CC Tool

The next five years of warming from human activity will add around 1% to the SCR for a typical insurance company. By contrast, uncertainty in baseline risk is around 10 to 20% and currently amplified by inflation, supply chain disruption and war. The CC Tool provides CC solutions which are proportionate to impacts on insurance companies. 

The user inputs a baseline climate YLT with aggregate losses and their scenario hazard choices for the peril, then clicks 'submit' and after a few seconds, a scenario YLT is output containing the same events with modified loss severities. The loss changes are validated using independent estimates (including from main vendors) together with numerical testing of modelling assumptions.

The key is an 'EP-matching-and-mapping' technique used by some NatCat modellers at smaller analysis scales, and developed to work at aggregate scales in the CC Tool. The powerful method is agnostic of model vendor and exposure, and speed is assured by its agg-scale analysis.

The Tool’s scenario hazard changes and estimates of loss impacts are outlined in the table to the right, for a climate with 2°C global-mean warming.

Documentation was a priority for the CC Tool. It includes a beginner’s guide to climate change and its impacts, a section on 'EP-matching-and-mapping', then sections for each of six different perils, with methods, results and validation fully described. It is fit for regulatory purposes.

Please contact us if you want to know more.

Europe Windstorm History, 1940-2023

• Industry losses for top 1000 events for 16 countries (+EU)
• Wind gusts at 25km resolution for top 1000 events

Based on ERA5 reanalyses data. Key features include:

1. Using 10-min mean winds rather than gusts from ERA5
2. With assimilation of long records of observed gusts
3. Klawa and Ulbrich (2003) relation converts wind to loss
4. Full set of trended historical losses for validation
5. Documented in a manuscript currently in review (available on request)

ERA5 timestep-mean winds accurately simulate the spatial variations of storm winds. However, they have a natural offset from the gusts that cause damage, and also contain misleading longer-term trends due to upgrades in observational systems over time. Stormwise have overcome these limitations to realise the full potential of ERA5's skilful simulation of storm winds.

Hazard for two major storms are shown: Lower Saxony in Nov 1972, and 87J in Oct 1987. Both storms deepened explosively just off mainland Europe, and, after correction with observed gust data, the quite remarkable accuracy of ERA5 simulations for these  intense events is typical of many storms.

The resulting national industry losses provide an extra view of historical industry losses, and a little processing gives insights into climate variability. For example, the barplot to the right shows how the longer-term 1940-2023 average loss in Europe is 20 to 25% lower than a baseline climate based on the past 50 years of windstorms. The past 3, 4 or 5 decades is not enough to form a robust baseline loss climate for this thick-tailed peril.

These datasets can help with both loss and hazard validation of a peril model, including the stability of its baseline climate.

Please contact us if you want to know more.