How is Google Using AI to Scale Up Flood Forecasting?

“Valencia: 220 and rising. Hurricane Helene: 230. Hurricane Milton: 250. Storm Boris: 26. Rio Grande do Sul in Brazil: 176. Over 1000 victims across West Africa. I could go on. These are the people that have died in floods in the last three months with unprecedented rainfall.”

Pronounced calls for urgent action come from Paul Simpson, a Partner at ERM, who strongly voices the need for immediate measures to counteract the severe impacts of climate change globally.

A leading answer to these urgent calls is from Google. The tech giant has just announced a major initiative to tackle the intensifying challenges brought by climate change through a significant expansion of its AI-driven flood forecasting capabilities. by any crisis or any sort of extreme weather event," reflects Yossi with a note of optimism.

This expansion aims to widen coverage to over 100 nations, potentially benefiting around 700 million people worldwide—up significantly from the previous tally of 460 million.

Kate Brandt, Chief Sustainability Officer of Google, says that floods are a growing concern for people all over the world.

“Floods and storms were amongst the top natural disasters searched for in 2023, showcasing that the world is searching for this kind of information like never before,” she explains.

The scenario seemed even grimmer in 2024.

Google is now stepping up amidst widespread anxiety and scepticism, hoping to mitigate the effects of these increasingly frequent and severe weather occurrences. But what methods will they employ to effect this change?

Advancements in flood forecasting technology

Yossi Matias, Vice President & Head of Google Research, is best positioned to talk about the transformative potential of this technology. "Our mission is to use AI to make flood forecasting information globally accessible,” he says.

“This expansion will bring critical flood information to millions of people across the globe, as well as more data to researchers and partners." 

Improvements in modelling now promise enhanced accuracy up to seven days ahead, surpassing the previous model's five-day forecast but with better reliability. This leap in performance stems from groundbreaking research that integrates a greater volume of labelled data and a novel model architecture.

By tripling the number of locations used during training phases and incorporating a new weather forecasting model, Google significantly bolsters the robustness of its predictive capabilities.

This advancement arrives at a crucial and sensitive juncture, particularly after Spain's deadliest floods were partly attributed to inadequate early warning systems.

Widening access for researchers and partners

To further drive research and response strategies concerning flooding, Google plans to make its flood forecasting data accessible through an upcoming API and the Google Runoff Reanalysis & Reforecast (GRRR) dataset.

This provision will assist researchers and partners in sourcing hydrological forecasts and prospective flood data, especially valuable in regions with scant local data.

The integration of 'virtual gauges' in the system will enhance the availability of flood forecast data in over 150 countries, even extending to areas bereft of physical measuring instruments.

How the past can inform the future in flood forecasting

Intriguingly, Google is also making public its historical outputs from the global flood forecasting model, dating back to 1981. This dataset is set to help researchers recognise long-term flooding patterns and aid in mitigating future impacts.

Yossi says: "We believe that access to this data will help researchers around the world understand how flooding is affecting communities, and potentially reduce the impact of devastating floods."

The practical impacts of these advancements are evident. In May 2024, amidst severe flooding in Rio Grande do Sul, Brazil, Google collaborated with the Geological Service of Brazil to activate more than 200 new monitored locations on its Flood Hub platform.

This timely data enabled local authorities and residents to devise effective response strategies.

Early warnings, fewer disasters

With climate change fuelling an increase in extreme weather events, the role of advanced AI technologies in disaster preparedness is becoming ever more critical.

Google's expanded flood forecasting capabilities are tightly aligned with global initiatives such as the UN’s Early Warnings for All programme, which strives to provide timely safety information for crisis reactions worldwide.

For Yossi, these developments are cause for optimism. He says: "When you think about other natural disasters... ideally, I’d like to get us to the place that nobody is surprised by any crisis or any sort of extreme weather event."

Receive the next edition of ClimateTech Digital by signing up for its newsletter. 

As part of this portfolio, make sure you check out Sustainability Magazine and Energy Digital and also sign up to our global conference series - Sustainability LIVE. 

ClimateTech Digital is a BizClik brand