2025/11/24 by Matthew D. K. Priestley, Hannah Bloomfield · 1 voice
Agricultural and Biological Sciences · Economics, Econometrics and Finance · Environmental Science · #Agricultural risk and resilience #Insurance and Financial Risk Management #Flood Risk Assessment and Management
paper · pdf · doi:10.1002/wea.70009
Most of us hold some form of an insurance policy. Insurance serves to protect people and businesses from the risk of financial loss when unexpected or impactful events occur. At its simplest, insurance works by having people pay a regular amount, called a premium, to an insurance company. In return, if a covered event (e.g. flooded property, car crash) does happen, the insurance company will help cover the costs. This makes it easier for people to recover by mitigating the financial impact. Insurance is offered for almost anything that is at risk. The most common policies are to cover risks to cars, health, life, home/contents and other personal possessions. Insurance companies calculate premiums based on risk (or likelihood) assessments and the value of what is being covered. Factors like age (of the policy holder, or their items), health or geographic location are considered to determine the likelihood and impact of an event happening. For example, young men often have the highest car insurance premiums as they are considered more likely to be involved in accidents and therefore would require financial reimbursement more often. Climate and weather conditions also play a significant role in this process, as areas prone to natural hazards carry higher risks and therefore higher premiums. In the event of extreme weather events there can be major impacts on the insurance industry, especially for property and home insurance. Severe weather events – like hurricanes, floods and wildfires – can cause huge losses. For example, Hurricane Katrina in 2005 led to approximately 65 billion in insured losses (Swiss Re, 2020), the 1999 European windstorm series of Lothar and Martin caused nearly €10 billion in losses (PERILS, 2024) and the January 2025 California wildfires are estimated to cost insurers 20–30 billion (Moody’s, 2025). With a warming climate, many hazards are increasing in both frequency and severity, which means insurers are facing higher costs and customers in high-risk areas may see increased premiums to cover this. Insurance companies analyse past weather data and historical climate trends to assess which areas are at highest risk. If a region is particularly prone to natural disasters, insurers may charge more for policies there, or even limit the coverage they offer. In 2012, for instance, Hurricane Sandy caused about 30 billion in insured losses, prompting insurers to re-evaluate policies and raise rates in flood-prone areas. Recently it was reported that several companies withdrew coverage for wildfires in California,1 citing too high risk and costs. This region was subsequently ravaged by wildfires in early 2025. This insurance withdrawal can make it harder and more expensive for people in these locations to get coverage, and sometimes governments step in to help provide insurance where private options are limited. As climate change drives more extreme weather, the insurance industry is working to adapt by focusing on ways to reduce risks before disasters strike. Some companies invest in technologies like early warning systems or in projects that help communities plan for extreme weather. Other companies are exploring how their catastrophe risk models can be used to design adaptation solutions to climate change (Galloway et al., 2025). These proactive approaches aim to lower potential losses, keeping costs down for both insurers and customers and helping communities stay safer. There are numerous tools for understanding how high-impact weather events and changes in climate can impact losses. All insurers have a list of historical claims, so they can understand where/when losses have occurred. These losses are ascribed to weather events using historical meteorological observations, or gridded atmospheric data products such as reanalyses. By examining these losses, and how they vary with relevant meteorological hazards (e.g. wind gust, storm surge level, surface water flood depth), relationships can be constructed as to how changes in weather phenomena drive variability in losses across space and through time. However, the range of plausible extreme weather events is not limited to those we have observed. The events that occur next week or next year will be unique and an understanding of those is required. This is where data products such as large ensembles, or historical climate model runs (e.g. Thompson et al., 2017) can help boost the number of possible weather events for examination. Through the examination of alternate histories, or ‘what-if’ scenarios, the true risk can be explored. For example, Rye and Boyd (2022) found that alternate versions (e.g. landfall in more impactful locations, higher intensity) of some impactful US hurricanes could increase losses by as much as 300× the observed values. Assessments of changes or trends in physical hazards are also of importance for understanding how an insurance portfolio may be at risk in the future. These assessments are often uncertain due to the difficulty of representing these phenomena in global climate models due to their limited spatial and temporal resolutions (Strandberg and Lind, 2021). Some examples of hazard change assessments are for tropical cyclones (Knutson et al., 2020), European windstorms (e.g. Priestley et al., 2024), flood risk (Wing et al., 2018), hailstorms (Rädler et al., 2019) and wildfire (e.g. Liu et al., 2013). Once hazard changes have been quantified, a full assessment of how climate change will affect risk can be made by translating these changes into adjustments within a catastrophe model (e.g. Jewson, 2021). An insight into the insurance impact of weather events can often be made using publicly available data. For extratropical cyclones (which often bring extreme wind gusts and precipitation accumulations (Jones et al., 2024)), metrics such as the storm severity index (SSI; Klawa and Ulbrich, 2003) and flood severity index (FSI; Bloomfield et al., 2023) can be used to infer damages, or to compare the impact of different historical storms. These metrics use simple formulations that relate the wind gust (or precipitation accumulation), above a locally varying damage threshold, to the level of damage. This can be demonstrated using Storm Daria (also known as the Burns’ Day Storm; Figure 1a), which impacted the UK and Western Europe on 25/26 January 1990 (McCallum, 1990). During this event, 47 people lost their lives in the UK, with insured damages of >8 billion (Roberts et al., 2014). From Figure 1(b) it is evident that Daria had the most damage potential for southwest England and southern Wales, as well as the Dutch coast. This does not necessarily reflect the areas of greatest impact, as for there to be damages, the density of insured property (e.g. the exposure) and the vulnerability of these buildings must also be high. With a changing climate, it is highly likely that the risk from extreme weather events will increase, putting more pressure on the insurance sector. To maintain resilience and financial stability, insurers must integrate high-quality weather and climate data into their risk assessment frameworks. The effective use and continual improvement of these tools are essential for accurately quantifying evolving risks, guiding adaptation strategies and ensuring the sector’s ability to safeguard communities and economies in an increasingly uncertain climate future. Matthew D. K. Priestley: Conceptualization; writing – original draft; writing – review and editing; visualization. Hannah C. Bloomfield: Writing – original draft; writing – review and editing.