Thunderstruck

Summertime in Canada can bring hot and sometimes humid weather to many of the country’s most developed areas in the south central and southeastern regions.

The jet stream tends to migrate toward Canada at that time of year, flowing across many of the country’s major cities. It often combines with unstable weather systems that form near the Rocky Mountains or the Great Lakes, or are carried north and east from the Gulf of Mexico across the United States. This interaction can result in severe thunderstorms that bring damaging hail, tornadoes and straight-line winds.

The risk that severe thunderstorms in Canada pose to insurers is, as in the United States, largely the result of the accumulation of losses from multiple events, contributing to a higher annual aggregate loss. However, several historical events stand out as causing significant loss in Canada on their own, many producing significant losses from a combination of perils.

PERILS OF HAIL, TORNADOES AND STRAIGHT-LINE WINDS

Most costly thunderstorm peril

To date, the costliest peril associated with severe thunderstorm events is hail – and Alberta is the site of one of North America’s most volatile hail zones. Hail damage in this region is so great that some companies have used the controversial (and scientifically unproven) practice of seeding clouds with crystalline silver iodide in the hopes of reducing the amount of hail in high-exposure areas.

Hailstones – which can be as small as a pea or as large as a softball – are capable of causing significant damage to buildings, automobiles and crops in Canada. The July 12, 2010, hailstorm in Calgary, for example, caused insurable damage of approximately $450 million to buildings alone.

Canada has the world’s second highest frequency of tornado events, with about 80 to 100 reported tornadoes each year. (The United States ranks first, with an average of 1,000 to 1,200 annually.)

Recent research by scientists at Environment Canada indicates that approximately 100 additional unreported tornadoes may occur annually in Canada; population bias is a crucial consideration when analyzing the severe thunderstorm reports.

Most intense thunderstorm peril

One of Canada’s most significant tornado events occurred on June 22, 2007, when the country’s only F5 tornado on record – F5 is the highest level of intensity on the Fujita scale, which is based on the damage from a tornado, a rare event in any part of the world – touched down 40 kilometres outside of Winnipeg. The intense twister, measuring 300 metres wide, left a 5.5-kilometre trail of destruction as it tore homes from their foundations and sent cars and trucks spinning off highways.

That year, eight tornadoes occurred in Manitoba, including an F3 just south of Baldur, a small community about 200 kilometres southwest of Winnipeg. And on August 21, 2011, an F3 tornado with a width of 1,500 metres formed over the Great Lakes and, accompanied by hail and violent straight-line winds, tore through the picturesque and historic town of Goderich, Ontario, causing more than $130 million of damage.

Although tornadoes are capable of producing damage and losses far greater than hail, they usually do not because they generally have narrower swaths than hailstorms.

Most frequent thunderstorm peril

Often, damage attributed to a tornado appears to be spread in one direction. In these cases, the true culprit of the damage is, in fact, straight-line wind, which is the most frequent peril associated with severe thunderstorms. While they lack the dramatic appearance of tornadoes, straight-line winds can cause a similar amount of damage that can stretch across hundreds of kilometres.

UNDERSTANDING SEVERE THUNDERSTORM POTENTIAL

Insurers and reinsurers concerned with the exposure of their portfolios to severe thunderstorms need to understand where future storms are likely to occur, how strong they may be and the potential damage that could result.

In the U.S., the Storm Prediction Center collects point reports of historical storm data – containing the time of the storm, both its location and intensity, and information on damage and other observations – from the public, the media and trained weather spotters.

Currently, storm reports in Canada are not very comprehensive or consistent, although efforts are under way by scientists at Environment Canada to create a comprehensive database, starting with the tornado subperil.

An alternate method of analyzing severe thunderstorm risk that supplements the historical data is the use of reanalysis data. The National Centers for Environmental Prediction, an organization within the U.S. National Oceanic and Atmospheric Administration, initiated a project to provide atmospheric data, at high spatial and temporal resolution, for climate studies.

The result of this project is the Climate Forecast Systems Reanalysis (CFSR), which is a worldwide dataset that represents a “best estimate” of the state of the atmosphere through time.

CFSR is based on a type of atmospheric simulation model known as a numerical weather prediction (NWP) model. Available observational data from surface-based weather stations, ocean buoys, weather satellites, precipitation gauges and weather balloons are input into the model.

This detailed information allows an NWP model to estimate data values in locations where observational data is not available, thereby providing complete and seamless spatial coverage.

CFSR data, which include records as far back as 1979, provide key information about the state of the atmosphere at the time of past severe thunderstorm events.

Conditions that are conducive to severe thunderstorm formation – including moisture, instability, rotation and lift – can be used to determine where events of each type of peril were likely to have occurred.

For example, reanalysis data from July 12, 2010 was used to indicate regions of high potential for hail formation (see Figure on page 27) and these results match well with areas where actual hail events were noted.

CLOSING THOUGHTS

Damage footprints of severe thunderstorms can be small, and given Canada’s vast land mass, observational data and damage claims from historical events do not provide a complete view of the risk across the country (or even within a province).

Using reanalysis data, a comprehensive view of risk in Canada can be achieved even without a large database of historical storm data.

And since reanalysis data are not based on human-reported events, the information is unbiased by both population growth and population distribution, which is a major advantage when analyzing storm risk in Canada.

Studies that incorporate both observations and CFSR reanalysis data indicate that severe thunderstorm risk exists to some degree in all of the Canadian provinces. The atmosphere is less conducive to severe thunderstorm risk in the far northern areas, which are sparsely populated, as well as in the Maritimes, although the latter is susceptible to tropical cyclones.

As Canada’s population is concentrated in some areas and sparse in others, the risk varies a great deal depending on the location. A worst-case scenario could result from either a single large tornado in a major city, or an extended outbreak that affects large regions.

Effective management of severe thunderstorm risk, therefore, requires a robust view that takes account of a full range of scenarios that reflect the true risk in all areas of the country.