Google creates AI weather simulator that measures climate change

Google has developed an AI-enhanced weather simulator that can accurately predict warming from climate change.

A research team from the company published their findings in an article on Monday (22 July) sharing how the experimental NeuralGCM model looks at historical sea surface temperature and sea-ice concentration to determine extended climate simulations. It has also been trained on 2.8° and 1.4° deterministic models.

The simulator can do the bread and butter of a weather forecast too, as it’s been found to match or improve on existing predictions.

The hybrid model combines the strengths of traditional forecasts with machine learning and is believed to be the first of its kind to use this technology to make accurate predictions.

“It is also, to our knowledge, the first hybrid model that achieves comparable spatial bias to global cloud-resolving models, can simulate realistic tropical cyclone tracks and can run AMIP-like simulations with realistic historical temperature trends,” writes the researchers behind the project.

Google’s hybrid AI weather simulator could predict hurricanes in the future

For further climate projection, NeuralGCM will be reformulated to include other components like ocean and land and data from greenhouse gasses and aerosols.

While the current research is focused on generalizations, the team notes how the hybrid modeling approach of NeuralGCM has the “potential to transform simulation for a wide range of applications, such as materials discovery, protein folding and multiphysics engineering design.”

The model will be expanded on in the future as the researchers are working on a new feature that could generate year-ahead hurricane predictions. This would allow for greater preparation measures to be taken if extreme weather is known about with better accuracy.

In a conversation with Bloomberg UK, a professor of atmospheric sciences at Texas A&M University who wasn’t involved in the research, R.Saravanan called it “an important advance in atmospheric modeling and long-term weather prediction, but not necessarily a giant leap in climate prediction.”

Featured Image: Via Ideogram