How Snowflake's AI Identifies Buildings Facing Flood Risk

A cloud data platform has worked with the national mapping service to analyse flood vulnerability across England. Snowflake and Ordnance Survey identified approximately one million buildings that lack flood defences and face water damage risks.

The analysis appears in the Intelligent Flood Readiness Model. According to the project, the model uses Ordnance Survey's mapping data and Flood Risk Management Plans from government sources. The collaboration demonstrates how technology can integrate multiple data sources to address complex environmental challenges.

Combining multiple data sources

The model integrates six data streams into what the organisations call a structural intelligence layer. It compares Ordnance Survey building datasets with the Indices of Deprivation in England.

This comparison identifies where building characteristics such as height and construction type overlap with areas of social disadvantage. The model then adds Environment Agency flood data and defended and undefended flood risk areas.

An analysis of more than 3,000 pages of statutory Flood Risk Management Plan documents also feeds into the system. According to the project, this text analysis uses machine learning methods to extract relevant information that would be difficult to process manually.

The findings could help policymakers understand flooding probability in specific areas. They could also show which communities might face the greatest recovery challenges after flood events. This intelligence enables more targeted resource allocation and emergency planning.

Older buildings lack protection

According to the model's findings, 1.2 million buildings in England sit outside existing flood protection systems and face flooding risk. Approximately 68% of these buildings are in areas classified as deprived.

These areas often lack resources and infrastructure that support recovery after flooding. The analysis suggests 84% of undefended buildings were constructed before 2001.

Planning legislation introduced after that date requires flood risk assessment for new developments. The model shows 15% of at-risk buildings date from before 1919 and 23% from between 1919 and 1959.

Many of these structures were built before their locations were classified as flood zones. The findings also show England's varied housing stock, from tower blocks to Victorian homes. This diversity presents different challenges for flood protection and emergency response planning.

Data-driven flood planning

Fawad Qureshi, Global Field CTO at Snowflake, says: "Data is at the heart of making informed decisions. As this project shows, it's rare that one body holds all the relevant data or that this data is in the same format."

He adds: "But we're now in an era where technology can bring together the right people and the right data to collaborate on making better informed decisions."

Tim Chilton, Managing Geospatial Consultant at Ordnance Survey, says the collaboration aims to "help local authorities better understand, plan for and manage floods".

According to Chilton, the system provides "geospatial intelligence difficult to derive manually". He says: "Decision-makers can access data-driven, actionable insights – without the burden of analysing endless spreadsheets."

Flood planners face the challenge of tracking changes to natural and built environments. Flood Risk Management Plans are produced every six years and cover broad areas. This timeframe can mean plans become outdated as development patterns and climate conditions change.

Sustainability leaders won’t want to miss Sustainability LIVE: The Leadership Summit at London Climate Action Week, taking place at Code Node on 25 June 2026.

Register now for this exclusive invite-only event. 

The model's creators have outlined five recommendations for policymakers:

• Use granular data analysis to assess neighbourhood or individual building vulnerabilities rather than treating wide geographic areas as uniform
• Identify clusters of vulnerability, particularly where these cross local authority or Flood Risk Management Plan boundaries
• Invest in surface water infrastructure such as drainage systems, as most at-risk properties face surface water flooding
• Conduct vertical risk assessments that factor building height and develop specific protocols for high-rise structures
• Include social deprivation metrics in planning, as this influences community resilience and recovery capacity after flooding.

According to the project, two areas could have similar geography and building types but different levels of wealth or deprivation. This difference could affect recovery outcomes significantly.

The analysis suggests policymakers could use regularly updated data to create structural intelligence layers. This approach could replace reliance on static maps and separate databases that quickly become outdated.

Fawad says: "It's not the final answer, but it can inform the next question and help offer more protection to some of our most vulnerable neighbourhoods before the first drop of rain falls."

The model combines mapping, flood risk and deprivation data to support community protection planning. The approach could help authorities identify vulnerable properties as geographic data updates.

This could support more targeted and resource-efficient flood management strategies in areas facing increased flooding frequency. As climate patterns continue to shift, such dynamic planning tools may become essential for protecting vulnerable communities.