Natural Catastrophes

 

 

Regional Extreme Weather Risk assessment
(Walker Institute; University of Reading; University of Colorado)

Climate risk is regional in its impact and scientific research must provide regional and local assessments of the potential variability in extreme weather events to be of use to risk managers. Willis Research Fellows and WRN Senior Academics from partners including the University of Reading and Colorado University are working to define a regional risk assessment project that will for the first time provide insurers with consistent and robust risk information based on state of the art climate science.

 

 

Flood correlations and clusters
(Durham University)

Willis Research Fellow Emma Raven at Durham University is conducting new research into the temporal and spatial clustering of flood events, using long-period historical records and statistical methods to identify climate drivers of flood risk and flood poor periods, plus the potential for clustering over insurance periods.

 

Mega city flood risks – Asia and Europe
(National University of Singapore; Bristol University)

Urban areas are poorly represented within current catastrophe models. WRN members are adopting novel computational methods to improve the representation of urban areas within flood models and the generation of loss estimates that reflect the variability within urban flood propagation. Jakarta and Central Europe are being used as case studies to explain the ways in which this work can be effectively employed for insurance risk assessment.

 

Seismic post event damage assessment – using remote sensing tools
(Cambridge University)

Willis Research Fellow Keiko Saito is undertaking work at Cambridge to develop methods for remote assessment of damage after extreme events. Recent work uses the example of the L’Aquila earthquake and immediate post event monitoring of damages from high resolution satellite imagery.

 

Insurance Risk from Volcanic Eruptions

(Cambridge University; University of Naples)

WRN partners from Cambridge have been working with in-house Willis experts to develop a risk ranking for European volcanoes identifying in a uniform way the populations which may be at risk from the expected eruptions of Europe’s most dangerous volcanoes. This research has already identified the 10 European volcanoes with potentially affected populations greater than 10,000, and with an aggregated exposed property value at risk of US$85 billion. Future research will examine the possibility of modelling volcanic risk for the insurance industry and working towards a better understanding of scale of future impacts.

 

 

 

Vulnerability, Exposure & Risk

 

Global exposure inventories
(Cambridge University)

Partners from Cambridge University are developing methods for extraction of key building stock parameters for catastrophe model input from earth observation datasets. The work includes case studies being used to develop new methods for the characterisation of key variables such as roof type, age and construction, consistent across regions and territories.

 

Time-varying exposure
(City University London)

Catastrophe risk is not constant – exposure varies in time and space but is fundamental to the representation of population and transport risks. For some perils, time can dictate loss potential – and the coincidence between time varying hazard (e.g. hail) and exposure (e.g. traffic flow) become the key risk factor. The WRN is conducting research into time varying exposure, focussing on traffic flows and hail as a key peril, population movement within urban centres and the data sources (including mobile communications data) that can potentially be used to represent population and transport risks.

 

Visualising and communicating risk – desktop tools for better risk reporting

(City University London) 

As data from catastrophe modelling  becomes more and more complex, so the need to represent the patterns of risk, particularly to non-specialist decision makers, become increasingly critical to successful risk analysis. Work being undertaken by WRN partners at City University is examining complex data visualisation including building stock and geo-demographic data, GCM outputs and eventually tsunami risk mapping.

 

 

Demand surge

(University of Colorado, Kyoto University)

Willis Research Fellows from the University of Colorado and Kyoto University are working to build robust models of demand surge based on a fundamental review of historical events and their impacts on demand surge loss potential. The work also encompasses modelling the role of infrastructure and complex systems on loss potential in large urban areas.