Google AI Forecast Floods Up to 7 Days in Advance

While Google’s generative AI tools like Circle to Search and Search Generative Experience (SGE) have garnered significant attention, the tech giant is also utilizing AI in lesser-known areas such as flood prediction. Recently, Google leveraged machine learning (ML) to forecast floods with remarkable accuracy, providing warnings for up to seven days.

Google unveiled a substantial enhancement in global-scale flood forecasting through the integration of machine learning technologies. The company reported a notable improvement in the reliability of “global nowcasts,” extending from zero to five days on average. Remarkably, in certain instances, Google achieved forecasts predicting floods up to a week in advance.

Predicting floods in advance poses challenges due to the absence of streamflow gauges in many rivers, which are crucial for collecting precipitation and watershed data. However, Google overcame this hurdle by leveraging its ML technology, which processed all accessible river data.

84% of Google Search Insights Are Driven by the Power of Generative AI

Additionally, the ML model was applied to basins lacking data, enabling more comprehensive flood forecasting. The significance of Google’s achievement is underscored by the publication of its findings in Nature, a prestigious multidisciplinary science journal known for its peer-reviewed research.

Google’s ultimate objective with this technology is to enhance the accuracy of flood forecasting on a global scale, even in regions where local data may be lacking. Through its Flood Hub, Google has already extended forecasts to over 80 countries. Moreover, it disseminates alerts via Google Search, Maps, and Android notifications, ensuring that vital information reaches a broad audience efficiently.