Google's newest AI tool, Scalable Ensemble Envelope Diffusion Sampler (SEEDS), can reportedly help make weather probability forecasts cheaper and faster compared to current forecast technologies like supercomputers.
According to Google, its recently developed SEEDS tool aims to produce scale weather forecasts.
However, SEEDS differs from existing technologies in that it is less expensive and can produce equivalent results without requiring a system that is as sophisticated as a supercomputer.
In the long run, this can lower the operation's total cost while improving long-term storm and weather event prediction accuracy.
(Photo: Photo by NASA via Getty Images) ATLANTIC OCEAN - SEPTEMBER 10: In this NASA handout image taken by Astronaut Ricky Arnold, Hurricane Florence gains strength in the Atlantic Ocean as it moves west, seen from the International Space Station on September 10, 2018.
The high number of forecasts that SEEDS can produce increases the potential for more analysis. This can also provide meteorologists with more data to work with. With access to more climate projections, users should be able to provide a more accurate image of the climate and weather to expect.
Google claims that because generative AI can produce incredibly detailed images and movies, it is even more appropriate for weather forecasting. This feature is particularly helpful in producing ensemble forecasts that are in line with likely weather patterns, which eventually yield the greatest additional value for applications further down the line.
Related Article: Earth 2: NVIDIA Unveils AI-powered Earth Climate Digital Twin
Google's Weather Initiatives
Google's track record with machine learning models used to predictive assessments reportedly suggests a favorable reaction for weather experts, despite certain reservations regarding the incorporation of AI into weather apps.
The effectiveness of GraphCast, a related machine-learning model developed by the DeepMind unit, demonstrated how AI might leverage past data to forecast meteorological conditions over time.
Google's advancement of artificial intelligence algorithms used in weather prediction tools proves there is increasing potential for more accurate and reliable weather predictions. This will be the case particularly when taking into account upcoming projects utilizing SEEDS as we move into what meteorologists anticipate to be exciting times ahead.
AI-Powered Climate Projects
AI, in general, continues to be used in climate action as it was also recently proven able to predict optimum solar panel spots. The AI algorithm that can anticipate and pinpoint the best locations for double-sided solar panels to be installed to maximize their solar energy output was created by a team of Chinese scientists.
Using the model, the scientists have already determined that the best places to maximize solar energy output are in northwest China, specifically in the eastern Tibetan Plateau.
To overcome the lack of field data, researchers from Tsinghua University in Beijing and the National Tibetan Plateau Data Centre created an artificial intelligence model based on sunshine data from 2,453 meteorological stations across China.
To assess both diffuse and direct sun radiation, the researchers used data augmentation with the LightGBM machine learning model. This strategy evades the traditional problems associated with scarce and unevenly distributed ground-based observations by creatively using sunshine duration data gathered from over 2,453 meteorological stations.
(Photo: Tech Times)
