Integrating expert opinion with modelling for quantitative multi-hazard risk assessment in the Eastern Italian Alps

Extreme rainfall events are the main triggering causes for hydro-meteorological hazards in mountainous areas, where development is often constrained by the limited space suitable for construction. In these areas, hazard and risk assessments are fundamental for risk mitigation, especially for preventive planning, risk communication and emergency preparedness. Multi-hazard risk assessment in mountainous areas at local and regional scales remain a major challenge because of lack of data related to past events and causal factors, and the interactions between different types of hazards. The lack of data leads to a high level of uncertainty in the application of quantitative methods for hazard and risk assessment. Therefore, a systematic approach is required to combine these quantitative methods with expert-based assumptions and decisions. In this study, a quantitative multi hazard risk assessment was carried out in the Fella River valley, prone to debris flows and flood in the northeastern Italian Alps. The main steps include data collection and development of inventory maps, definition of hazard scenarios, hazard assessment in terms of temporal and spatial probability calculation and intensity modelling, elements-at-risk mapping, estimation of asset values and the number of people, physical vulnerability assessment, the generation of risk curves and annual risk calculation. To compare the risk for each type of hazard, risk curves were generated for debris flows, river floods and flash floods. Uncertainties were expressed as minimum, average and maximum values of temporal and spatial probability, replacement costs of assets, population numbers, and physical vulnerability. These result in minimum, average and maximum risk curves. To validate this approach, a back analysis was conducted using the extreme hydro-meteorological event that occurred in August 2003 in the Fella River valley. The results show a good performance when compared to the historical damage reports. (C) 2016 Elsevier B.V. All rights reserved.