1. Mountain Risks: 2007-2010 A Marie Curie Research & Training Network THE MOUNTAIN-RISKS RESEARCH PROJECT: CHALLENGES IN QUANTITATIVE RISK ASSESSMENT P. Giacomelli (1), S. Sterlacchini (2) and the ‘Mountain-Risks’ research team (1) Department of Economy and Agricultural Politics, University of Milano-Bicocca, Milano, Italy. (2) Department of Environmental and Territorial Sciences, University of Milano-Bicocca, Milano, Italy.  SCOPE: MULTI-DISCIPLINARY ANALYSIS OF LANDSLIDE RISK Usually, risk assessment comprises three essential components: risk analysis, risk evaluation and risk management. Among these components, vulnerability analysis is an essential part of risk assessments. Ideally, Quantitative Risk Assessments (QRA) requires that both hazard and vulnerability of the elements at risk be defined as independent probabilities (of occurrence, and of damage, respectively). In practice, it is rarely possible to obtain such probabilities and apply the generally accepted risk definition. Therefore, for many years, a qualitative or semi-qualitative evaluation of consequences, based on expert judgement and interviews with private households, firms and land owners, has been preferred as the more convenient method in terms of application. Subsequently, statistics of detailed historical records, matrices or mathematical frameworks have become more commonly used. A recent literature review has demonstrated that within quantitative risk research there is a serious lack of studies related to vulnerability, especially concerning both the social and economic aspects. Contact: Paolo Giacomelli, Simone Sterlacchini Email: paolo.giacomelli@unimi.it Email: simone.sterlacchini@unimib.it An entire risk assessment requires the combination of different techniques and methodologies, and the interplays of various experts. Thus, a more rigorous quantitative ‘cause-effect’ correlation has to be investigated in order to develop a systematic approach to multi-risk problems. This should be incorporated into a QRA, which robustness depends upon: Fig. 1: The concept of risk assessment (from Bell & Glade, 2004). Risk analysisRisk evaluation Risk management What could happen? What could happen if something changes? What could happen, and where (and if something changes)? What could happen, and when (and if something changes)? What could happen, when and where? (and if something changes?) What is allowed to happen? What must not happen? Who is affected? Who has to decide? What is to be done? What can be done? What are the alternatives? Who is paying? -a quantitative evaluation of multihazards consequences through the collection of relevant data on past phenomena; -a translation of the damage assessment into economic direct and indirect losses; -an analysis of the trends acting in the study area in order to translate the physical effects into economic and social losses; -an understanding of the importance that society attaches to the level of hazards. Therefore the ‘Vulnerability analysis and QRA’ working block of the project will investigate the following themes.  Estimation of risk scenarios  Identification of the source of uncertainties in the analyses Fig. 3: Example of procedure used to assess landslide risk on the Municipality of Corvara in Badia (from Sterlacchini et al., submitted). Fig. 2: Matricial risk computation scheme (from Amatruda et al., 2004) Fig. 5: Example of matrix of individual risk values expressed as a function of different specific vulnerabilities and scenarios (P: physical, S: social, E: economic, En: environmental) (from Amatruda et al., 2004). Fig. 7: Example of a calculation of direct and indirect damages for different scenarios for the Municipality of Corvara in Badia (Plan Pezzie landslide). Values were estimated by economic analysis (E) or by interpolation (I) (from Sterlacchini et al., submitted). Many authors have stressed the gap between the theoretical definition of risk and the practical assessment of risk. Therefore expert’s knowledge and professional judgements are deeply involved in a risk assessment procedure, given that each mathematical model, used to represent complex natural mechanisms, has to be integrated with guidelines able to synthesise human behaviours, choices and actions. As a consequence, uncertainties associated to both the raw data used in the analysis and the somewhat qualitative nature of vulnerability will be investigated and quantified.  Inventory of the elements at risk and estimation of vulnerability functions The inventory of the elements at risk and the associated vulnerability functions has to be performed using territorial data (landuse, population distribution, etc.). The elements at risk are then classified and counted. These elements at risk can be monetarily quantified or globally assessed in order to obtain a consequence calculation and/or a vulnerability function. Within the project, the value computation will be performed using several methods (computation of a specific value for the individual elements, use of utility functions, semi- quantitative assessments of a global value for a certain area, eg. index-based approach). Fig. 4: Example of a GIS database of the elements at risk of the Municipality of Barcelonnette (South French Alps) and related socio-economic attributes. QRA will be performed exploiting the idea of scenario mapping: the number of people killed and injured, and the losses arising in other elements (buildings, infrastructures), due to the impact of a damaging event will be assessed within ‘realistic’ hazard and vulnerability scenarios. All of the scenarios will have to consider the simultaneous occurrence of rare events. The economic consequences (in terms of direct and indirect losses) will be evaluated through scenarios taking into account the social and economic trends acting in the study areas. On one hand, the social analysis will characterize the human presence and its evolution over time; on the other hand, the economic analysis will evaluate the assets allocated to an exposed area, study the flow of activities and finally define the flows of goods and commodities. Fig. 6: Example of risk maps obtained by GIS techniques concerning the economic risk for a scenario in the Anzasca valley (Pennine Alps, Italy) (from Amatruda et al., 2004) The results of the QRA may be considered as the key to estimate the level of future prospective losses, becoming the leading information for risk evaluators and risk managers. C1C2C3C4 H1R0 R1R2 H2R0R1R2R3 H3R1R2R3 H4R1R2R3 Consequences Hazard 4 risk classes: R1: Null; R2: Low; R3: Moderate; R4: High Fig. 8: Example of a risk map obtained for a hazard and a consequence scenario over the Municipality of Enchastrayes (South French Alps), by combining probabilistic analyses of landslide susceptibility, process-based modelling to assess landslide occurrence, and an evaluation of potential consequences with an index-based approach (Malet et al., 2006).