Workshop “Multirisk Approaches: Quantitative probabilistic methods, comparable assessment procedures, domino effects”, Bruxelles (EC CDMA Building), May 31 and June 1, 2006

The state of the art and the problems related to the implementation of probabilistic approaches in a multi-risk framework have been discussed during a Workshop held at the EC CDMA Building (Bruxelles) on May 31 and June 1, 2006.
The Workshop dealt with “Multirisk Approaches: Quantitative probabilistic methods, comparable assessment procedures,domino effects”.
Its main objectives have been:

1. to assess the state of the art of multirisk approaches for the evaluation of natural hazards and risks, with special reference to domino (or cascade effects) triggering other natural and industrial risks;
2. to review the contribution by relevant EC projects;
3. to identify comparable procedures to the assessment of global risk useful for land use planning, town planning and emergency management;
4. to work out a document indicating future research direction and key arguments.

The event was organized in two sections. In the first section the activities developed in the frame of selected EU projects were presented, whereas in the second section experts were invited to present frontier lectures on relevant aspects of quantitative probabilistic risk evaluation and multi-risk approaches. It was solicited the participation of those projects funded in 5th and 6th FP where a multi-risk and/or quantitative approaches of hazard and risk was a relevant part of the project.
Researchers involved in 13 EU projects attended the meeting.
The results of the meeting were summarized tin the following document.
The last decades were characterized by a widespread increase of fatalities and economic losses due to natural catastrophic events. This escalation is not only due to the growth of inhabitant density in risky areas, but also to the consequent (and concomitant) increase of possible “cascade effects” due, for example, to the growth of man made structures which can produce explosions, pollutions, fires when hit by a natural catastrophic event.
In order to mitigate the effects of catastrophic events innovative approaches to the assessment of hazard and risk in a particular area are required. Although for industrial and man-produced risks, quantitative probabilistic estimates are routinely applied, the same is not generally true for many natural risks, with the exception of modern estimates of seismic risk. In fact presently available scenarios are frequently related to a reference event and rarely account for the related uncertainties. Moreover different types of risks (volcanic, fast mass movements, floods, earthquakes, etc) are often estimated using different procedures so that the produced results are not comparable.
This raises a challenging problem that can be tackled developing methods able to provide reliable quantitative estimation of individual and coupled events. Compared to classical analysis of single risks, these methods may provide a formal scheme to compare and to rank different kinds of hazardous phenomena (natural, industrial, etc...), and account for “cascade effects” that are usually neglected in single risks analysis.
This is of particular interest to assess and mitigate risks within mega-cities and in generally densely inhabited areas.
It was stressed that multirisk is not alternative to single risk analysis. Multirisk is something more. Reliable single risk analysis is necessary (but not sufficient) to have a reliable multirisk analysis (modelling some relevant “cascade effects”).
Perhaps the most important feeling from the workshop is that there is a strong research effort that has produced valuable results in predicting reference scenarios for individual typologies of risk. A significant improvement would be the use for all natural risks of probabilistic approaches, such as those used for earthquake and anthropogenic risks, allowing the development of several scenarios each characterized from a probability and uncertainty levels. True “multirisk” analysis (in terms of added value to the individual risks) is still scarce. FP7 needs a concentrated effort (perhaps a joint project) in this field, in particular when integrating phenomena with causal physical relations.
It must be emphasized that, as a rule, occurrence of hazardous phenomena is governed by a long-term evolution of a physical system and (sometimes) also by various triggering mechanisms. Special attention should be paid to understanding statistics and physics of the triggering effects and including them into the hazard analysis. Triggering is of key importance mainly in cascaded hazardous phenomena. New mathematical tools have to be developed to modify the hazard assessment methods for including triggering.
Attention should be paid to studies of the effect of early-warning systems on the risk at a site, and how to include such an effect in the mathematical assessment of the risk.
AMRA has gathered a group of experts on natural and anthropogenic risks who has worked out a Guide-book for the implementation of quantitative estimates of risks in a multi-risk approach. The application of this principles to a test case in Campania, Southern Italy, has been reported and discussed at the Workshop on Multi-risk approaches held in Bruxelles.