Google continues to offer different alternatives to know the weather, as happened recently with the update of Google Maps. Now presented a new generative artificial intelligence model dedicated mainly to weather forecasting.
Weather forecasting based on highly trained artificial intelligence models and the use of supercomputers leads to excessive expenses and costs. This is where SEEDS comes into play. (Scalable Ensemble Envelope Diffusion Sampler), the new AI model presented by Google a few days ago.
SEEDS is capable of giving the same results as current forecasting systems using a supercomputer. Which will reduce the overall cost and resources required in order to provide extreme reliability when forecasting storms or other types of weather events.
Introducing SEEDS, our latest generative AI technology that advances medium-range weather forecasting. We can now generate ensemble forecasts more efficiently, helping us better predict rare and extreme weather events. 🌩️ #Weather forecast Learn more at https://t.co/wcVXjpS1dx pic.twitter.com/iLOEWVZelS
– Google AI (@GoogleAI) March 29, 2024
The arrival of SEEDS could generate forecasts on a larger scale, since this would give experts more data to work with, while increasing the potential for larger analyses. Which would result in a better representation of climate and weather.
In fact, Google’s comments indicate that the results of its new A.I. They are similar to the system currently used in the United States for weather forecasting, as happened with a demonstration of GraphCast, a DeepMind machine learning model.
Image from a Google SEEDS video demonstration
Free Android
The tech giant believes that advances in the efficiency of generative AI could enable for institutions dedicated to weather reporting to access investments in advanced forecasting the weather. That is, if SEEDS is able to take a few physics-based models and convert them into a multitude of weather distributions, the capital and resources saved by moving away from numerical weather forecasts would help fund a larger number of published forecasts . with more frequency.
By saving resources, as Google claims on its blog, these could also be used to create better physics-based models with greater precision in their results. For now, the SEEDS model will be part of MetNet-3 and GraphCast as well as other new weather-based technologies from Google.
