In Series: Windstorms: Hindcast is a wonderful thing


Use of the right meteorological data is vital for insurers.

Post and the Met Office recently combined efforts to research how insurers use natural hazard science to manage property exposure.

The results showed that meteorological analysis is valued at every step in the insurance process. Post’s recent articles on European windstorm show that this is potentially Europe’s most important peril; so how does meteorological information help insurers manage the risk?

Somewhere off the New England coast, riding on the Labrador current, a mass of cold Canadian air meets with tropical moisture drawn up from the western Atlantic. As these air masses collide, the instabilities begin to form an extra-tropical cyclone that is soon observable from satellite and surface observation systems.

A week later in northern Europe, it could just be a typical damp day. Or it could become a catastrophic windstorm, causing injury, displacement of people and greater insurance loss than from any other peril.

Scientists studying storm frequency, severity and tracks have developed techniques to better understand the risk. At every point in the insurance process, from capital reserving through to the settlement of claims, understanding the meteorology will help.

For catastrophe modellers, the ‘hindcast’ – a re-run of a numerical weather prediction model over an historical data set – is an invaluable source. It can be used to identify and describe past extreme events at a high resolution. An historical catalogue of thousands of real events can be developed in order to understand variability over a period and reveal characteristics such as clustering, a topic currently of significant interest among catastrophe modellers. The most severe named storms for which insurance loss records exist, such as Kyrill or 87J, can be modelled and manipulated to test the accuracy of a loss analysis.

Underwriters, actuaries and analysts can also benefit from hindcast output. A statistical analysis of the most extreme values in the data can be used to create hazard maps representing multiple return periods, up to one in 200 years. Hazard maps are easily used in geographic information systems to compare gust values with local exposures, enabling portfolio balancing between high and low risk areas.

Meanwhile, our windstorm is developing and tracking north-east. Ships and aircraft report exceptionally strong winds. Met services ‘ensemble’ forecasts model multiple possible outcomes, and meteorologists compare them to reach consensus on the evolution of the event.

Days before any public weather warnings are published, alerts are available to insurers, together with high-resolution footprints showing the likely severity and track of our storm. Footprints are updated daily and can be compared to historical events for a better understanding of likely loss. Typically one to three days beforehand, operational decisions such as emergency staffing will hinge on the skill and detail of the forecast.

Our storm has now been named and public weather alerts issued. In common with many events from the last two winters, it crosses Scotland and moves south, causing damage on the UK east coast and northern Europe.

Attention turns to analysis of what happened. A high resolution footprint, issued 24-48 hours after the event, can be used for a rapid loss forecast. For the next few weeks claims departments can use this footprint and weather station observations for the validation of claims. Our storm now enters the catalogue for future analysis and the clock is reset to zero.

The average windstorm claim may not be as big as a flood claim nor as well publicised, but there are so many more of them that a single event could have a huge impact on the industry. This makes meteorological proof valuable and even essential to ease the burden on loss adjusters.

In short, use of the right meteorological information at every stage of the insurance cycle will help control the risk, increase customer responsiveness and ease the bottom line.

Paul Maisey
Science team leader for insurance, Met Office

  • LinkedIn  
  • Save this article
  • Print this page  

You need to sign in to use this feature. If you don’t have an Insurance Post account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an indvidual account here: