Google Researchers Use AI To Forecast Floods With Small Data Sets
Climate change is increasing global flooding, and artificial intelligence can help mitigate this damage by providing accurate warnings, even in flood basins lacking water gauges. Floods, the most common natural disaster, cause an average of $50 billion in global economic damages annually, with 90% of the 1.8 billion susceptible individuals in low- and middle-income countries.
An AI model now operational in 80 countries provides more accurate predictions of river flooding than the previous dominant system, according to a paper published in Nature. Researchers at Google developed the model, which improved forecasts in Africa to be similar to what is currently available in Europe.
Google AI modelers used diverse, publicly available data sources, such as weather forecasts, satellite imagery, topography, and soil type, and used AI to predict flood areas and water depth.
(With inputs from Shikha Singh)
You need to login in order to Like
