Google’s artificial intelligence beats supercomputers at fast, accurate weather forecasts

Advertisement · Scroll to continue

NEW YORK, BRONX (ORDO News) — Weather forecasting has long been a challenging task for scientists due to its complex nature and countless variables. But now Google has revolutionized this field by developing an artificial intelligence system that can accurately predict the weather 10 days ahead in just one minute on a single machine.

This breakthrough has the potential to change the way we plan our lives and prepare for severe weather events.

The famous butterfly effect shows that even the smallest actions can have a significant impact on the weather. The task of weather forecasters is to take all these factors into account and create accurate models to predict future weather conditions.

This process, known as numerical weather forecasting (NWP), involves taking current weather data and processing it using complex physics equations on supercomputers.

Google’s GraphCast artificial intelligence system was trained on 40 years of weather reanalysis data collected from various sources such as satellite imagery, radar and weather stations. The system receives the current weather and, based on an extensive array of data, makes a forecast for the next six hours with a lead time of up to 10 days.

The model covers more than a million grid points around the Earth‘s surface and takes into account different variables at different heights.

The tests compared GraphCast to the existing weather forecasting standard, the HRES supercomputer-based modeling system.

The AI ​​system was able to make a more accurate forecast in less time, outperforming HRES on 90% of test variables and forecast latency. In the troposphere, the layer of the atmosphere closest to the Earth’s surface, GraphCast outperformed HRES by 99.7%.

One of GraphCast’s most impressive achievements is its ability to detect severe weather events earlier than traditional forecasting methods. In a real-life example, AI accurately predicted the location of a hurricane’s landfall in nine days, while traditional forecasts could only confirm it in six days.

Google has made GraphCast open source, allowing scientists around the world to experiment with it and implement it into their weather forecasting systems. Through this collaboration, the accuracy and efficiency of weather forecasting is expected to improve significantly.

Google’s AI technology has made weather forecasting faster, more accurate, and more accessible. This achievement has the potential to save lives and prevent damage from severe weather events. As for the future, who knows what other complex problems AI will be able to solve in just one minute on a single computer.


News agencies contributed to this report, edited and published by ORDO News editors.

Contact us: [email protected]

Our Standards, Terms of Use: Standard Terms And Conditions.

To eliminate any confusion arising from different time zones and daylight saving changes, all times displayed on our platforms are in Coordinated Universal Time (UTC).

Advertisement · Scroll to continue
Advertisement · Scroll to continue