Supplementary material to “Probabilistic Volcanic Hazard and Risk Assessment”

W. Marzocchi, Istituto Nazionale di Geofisica e Vulcanologia, Bologna, Italy; A. Neri, Istituto Nazionale di Geofisica e Vulcanologia, Pisa, Italy; C. G. Newhall (Ret.), U.S. Geological Survey, University of Washington, Seattle; P. Papale, Istituto Nazionale di Geofisica e Vulcanologia, Pisa, Italy

Citation:
Marzocchi, W., A. Neri, C. G. Newhall, and P. Papale (2007), Probabilistic volcanic hazard and risk assessment, Eos Tran. AGU, 88(32), 318. [Full Article (pdf)]

Volcanic hazards can affect millions of people, and their reliable evaluation is a primary ingredient for a sound emergency management, land use planning, and other risk mitigation strategies. Regrettably, we are still far from having complete volcanic hazard and risk assessment for most dangerous volcanoes. This dearth is due partly to the intrinsic complexity and variability of volcanic processes and partly to the fact that a full volcanic risk analysis requires multidisciplinary competences and expertise, ranging from classical volcanology, physics, engineering, statistics, medicine, communications, educations, etc. A further important issue is the large variability in the level of risk assessment worldwide, with few well monitored volcanoes and many dangerous volcanoes worldwide almost lacking any instrumentation.

These aspects of the problem were discussed in a recent meeting held in November 2006 at Erice (Italy) entitled “Quantifying long- and short-term volcanic hazard: building up a common strategy for Italian volcanoes” (http://www.bo.ingv.it/erice2006). The main goal of the workshop was to discuss the role of quantitative volcanic hazard assessment in managing volcanic crises and in assisting the land-use planning of volcanic regions. Issues that emerged in invited talks are in two main (related) groups:

  1. defining the role of uncertainty and probability in hazard/risk assessment;
  2. how volcano hazard studies are communicated to society and decision makers.

In the following we report the main outcomes that emerged from the workshop.

The term hazard can lead to some misunderstanding. In English, hazard has the generic meaning “potential source of danger”, but for more than thirty years (e.g., Fournier d’Albe, 1979), hazard has been also used in a more quantitative way, that reads: “the probability of a certain hazardous event in a specific time-space window”. This discrepancy has been also addressed in the recent book Statistics in Volcanology (2006), where volcanic hazard is adopted as the label for the latter, more quantitative definition. However, many workshop participants still use “hazard” and “volcanic hazard” in purely descriptive and subjective ways. For this reason, it has been suggested that a more suitable term for the estimation of quantitative hazard is “Probabilistic Volcanic Hazard Assessment” (PVHA, hereinafter).

The extreme complexity, nonlinearities, limited knowledge, and the large number of degrees of freedom of a volcanic system prevent a deterministic prediction of the evolution of the processes involved. In other words, volcanic systems are stochastic and hazardous volcanic phenomena involve so many uncertainties that only a probabilistic approach can cope with them properly. In general, probabilistic and stochastic modeling have a two-fold application: to set up evidence-based models, and to build a framework that merges all kinds of available information (theoretical, empirical, geological, volcanological, geophysical, historical, etc…) in order to achieve the most reliable PVHA which, in turn, becomes the rational basis for critical decision-making, in terms of safety and mitigation. A major outcome of the Erice workshop was essentially full advocacy of the use of the PVHA approach as a tool in mitigating volcanic risk to communities. In this context, PVHA methodologies are useful in both in the long-term, prior to onset of volcanic unrest and in the short-term, during volcanic activity and during “volcano crises”.

The probabilistic approach does not in any way reduce the importance of deterministic studies and the analysis of specific scenarios. Instead, as mentioned above, the probabilistic methods discussed during the workshop seek to merge together all the available theoretical knowledge, physics/stochastic modeling, and empirical evidence in order to provide the best possible picture of the phenomenon. This procedure contrasts with what is sometimes encountered in seismic risk analysis, where deterministic and probabilistic approaches are often considered irreconcilable (e.g., Castanos and Lomnitz, 2002). In seismic hazard assessment, the terms “probabilistic” and “deterministic”, contained in acronyms PSHA and DSHA, reflect the kind of strategy adopted, mostly evidence-based for PSHA and mostly based on physical models for DSHA. In volcanology, we do not see this conflict; we attempt to use all the information we have (models, data, and expert beliefs), and the term “probabilistic” in PVHA only emphasizes that the quantification of volcanic hazard takes account of associated uncertainties.

