing and location of rockfalls and pyroclastic flows, as well as the nature of other precursory activity. The project will address three issues that require further attention: differentiating failure modes; identifying structural controls; and treating collapses as part of a continuous process of mass wasting rather than discrete independent events. By investigating along these themes, this work will validate existing models and constrain a number of current limitations. The results of this work will provide the basis from which statistically-driven models can be developed, in order to construct a short-term forecasting model of mass wasting at lava domes. The novelty and impact of this study hinges on: Innovative techniques for comparing disparate data sets of monitoring data, and the generation of an important record of dome activity that will be useful to the volcanological community; Understanding continuous growth and wasting processes allowing assessments to be made about physical mechanisms but also providing a basis for testing statistical models (the quality of which is enhanced by collection of large data sets); The fusion of data from digital elevation models, thermal images, and gas plume maps, developing a truly integrative approach to understanding dome instability.
One of the most important types of volcanic hazards are pyroclastic density currents. These are hot avalanches of ash, volcanic gas and debris that sweep down the flanks of a volcano destroying everything in their path. They are frequently generated at lava dome volcanoes, where large volumes (tens of millions of cubic meters) of viscous lava pile up around the summit of a volcano and periodically collapse to form these flows. This research has focused on improving the understanding of three aspects of lava dome eruptions: 1) How the structure of the lava domes affect their propensity to collapse; 2) how frequently the collapses generating pyroclastic density currents occur, and what the forcing mechanisms are and; 3) understanding the mechanisms of collapse and flow propagation in order to forecast the area of their impact. The principal outcomes of the work are the following: We have identified particular growth styles and indicators of inherent instability associated with lava domes from several volcanoes that are formed by either slow or rapid growth respectively. We have used direct sampling, aqueous geochemistry, and remote sensing techniques to detect hydrothermal alteration of a lava dome in Guatemala to assess collapse hazards at that particular volcano. A database has been constructed of hazards associated with lava dome eruptions around the world, and this has been used to undertake a review and statistical analysis of those hazards as well as improve aspects of empirical models used for forecasting the inundation extents of pyroclastic density currents. A key outcome of this work has been determining factors that control the dynamics and emplacement of pyroclastic density currents. In particular, these currents sometimes separate into high and low-density components, which react to the topography over which they are flowing in very different ways. The travel path of the upper, lower density, part of these currents, is particularly difficult to forecast, and it is this part of the current that is responsible for high fatality numbers. Our work has identified different flow regimes that come into play, during propagation of these currents, which controls the extent to which this flow separation occurs. Once a fuller understanding of the mechanics of these processes can be incorporated into computational models, fundamental improvements will be achieved in the reliability of model results for hazard mapping. Lastly, a new methodology has been developed which provides fully probabilistic inundation maps of pyroclastic density currents using an existing flow model but with much-improved probabilistic capabilities. This method has been developed initially for the Soufriere Hills Volcano, Montserrat, but is applicable at other sites, as well as for other hazard types. The funding has supported the work of one investigator and three graduate students collaborating with a number of institutions around the world. The graduate students have benefited from key first-hand experience in the theme of applied volcanology working alongside and developing an understanding of the work done in volcano observatories and by government agencies responsible for real-world hazard management.
