Finance and insurance
Extreme weather events are a major source of insured losses across the globe. For example, the extreme winds and precipitation associated with tropical cyclones (hurricanes and typhoons) result in huge insured losses in parts of the US, Asia, Latin America and Australia. In Europe and at higher latitudes of the US, winter wind storms (extra-tropical cyclones) can cause considerable losses through their high winds and intense rain or snow. Convective storms which give rise to hail, thunder and lightning and tornadoes also cause large insured losses in many regions, and in fact, in 2014, a series of such storms in the US resulted in the largest insured loss event of the year.
Insurance and re-insurance companies need to understand the risk associated with extreme weather so they can make appropriate financial plans. In Europe, insurers are required event by the Solvency II regulations to hold enough capital to withstand the losses of a 1 in 200 year event. For weather events such as those mentioned above, estimating these losses requires a thorough knowledge of the long term variability of the event, and how it may respond to climate change.
In Europe, large insured losses are caused by extra-tropical cyclones (ETCs), flooding (which is often also associated with ETCs) and hail from convective storms. The ETCs usually originate off the west coast of the US and travel along the North Atlantic storm track towards Europe. Climate models are a useful tool for understanding the variability of the frequency, intensity and location of the ETCs, but low spatial resolution models tend to simulate ETCs which are too low in intensity, and a storm track which is too zonal, meaning that the ETCs impact Europe further south than observed. Some higher resolution models, however, have been found to improve these biases. The high resolution PRIMAVERA models will therefore be very useful for improving assessments of the risk posed by ETCs.
Convective storms cannot be explicitly modelled at the typical resolution of global or even regional climate models (10—300km). Instead, researchers look at the larger scale atmospheric environments favourable to convective storm generation, such as high convective available potential energy (CAPE). With the high resolution of PRIMAVERA models, it is expected that the spatial variability of these proxy measures will improve, giving a better understanding of convective storm risk.
Many insurance and re-insurance companies are already engaged with the weather and climate research community. For example, several companies have collaborated with universities to produce European ETC event sets. These event sets are often created from low resolution, long-running climate simulations which are then regionally downscaled. Event sets created from PRIMAVERA models provide a useful benchmark for event sets developed within the industry, and perhaps improve understanding of the biases present in event sets created from lower resolution models
The PRIMAVERA models could also be used to answer several important questions for the industry on the long term behaviour of extreme weather events. In the case of European ETCs, temporal clustering of storms has been found to have a large impact on insured losses, yet the process is not yet fully understood. In addition, it is also useful for insurers to know ahead if a particular stormy winter is expected, so understanding how ETC frequency and intensity is related to predictable large scale climate indices such as El Niño Southern Oscillation and the North Atlantic Oscillation would be particularly useful.
It is essential to work closely with insurance and re-insurance companies to ensure that the meteorological data given is compatible with insurers’ vulnerability functions, and that it is in a usable format.
Some organisations have developed their own resources to explain how weather and climate affects their businees in this sector. Please contact us to share any resource (text, videos, slideshows) on this topic.