For Floods, Water Levels; for Drought, Satellites... Climate Insurance that Reads the Weather
More Data Isn't Enough... Basis Risk Must Be Reduced
"Advancing Risk Forecasting with AI and Satellites Is Crucial"

On the 26th of last month (local time), parts of Nuwakot, Nepal, were covered in mud due to flooding. Photo by Reuters Yonhap News Agency

On the 26th of last month (local time), parts of Nuwakot, Nepal, were covered in mud due to flooding. Photo by Reuters Yonhap News Agency

View original image

#Recently, large-scale flooding occurred in the border area between Nepal and China's Tibet, sweeping away villages and roads in the torrents. In downstream areas, the river level rose by up to 9 meters in just 30 minutes, rapidly spreading the damage. Critical infrastructure such as hydropower plants was also repeatedly impacted. Satellite image analysis revealed that part of a glacier located at an altitude of about 5,200 meters collapsed, plunging into a valley 1,200 meters below and causing a torrent mixed with ice, rocks, and mud. There were also indications that high temperatures immediately before the incident caused snow and ice to melt, possibly contributing to the destabilization of the mountainous terrain. Whether climate change directly triggered this collapse or not is still under investigation.


As climate-related disasters become increasingly unpredictable and severe, the insurance industry’s response is also being put to the test. With heatwaves and heavy rainfall occurring more frequently, traditional methods—such as visiting the site after an incident to calculate damages—are proving insufficient to fill the protection gap. Overseas, insurance companies are beginning to use not just weather indicators like temperature and precipitation but also satellite and sensor data for insurance payouts and risk assessments, thereby increasing their response speed. In Korea as well, efforts to build infrastructure that combines meteorological and insurance data are underway, with one of the key challenges being how precisely increasingly complex climate risks can be reflected.


As Disasters Grow, Insurance Changes Too… Climate Insurance Spreading Overseas

[Climate Insurance Blind Spot]②Overseas Climate Insurance Evolves with 'Weather Data'...Will Korea Also Fill Coverage Gaps? View original image

According to the insurance industry on September 1, the amount of payouts for domestic storm and flood, crop disaster, and livestock disaster insurance increased from 1.1891 trillion won in 2023 to 1.5701 trillion won last year. This represents a rise of about 32% over just two years. In particular, the proportion of storm and flood insurance claims due to heavy rain and flooding rose from 14.9% to 20.4% over the same period. For livestock disaster insurance, the share of policies triggered by heatwaves climbed from 66.9% to 80.3%.


As the number of heatwave days and the frequency of heavy rainfall increase, and because 'climate disasters' are now occurring in regions previously thought to be relatively safe, the insurance industry is seeking ways to identify and compensate for natural disaster risks more quickly. This is due to the fact that, in cases where the affected area is large or difficult to access, it can take a long time to assess damages through on-site investigations. A source in the financial sector stated, “These days, there are more instances where heavy rain falls over a certain area in a short period or where damage occurs in unexpected places. Disasters don’t wait for anyone, but if insurance payouts take a long time to calculate, timely compensation is impossible.”


In response, parametric insurance is being increasingly advanced abroad. Parametric insurance is different from conventional insurance in the way payouts are calculated. While conventional insurance investigates flooded buildings or damaged facilities before paying out according to the loss, parametric insurance pays an agreed amount if objective indices such as temperature, precipitation, wind speed, or water level exceed predetermined thresholds.


In Nepal, which recently suffered from large-scale flood damage, parametric flood insurance is being utilized mainly in western agricultural areas. Instead of requiring each farmer to individually prove the extent of their damage, insurance payouts are made if a pre-established flood risk index exceeds certain standards. For countries like Nepal, where large-scale natural disasters are frequent and mountainous areas abound, the advantages of this approach are even more evident. It is difficult to dispatch survey staff to every affected region, and if roads and communications are cut off, insurance assessments at the site are delayed or impossible.


