Artificial intelligence (AI) is already well known for its wide range of applications, from personal use to complex business solutions. It is also known that AI forecasting has already been tested in the field of climate change. However, Google is further expanding the potential of this technology in the climate sector by introducing innovative AI forecasting tools.
Google DeepMind Launches New AI Forecasting Tool for Monitoring Meteorological Phenomena
On June 12, Google DeepMind released a statement announcing, together with Google Research, the launch of a public and interactive website called “Weather Lab.” The platform will showcase shared AI weather models and the latest experimental AI-based forecasts.
The goal of this initiative is to enhance the capabilities of weather agencies and emergency service experts in better predicting cyclone trajectories and intensity. It also aims to improve decision-making processes and ensure timely communication about related risks and the management of potential cyclone impacts.
Although this project is still experimental, the AI forecasting tool has already demonstrated high-quality performance. In early March of this year, “Weather Lab” successfully predicted the exact trajectories of cyclones “Hone” and “Garance” south of Madagascar and tracked the paths of “Jude” and “Ivone” in the Indian Ocean nearly seven days before their appearance.
The photo shows the working principle of the AI forecasting tool in predicting the above-mentioned cyclones. Image source: Google DeepMind.
Collaboration with External Institutions on AI Forecasting
Google states that the website is being developed in collaboration with the U.S. National Hurricane Center, which will help scientifically validate the developed method and its results. Overall, the experimental AI-based tropical cyclone model can predict the formation, trajectory, intensity, size, and shape of a cyclone by generating 50 possible scenarios up to 15 days in advance, significantly surpassing traditional forecasting methods.
The AI forecasting system is capable of modelling based on two types of data:
- Large reanalysis datasets, which reconstruct past weather across the globe from millions of observations.
- A specialized database containing essential information on nearly 5,000 observed cyclone trajectories, intensities, sizes, and wind radii over the past 45 years.
Google also openly states that it is eager to contribute to public sector initiatives after U.S. President Donald Trump reduced the staff and capacity of the National Weather Service responsible for federal climate and weather research.
“Weather Lab” Will Offer Several Types of AI Model Forecasts
The mentioned system will display both live and historical cyclone forecasts generated by different AI weather models. Some of these models will operate in real time: “WeatherNext Graph,” “WeatherNext Gen,” and the latest experimental cyclone model.
Another portion of the forecasts is historical – showcasing over two years of archived forecasts, which experts and other researchers can download and analyse.
Final Thoughts
For now, scientists warn that this AI forecasting tool presented by Google DeepMind is still a pilot project, so the information it provides should not be interpreted as 100% accurate scenarios, as errors may still occur, and other official meteorological sources should be relied upon. While it may not yet replace traditional tools, but rather assist them, it signals that AI will continue to improve and will increasingly contribute to the monitoring of extreme and hazardous climate phenomena.
If you are interested in this topic, we suggest you check our articles:
- AI Climate Solutions: Will We Finally Solve the Ongoing Battle Against Climate Change?
- AI Weather Forecasting: New Dawn of Meteorology?
- How are AI Solutions for Climate Change Making an Impact?
Sources: Google DeepMind, The Verge
