[ad_1]
Floods are becoming more frequent and severe as the climate crisis intensifies, and are likely to get worse. That can make the task of predicting where and when these disasters will occur challenging, especially in areas with little historical or real-time data from which to draw.
TO new paper of Google researchers posit that machine learning might hold some answers. The tech giant claims that an artificial intelligence model, which is based on publicly available data, can predict river flooding up to five days in advance with “similar or better” reliability than current systems.
Google Vice President of Engineering and Research Yossi Matias said that what is particularly notable about this model is that it is widely applicable on a global scale, which could provide more accurate forecasts for developing areas that do not have current meters or water measuring stations along the rivers.
“The real question was, how can we build a model that won’t be built for one location, but is a global model, where we can take any data (about what) is happening anywhere in the world and put it into the prediction? modeling,” Matías told Tech Brew. “The big breakthrough is really, first, showing that we can really take a global model and do it and, second, measuring it systematically.”
The findings come as Google has been expanding its maps and search tools to provide consumers and developers with more data on climate-related phenomena such as wildfires, air quality and flooding.
“This is a business-to-business effort,” Matías said. “In everything we do, we’re not just looking to build models, but also to see how they really impact people through our platforms and other platforms.”
Google first piloted their flood forecasting work in 2018 in the Indian city of Patna, located in the flood-prone state of Bihar, Matías said. From there, the program gradually expanded throughout the country and into Bangladesh. But the location-specific nature of the model initially made it difficult to expand the approach to areas without flow data, according to one study. blog post written by Matías.
Stay up to date with innovative technology that transforms business
Tech Brew keeps business leaders up to date on the latest innovations, advancements in automation, policy changes, and more, so they can make informed technology decisions.
That is where the model with a global focus has been useful, Matías stated. The company has already implemented it on a platform called Flood Hub that now offers alerts to 80 countries and covers around 460 million people, according to Matías. It can also help with coverage, as the climate crisis creates flood risks in areas that historically have not been prone to flooding, Matias said.
“It turns out that when we started thinking about global models, we were in a situation where we asked ourselves, ‘How can we predict flooding in places that don’t have much history?’” Matías said. “What happened was that because of these new floods that we see in places that never actually experienced flooding, it turns out that this model is even more important than we anticipated.”
Currently, Google’s alert system focuses specifically on river floods, which according to Matías are the “most common, in terms of impact on society.” But more research is still needed to understand and predict other types, such as flash, coastal and urban flooding, he said.
“When you think about other natural disasters, again, with the progress of technology and the ability to solve difficult problems, the ideal would be to get us to the point where no one is surprised by any crisis or any type of (weather) extreme, simply because “We’re going to have enough technology that will be able to give predictions and alerts in due time… I’m very optimistic that it’s a matter of time before we can do that.”