Natural Disaster Risk Analysis in the Big Data Era
Prof. Dr Seth Guikema | University of Michigan, USA
Natural disasters regularly impact society leading to large economic losses, loss of life, and loss of critical infrastructure service.
Modern data analytic methods combined with an increasing amount of data on the impacts of these events allows for the development of predictive models that can be used to better estimate risk, both in the short-term before a forecast event and over the longer-term considering the potential influences of climate change on some types of hazards. This talk discusses the strengths and limitations of adopting a data-driven approach to disaster risk analysis. It uses as an example power outage risk models developed over the past decade for hurricanes and other types of weather events. These models leverage large, diverse data sources and statistical learning theory methods to provide estimates of event impacts. The importance of considering uncertainty in the impact forecasts is highlighted. The talk closes by offering a perspective on possible future developments in the area of data-driven risk analysis.
Organizer
ETH Risk Center
Scheuchzerstrasse 7
8092 Zurich | Switzerland
More information on the seminar