Weather forecasting at a tenth of the cost is possible with Google’s new AI technology

Despite the availability of several tools and resources, no one can predict weather with 100% accuracy. These tools and resources are continually upgraded in order to achieve the best results. This should come as no surprise that weather experts are considering AI for the development of large-scale forecasting models. On the other hand, over time, the cost of running these new resources may rise dramatically. Google has now intervened to aid with this issue by providing a solution powered by AI.

The tech giant Google explained in a blog post about its newly developed AI model. It is named Scalable Ensemble Envelope Diffusion Sampler (SEEDS). The main focus of this model is to develop weather forecasts at scale. When it comes to the difference between SEEDS and the current technologies, it is the cost. SEEDS can deliver results without the requirement of a supercomputer that contributes to lowering the cost all in an effort to forecast storms and other weather phenomena more precisely throughout time.

SEEDS can generate large quantities of forecasts. The potential for analysis will increase due to the availability of more data. Users should be able to provide a more accurate image of the climate and weather to expect with access to more climate projections. Google just provided a demonstration in the journal Science Advances demonstrating the full potential of SEEDS as it exists.

Besides this, Google introduced GraphCast back in 2023. It is a machine-learning model from Google’s DeepMind branch. It demonstrated the ability of AI to create accurate predictions based on historical data as compared to numerical weather predictions. It seems like there is no limit to the potential of AI. Maybe we could come across wide adoption of SEEDS in the future based on its performance.