Skip to content

Search the site

Swiss Re buys Fathom, flood data specialist and The Stack's inaugural Tech for Good award winner

Globally, natural catastrophes caused $100 billion in insured losses this year alone, with “at least” $12 billion of these total insured losses attributed to flood-related events.

Swiss Re has agreed to buy Fathom – a data-led UK startup specialising in modelling flood risk and the winner of The Stack’s inaugural Tech for Good Awards in 2021 – for an undisclosed sum, it confirmed on Thursday.

Russell Higginbotham, CEO of Swiss Re Reinsurance Solutions, said: “We are very pleased to join forces with Fathom in our quest to narrow the protection gap for natural catastrophe risks, such as floods. Fathom’s market-leading research and innovative tools in this area create great synergies with Swiss Re’s risk knowledge and digital capabilities.”

Fathom, born out of the hydrology department at the University of Bristol, will retain its brand and continue its research activities, Swiss Re said.

Swiss Re Fathom acquisition comes as flood risk mounts

Swiss Re data shows that the re/insurance industry covered only 40% of the economic losses related to natural catastrophes in 2023, indicating a large protection gap. It estimates that, globally, natural catastrophes caused $100 billion in insured losses this year alone, with “at least” $12 billion of these total insured losses attributed to flood-related events.

That figure is over 30% higher than the past ten years’ annual average.

In a series of articles in The Stack over the past two years, Fathom has explained how it has built up its global flood risk modelling capabilities.

Machine learning transforms flood risk modelling

As COO Dr Andrew Smith explained in one: “Our data are used across multiple industries including (re)insurance, engineering, financial markets, corporate risk management and disaster response. 

“Until recently, our ability to build models at global scales was severely limited by the availability of accurate terrain data. In data poor areas, we were reliant on a legacy dataset that was derived from a radar product collected by NASA’s Space Shuttle, called the Shuttle Radar Topography Mission (SRTM). This dataset is over 20 years old and is riddled with instrumentation errors, alongside superfluous surface features... 

“These need to be removed in order to build an accurate flood model. This all changed last year when a new global terrain dataset called Copernicus GLO-30 (COPDEM30) was released by the European Space Agency. COPDEM30 is a ~30m resolution DSM collected by the TanDEM-X mission, a mission whose sensors operate at a native resolution of ~12m. 

He added in the guest article: “The precision and accuracy of the missions’ sensors ensured that COPDEM30 became the new gold standard in global terrain mapping, containing far more information than previous datasets. 

"However… to unlock the true functionality of the data, all the surface features needed to be stripped out. This is where the… flood hazard modelling community, came into the picture” – with a team later training machine learning algorithms to remove all the estimated surface artefacts in the COPDEM30 dataset, creating a new dataset that was a “significant step forward in our ability to model the bare surface of the Earth.”

Congratulations from The Stack to the Fathom team.

The Stack's Tech for Good awards return, free to enter, in early 2024. Get in touch today to learn more about sponsorship opportunities.