Classical volcanological studies also remain a primary ingredient for PVHA, providing a variety of information on volcanic scenarios which include: style and recurrence of eruptions, eruptive parameters, time sequences, frequencies of eruptions/eruptive phenomena, durations, intensities, magnitudes, etc. All of this information, joined with quantitative models that provide insights into causes and effects as well as information on dynamic quantities not preserved in the deposits, supply the basic scientific ingredients to be merged in a probabilistic scheme that works like a “glue” to combine all these aspects. Basically, models and empirical observations can be used to describe different possible eruptive scenarios for a specific volcano, and then PVHA estimates the relative probabilities of encountering each scenario in a future eruption.

In the eruption forecasting context, we considered the controversy linked to the term precursor. There was not an unanimous consensus about the meaning of this word among the participants at the workshop; the term precursor is seen either in a deterministic sense (a phenomenon that appears necessarily before an eruption, i.e., one-to-one correlation between precursor and eruption), or in a more statistical sense (a phenomenon that “mostly” appears before an eruption), and some also pointed out that a precursor can be defined only retrospectively, i.e., looking at the pre-eruptive processes only after the eruption. These are not just semantic differences, because they lead to very important practical implications. Civil authorities and society usually interpret the term precursor in a deterministic way, while for many researchers it makes more sense to think of precursors statistically. This discrepancy inevitably introduces misunderstanding that may lead to wrong practical actions to mitigate the risk (e.g., a premature call for an evacuation). In practice, some participants avoid the term precursor and instead estimate the probability of occurrence (and possibly of the size) of an eruption.

The average level of knowledge of pre-eruptive phases is significantly lower than for syn- and post-eruptive phases. The most obvious reason is that all pre-eruptive processes occur deep inside the volcano, inaccessible to direct observation and, historically, early signs of impending eruption may have been marginal and not documented. Our eruption forecasting ability overall, and especially for long-quiescent explosive volcanoes, is still rather poor. Initiatives like the WOVOdat project (http://www.wovo.org/WOVOdat/) will improve our capabilities. WOVOdat is a fledgling database that will serve as the primary resource for a new field of “volcano epidemiology” and will also aid associated research into how volcanoes prepare to erupt. During volcanic crises, it can be used to make queries such as “where else have X,Y,Z been observed and what happened?” or, more quantitatively, “how much do the given observations increase the probability of eruption today, tomorrow, and in the future?”

On the second group of issues discussed by the workshop, i.e., the interaction among volcanologists and society/public officials, all agreed that the achievement of a full and comprehensive risk assessment requires other competencies beyond volcanology. In particular, together with a sound PVHA, it is necessary to evaluate the vulnerability of exposed infrastructure, facilities and property, the impact of eruptions on human beings, costs vs. benefits of proposed mitigation measures, and the level of “acceptable risk” for society. In addition, we need educational programs to improve the “risk perception” of the people living around volcanoes, and improved ways to communicate risk and associated uncertainties to those people, mass media, and local authorities. All agreed that a multi-expert community (preferably established before periods of crisis) is needed to deal with all of these issues adequately, and that a common vocabulary should be created to simplify communications between experts, minimizing possible ambiguities and misunderstandings. The team, which worked on the recent European project EXPLORIS, represents an excellent example of such a collaborative grouping of experts (even then some specific competencies were lacking).

Another remarkable exercise, named MESIMEX (Major Evacuation Simulation Exercise), aimed at the evaluation of problems associated with a crisis at a high-risk volcano, was carried out in the Vesuvian Area in October 2006. The findings of that exercise were presented and discussed by the Department of Civil Protection of Italy, which coordinated the operations, as well as by researchers of the Istituto Nazionale di Geofisica e Vulcanologia (INGV), who had been involved. The exercise consisted of simulating a possible pre-eruptive scenario at Mt. Vesuvius, and the consequent evacuation. This took place according to the methods and procedures established by the National Plan of Emergency, involving about a hundred people from each of the 18 Municipalities concerned. The primary goal of MESIMEX was to check the viability of the whole procedure for managing the emergency, from monitoring data collection through to the decisions and testing of practical actions to manage and mitigate the impending volcanic risk. The exercise was highly successful, because it identified the strong and weak points of the chain of operations. Significantly, it was clear that periodic repetitions of similar exercises (also for other critical volcanic areas) would help to create a practiced multi-disciplinary community, and to determine the best procedures for risk mitigation. Establishing clear roles, and boundaries, for experts with different competences is important as well. For instance, the volcanologists themselves cannot order “evacuation” because they do not have full knowledge about costs and the risks of moving many people during emergencies, and because volcanologists deal mainly with scientific issues and not with social and political decisions. In other words, only a multi-disciplinary effort can give reliable answers to the demands of a society threatened by a volcano. The workshop outcomes are a significant step in this direction.

REFERENCES