[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries

On the LiDAR

As environmental risks become an everyday threat to the market, there could be huge value in contributed claims data and new mapping technology, says Richard Toomey


© iStock
© iStock

As more data is gathered from carriers across the market on property claims, the insights will grow

The World Economic Forum’s 2019 Global Risks Report stated that increasing instances of extreme weather stood out as both high-likelihood and high-impact, underlining the challenge actuaries face today when pricing for environmental risks.

According to the Association of British Insurers (ABI), the ‘Beast from the East’ cold wave of early 2018 cost UK insurance providers more than £300m in claims during the first quarter of that year. This was before an extended summer heatwave brought a peak in subsidence claims. When it comes to flood-related claims, while June 2019 was one of the wettest on record – with severe flooding in Lincolnshire – the winter of 2015 set the record that is yet to be broken, standing at more than £1bn.

Climate change is undeniable – and as changes in the UK’s weather patterns have become more apparent, the breadth of data around flooding, subsidence and fire risk, combined with geospatial mapping tools, have expanded, offering actuaries deeper insights into environmental risks to help them price for future losses. However, the current picture – created by the layers of information now available to actuaries, including their own historical claims data – omits the insights gained through access to market-wide claims data related to both properties and policyholders. 

Deep dive into data

Today, actuaries have access to detailed data on property characteristics such as square footage, construction, roof type, height, security locks, window size, and where it sits in proximity to other buildings. They are also able to access highly detailed risk models to determine how a property portfolio might be impacted by flood, fire and subsidence, based on geospatial analysis. For larger, usually commercial properties, it is even possible to determine the difference in flood risk between the front of the property and the back. This data can be visualised using mapping tools, allowing insurers to identify exposures when an extreme event occurs and accumulations to help determine underwriting strategies.

The most recent advances in the understanding of environmental risk can be attributed to light detection and ranging (LiDAR) technology. LiDAR is a remote sensing technology that uses the pulse from a laser to collect measurements that can then be used to create three-dimensional models and maps of objects and environments – it is one of the technologies underpinning the driverless car. LiDAR is already proving valuable in helping the insurance sector model for environmental risks using predictive analytics.

As the cost of using this technology has fallen, the flood, fire and subsidence risk models created through the images captured have become much more detailed. Three-dimensional images of the terrain can be built, showing buildings, trees and water levels for the whole of the UK. The insights provided are particularly valuable when it comes to flood modelling, as the data can be coupled with soil type, average rainfall data and river gauge monitoring, and used to accurately predict where water will go during a flood. It can also be used to show the effectiveness of man-made flood defences, and any potential knock-on effects they might cause.

As the ABI reported recently, these flood defences are extremely valuable, having prevented an average of £1.3bn worth of additional damage each year. However, to remain effective, they must be maintained, updated and extended, and their impact on other areas assessed. Floodwater must go somewhere, and LiDAR images can predict where it will go and what impact it could have in that area. 

Figure 1
Figure 1- LexisNexis ®Map View visualisation of flood risk from river for property marked 'A'. ®Map View can risk score the property based on the insurer's underwriting rules and the geospatial location of the property in relation to various hazards.

Using past claims data

However, even the latest technology can be defeated by nature. For example, flooding can be caused by blocked culverts – from illegal dumping, for example, which carries its own environmental risks – and this may not be seen using LiDAR or other mapping technologies. Other factors can also contribute to flooding, and this is where past claims data can fill in the knowledge gaps. Through accessing a picture of localised claims, as an added layer of risk data, an insurance provider can see that there is a problem – and therefore an increased risk of further claims in the area. 

An insurance provider looking at only its own data may not be able to see a pattern, make any predictions on future claims, or take preventative measures. Bringing together claims from all providers widens the view, and provides a more holistic and revealing picture of the risk.

The same is true for fire risks. Fire models predict how a fire might spread within a property and from building to building, based on proximity as well as building outlines and heights – but there will always be cases that deviate from the predictions. Here, past claims data can provide additional insight into the risk for an individual property, highlighting previously unknown risk factors such as a nearby takeaway or another high-risk commercial property.

Subsidence can also give rise to high-value claims. Currently, an actuary pricing for subsidence risk will review models based on weather patterns and soil data. The next stage, which is currently being explored, is the use of tree data to improve the accuracy of subsidence risk modelling. Tree data used in conjunction with soil data could create a clearer understanding of the subsidence caused by tree roots pulling moisture from the ground. Adding prior claims for subsidence to the model – not just for that property, but also for those around it – would bring a deeper understanding of the risk, allowing for more accurate pricing calculations for future losses. 

In the US, commercial property carriers have recently started to use past claims related to the property, and to the people behind the business, to understand risk.  In the early analysis, we have been able to identify that prior personal insurance claims correlate with higher frequency and severity of commercial claims. As more data is gathered from carriers across the market on property claims, these insights will grow.

Figure 2
Figure 2- LexisNexis® Map View visualisation of building blocks identifies accumulation of risk from insured building marked as 'A'.Using specific rules, the insurance provider can understand the potential impact from fire risk at various distances from neighbouring buildings and building heights.
Figure 3
Figure 3- A correlation between the number of personal insurance claims by business owners or directors in the US and commercial insurance claims cost and count.

Wider environmental risk factors

Of all the natural environmental risks facing the insurance sector today, flooding is perhaps the biggest threat. However, man-made environmental threats are also a major risk. Vacant buildings and plots are targets for fly-tippers, while waste disposal sites represent significant fire risks, so there is increasing demand to build specific models for these risks. 

Fundamentally, the more data added to all environmental risk models, the greater the ability to identify correlations between risks and potential claims. Access to industry-wide past claims data, coupled with valuable insights from the latest LiDAR and mapping technology, will enable actuaries to price more accurately for the annual average loss. It’s starting to work for our US counterparts; it’s time the UK followed suit and began reaping the benefits, too.

Richard Toomey is data analytics manager at LexisNexis Risk Solutions UK and Ireland.