Abstract
Exposure modelling is a critical factor in the assessment of risk from natural hazards. Athough often its role has been overshadowed by other risk components (most notably, hazard), an efficient estimation of exposure is key to improve confidence in impact analysis and forecasting and ultimately support decision makers to improve risk mitigation activities. This is particularly relevant in urban and metropolitan areas, where the density and complexity of the interplay between population, socio-economical assets and infrastructure is likely to foster non-linear risk amplification, possibly due to cascading phenomena.
In the last decade several innovative methodological approaches have been proposed, building upon statistical modelling, exploiting heterogeneous data from remote sensing, and integrating machine learning techniques in order to improve understanding, formal description and representation of exposure in a wide range of applications. These activities have been originally developed within the community of seismic risk, and later increasingly extended to other natural hazards, acknowledging the need for a more general and flexible approach to exposure modelling in the context of multi-hazard and multi-risk applications.
This is ever more important considering also the ongoing convergence of Disaster Risk Reduction (DRR) and Climate Change Adaptation (CCA) in the broader context of Comprehensive Risk Management (CRM).
In this contribute we aim at providing an overview of the most recent advances that the authors have proposed in the field, outline the current challenge and perspectives in the field of exposure modelling, and draw a tentative roadmap for the next future.