Risk assessment case history: the Soufrière Hills Volcano, Montserrat

Abstract

Introduction Volcanic hazard and risk at Soufriere Hills Volcano, Montserrat (SHV) has been assessed in a consistent and quantitative way for over 17 years (1997-2014), during highly variable eruptive activity involving andesitic lava dome growth (Wadge & Aspinall, 2014). This activity has placed serious stresses and constraints on the Montserrat population: about 12,000 people lived on this small Caribbean island prior to the start of the eruption in July 1995 and now (2014) this has stabilised at just over 4,000 souls. Over the years following 1995, a series of five very active dome growth episodes produced many pyroclastic flows, explosions and lahars, whose net effect was to destroy the main town, Plymouth, and most infrastructure, forcing people to leave Montserrat or live only in the northern part of the island. In June 1997, nineteen people were killed when a dome collapse pyroclastic flow caught a number of persons inside the exclusion zone. The risks faced by the people of Montserrat from volcanic activity are the responsibility of the UK government, and hazard and risk assessment work on Montserrat has been carried out by a Scientific Advisory Committee on Montserrat Volcanic Activity (SAC) (and the predecessor Risk Assessment Panel) appointed by them, working in collaboration with the Montserrat Volcano Observatory (MVO). While the administrative basis of the SAC has changed, the quantitative risk assessment methodology for enumerating risk levels (Aspinall et al., 2002, Aspinall & Sparks, 2002), has been kept the same since 1997 to ensure comparability of findings from one assessment to the next. In a protracted eruption crisis, continuity in scientific inputs to decision-making is essential: any major change in concepts, modelling or assumptions could entail large differences in evaluated risk levels and hence engender doubts for officials and confusion in the minds of the public. This series of multiple, repeated quantitative volcanic hazard and risk assessments must be unique in volcanology. In the case of Montserrat, by ‘volcanic risk’ we mean the probability that a person will be harmed by some volcanic hazard within some specified timeframe; assessing other risks and losses, such as damage to buildings or infrastructure, have had only a limited consideration in terms of framing scientific advice.