C. Scott Brown / Android Authority
long story short
- Google has been using artificial intelligence to improve flood forecasts around the world.
- By training a machine learning model, the tech giant was able to accurately predict some floods up to seven days in advance.
- Google already uses its technology to provide accurate flood forecasts in 80 countries.
You’ve probably heard a lot about Google’s various generative AI products, such as Circle to Search, its AI wallpaper tool, Search Generative Experience (SGE), and more. But you might not have heard of other ways it uses artificial intelligence, such as predicting floods. Recently, with the help of machine learning (ML), Google was able to accurately predict floods seven days in advance.
Today, Google announced that it can significantly improve flood forecasting around the world with the help of machine learning technology. According to the company, it is able to increase “the reliability of instantaneous forecasts globally from zero to five days on average.” In some cases, it is even able to predict floods a full week before they occur.
One reason it’s difficult to predict floods in advance is that most rivers don’t have what’s called a flow gauge. These gauges help provide relevant data such as precipitation and physical watershed information. However, the tech giant was able to solve this problem by applying its machine learning technology to all available river data and applying machine learning models to watersheds where no data was available.
The company’s findings were published in nitrogennature. For context, nature is a leading multidisciplinary scientific journal publishing peer-reviewed research.
The ultimate goal of this technology is to increase the accuracy of flood forecasts to a global level, even in areas where local data is not available. Google is already able to provide forecasts to more than 80 countries through its Flood Center. It also provides alerts for Google Search, Maps, and Android notifications.