Google first described ways it is using AI to address climate change in 2018 and on Wednesday announced an expansion of its flood forecasting and wildfire tracing services.
A platform called FloodHub is now global, offering a map display of flood forecasts that shows when and where floods may occur to help people prepare. Recent breakthroughs in AI-based flood forecasting models made it possible, according to Yossi Matias, vice president of engineering and crisis response lead at Google, in a blog.
The company sent 115 million flood alerts to 23 million people over Search and Maps in 2021. The new expansion of flood forecasts adds 18 countries across Africa, Latin America and Southeast Asia. Such alerts go to any Android phone and any phone with the Google search app installed. Users must enable Google location services on their phones to receive the flood alerts.
Before an alert is sent, Google predicts where flooding is likely. The company relies on AI transfer learning processes where a flood forecast model uses data from different drainage basins not specific to where the model is predicting an imminent flood. In the past, Google relied heavily on data from water level gauges to make alerts. Now, alerts can be done a week in advance instead of just two days in advance.
Google also uses new AI models to detect wildfire boundaries based on satellite imagery, with the location shown in Search and Maps. Since July, Google has covered more than 30 wildfires in the US and Canada, informing people with 7 million views in Search and Maps. The wildfire detection feature is available now in the US, Canada, Mexico and parts of Australia.
Also, Google uses SOS Alerts in Search and Maps to show a predicted trajectory of a hurricane. Machine Learning is used to analyze satellite imagery after disasters to identify areas needing help. The tech was deployed in September when Hurricane Ian hit Florida to help GiveDirectly allocate aid.
Matias said the blog was written in conjunction with the upcoming COP27 in Egypt, an international conference on sustainability that occurs annually and is attended by governments that have signed onto major climate change agreements.
Kate Brandt, Google Chief Sustainability Officer, also penned a blog on overall sustainability efforts at Google.
Google is active in other areas using AI; in one example,Project Green Light, it is using the technology to optimize traffic lights to minimize congestion.
Another project called Mineral uses AI to reveal insights about soil and weather data using drone and satellite imagery to detect the health of crops.
“At COP27 we hope to generate more awareness and engage in productive discussions about how to use AI, innovations and shared data to help global communities address the changing climate,” Matias wrote.
RELATED: US House passes IRA with major green tech; Biden to sign soon