In Kenya, Africa, parametric livestock insurance is offered to nomads exposed to risk of losing animals due to drought. Vegetation status is monitored by satellite, and if the condition of grasslands deteriorates below a set threshold, insurance is paid. Unlike previous systems that compensated after livestock deaths, this system provides payouts at the stage where grass becomes scarce due to severe drought, allowing nomads to purchase feed and water, or move livestock to other areas. By providing insurance payouts before a disaster reaches its worst stage, additional damage can be reduced.



On the 14th of last month (local time), amid ongoing heatwaves in Paris, France, people are swimming along the banks of the Seine River. Photo by EPA Yonhap News

On the 14th of last month (local time), amid ongoing heatwaves in Paris, France, people are swimming along the banks of the Seine River. Photo by EPA Yonhap News

View original image

In Europe, where heatwaves are becoming more frequent, the use of satellite and sensor technology in insurance is increasing. Examples include insurance based on soil moisture measurement via satellite or water level sensor data. Utilizing meteorological and environmental data makes it possible to design insurance coverage not only for visible damage such as facility destruction, but also for risks like reduced yields or higher costs due to drought. The scope of compensation for losses caused by natural disasters is therefore broadening.


In Japan, there is a movement to insure against even corporate revenue declines or additional costs that arise solely from changes in the weather. For example, if heatwaves reduce the number of visitors to outdoor leisure facilities, an unusually warm winter slashes winter clothing sales, or insufficient snowfall causes ski resort sales to decrease, these scenarios are all covered. Risk from reduced solar power sales revenue due to insufficient sunlight is also included.


'Filling Coverage Gaps'… Data Holds the Key for Climate Insurance

[Climate Insurance Blind Spot]②Overseas Climate Insurance Evolves with 'Weather Data'...Will Korea Also Fill Coverage Gaps? View original image

The insurance industry’s response to climate change goes beyond merely changing the way payouts are made. The use of data is also expanding in the evaluation of climate risks and the design of insurance products. Global property insurer FM has built a 'Global Hail Risk Map' by combining more than 500,000 hail observation records from 1955 to 2024 with satellite and reanalysis data. This allows risk to be estimated even for regions with insufficient past hail observations, and is used to assess the exposure of roofs, exterior walls, solar panels, and HVAC systems to hail. Angelika Werner, Vice President of Climate Risk and Resilience Research at FM, warned, “Extreme weather events are operational risks that companies have to manage on a daily basis.”


In Korea, the General Insurance Association of Korea and the Korea Meteorological Industry Technology Institute launched the project last month to establish the 'Weather and Climate Data-Based Climate Insurance Comprehensive Portal System.' The aim is to leverage meteorological climate data and insurance information on a unified platform to enhance insurers’ climate risk analysis and the development of climate insurance products. The new portal is expected to include features such as automatic determination of whether a claim meets compensation conditions for each type of climate insurance, functions for property insurers to utilize the Korea Meteorological Administration’s APIs, and weather statistics and analysis features specialized for insurance.


However, simply using weather data does not solve all problems. This is because of what is known as 'basis risk,' where actual damage and the criteria for insurance payouts may not always align. For example, within the same area, the extent of damage may vary depending on soil, terrain, or crop varieties, but basing insurance solely on a single indicator like precipitation can lead to a mismatch between real losses and compensation. In other words, insurance payouts might be triggered when heavy rainfall meets payment criteria even if actual damage is limited, or significant damage occurs but the relevant index fails to reach the threshold, resulting in no compensation. An insurance industry insider commented, “It’s just as important to reduce discrepancies between actual damage and payout standards as it is to gather lots of data. To improve compensation accuracy, it is necessary to segment characteristics by region and by risk type and to combine multiple indices.”



Experts point out that the larger the market for insurance covering climate disasters grows, the greater the need for more granular and accurate data. Rather than relying on single weather observation data, it is necessary to utilize micro-regional weather data and a variety of indices, as well as to incorporate data from satellites and Internet of Things (IoT) sensors. Kihong Lee, Research Fellow at Hana Institute of Finance, suggested, “It will be important to enhance the risk prediction system using artificial intelligence, satellite information, and weather data, and to expand the use of parametric insurance to improve the speed of insurance payouts and the operational efficiency of climate insurance.”


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing