Weather data for insurance risk modelling

Windstorm is the main peril responsible for property insurance loss in Europe.

Similarly, a succession of smaller events can cause significant aggregate losses in any season. Such exposure can be managed down by pricing risk at a local level and balancing the portfolio across high and low risk zones. This is only possible with an understanding of local hazard.

Our Euro Windstorm Historical Data Catalogue includes more than 6,000 real windstorm footprints and tracks to reveal the impact of a chosen storm or cluster of storms on a policy, local area or region.

Risk analysts, actuaries and underwriters can use this dataset to:

• reconcile historic losses against real storms such as Daria (1990) or Tini (2014)
• check on catastrophe model output
• use multiple storms to derive vulnerability curves
• evaluate Probable Maximum Losses (PMLs) when writing new policies

Windstorm Kyrill in 2007 caused €3.9-6.6Bn in insured losses.

Windstorm Xynthia in 2010 caused €1-4Bn in insured losses.

Why use Met Office data for risk management?

• Drawn from analysing 40 years of observations in 24 European countries
• 4km resolution to capture the effects of topography and land use and provide a detailed insight into loss
• Over 6,000 listed storms to help evaluate event frequency
• Footprints include 72-hour maximum gusts at every grid point of the Numerical Weather Prediction (NWP) model
• Tracks show the path of the storm
• Maximum wind gust at 10 metres above ground
• Combination of model skill and observations increases confidence in accuracy
• Delivered as model data output or in graphical format

Catastrophe modelling to estimate risk of natural hazards for insurance pricing

See our latest research on windstorm footprints, co-authored by Dr Laura Dawkins from the Met Office, and Prof David Stephenson from the University of Exeter.

Read the full published article in Natural Hazards Earth Systems Sciences.