Novel methods for large-scale dynamic probabilistic risk modelling

PROJECT BRIEF

Time-dependent risk modeling for long-term planning: Integrating changing vulnerability, exposure and hazards in the anthropocene

David Lallemant, Maricar Rabonza

The risk profile of cities – particularly Asian cities – is rapidly changing due to climate change, urbanization, and new patterns of vulnerability. However, current probabilistic disaster risk assessment approaches have been limited to static analyses (e.i. based on the current condition of hazard, exposure and vulnerability). This is a significant limitation for assessing risk in rapidly changing environments, or for analyzing the impact of policies and investments on future risk. Thus, the development of time-dependent risk analysis models that can account for dynamics in hazard, exposure and vulnerability will enable us to anticipate future trends in risk and guide our cities towards a resilient trajectory. 

The development of a fully time-dependent risk analysis framework will be accomplished by creating
and integrating models for dynamic hazard, dynamic exposure and dynamic vulnerability. Multi-level agent-based modeling, remote sensing and machine learning techniques will be used to develop large-scale dynamic risk simulations applicable to a broad range of cities. Other improvements in current state-of-the-art disaster risk analysis models include the development of multi-variate fragility models.