Floodly machine-learning flood prediction tool

Until now, understanding the relationship between rainfall and river flooding has involved building hydrological models. These require powerful and often costly modelling software and details of factors such as topography, river levels and the responses of all the influences in the catchment – data that is particularly challenging to obtain in complex urban areas. As a result, building a hydraulic model takes a lot of time and effort.

Our water and smart consulting teams took a different approach. They asked: could machine learning use historic data sources to predict river levels? 

For catchments in the south-east of England they matched many years of rain data pre-2000 to 2015 with river level data over the same period, then tracked patterns in the data. Using algorithms, they could accurately predict river levels from 2015 onwards, using only rain gauge data.

“In the catchments we examined we can select a specific time period and get a prediction of river level and timing within seconds. This is a task that typically takes minutes, hours or even days with traditional modelling.”

~ Andy Porter (Smart Consulting),  WSP

Floodly is not designed to replace hydraulic modelling, but to complement it by providing a more rapid response to predict river levels. In the future it could even be combined with satellite weather data to provide precisely targeted advance warnings of river flooding.

Our team have already demonstrated Floodly to lead local flood authorities in London, the Environment Agency, the Met Office and authorities in New Zealand. Having proved the concept, they will now work with partners to develop and refine the tool. In December 2019, WSP was the runner up in the Environment Agency’s hackathon and was the leading flood forecasting submission. This truly shows the potential and need for new and innovative, data-driven flood warning tools!

  “Floodly was demonstrated at the Environment Agency’s Flood Warning Hack in November 2019 as a potential solution to the problem of generating flood warnings in hard-to-reach locations.  Judges at the Hack were impressed with how Floodly could turn rainfall forecasts into river level predictions, even in catchments with very limited hydrometric data records.  The machine learning element, with the potential to automatically refine and improve forecasts through time, was a further interesting development.  Floodly was shortlisted to be considered for a trial as part of the Environment Agency’s Flood Warning Expansion Project (FWEP) Discovery phase.”

~ Ben McCarthy (Incident Management & Resilience),  Environment Agency


With climate change expected to make intense rainfall more likely, Floodly could play an important part in future flood response.