May 22, 2024
Google Unveils SEEDS AI for Accurate Weather Forecasts

Google Unveils SEEDS AI for Accurate Weather Forecasts

Google introduced the “Scalable Ensemble Envelope Diffusion Sampler” (SEEDS), an AI model designed to provide accurate weather forecasts on a large scale while being more cost-effective than traditional physics-based methods.

Source: Google AI

Revolutionizing Weather Forecasting

SEEDS operates similarly to popular large language models (LLMs) like ChatGPT and generative AI tools like Sora. It generates multiple weather scenarios, or ensembles, quickly and economically compared to conventional models.

In a paper published in Science Advances on March 29, Google’s team outlines SEEDS’ capabilities, highlighting its potential to improve weather prediction accuracy and save lives in the face of increasingly common extreme weather events due to climate change.

Challenges in Weather Prediction

Weather forecasting presents significant challenges due to numerous variables that can lead to devastating events like hurricanes and heat waves. Conventional physics-based models rely on collecting various measurements and generating an ensemble of predictions based on these variables.

However, accurately predicting extreme weather events is often beyond the capabilities of traditional methods. Deterministic or probabilistic forecast models introduce random variables to initial conditions, leading to higher error rates, especially for long-term forecasts.

SEEDS addresses these challenges by leveraging AI to produce prediction models from physical measurements collected by weather agencies. It focuses on the relationship between the potential energy unit per mass of Earth’s gravity field in the mid-troposphere and sea level pressure, two common forecasting measures.

Advantages of SEEDS

Traditional methods typically produce ensembles of only 10 to 50 predictions. In contrast, SEEDS can extrapolate up to 31 prediction ensembles based on just one or two “seeding forecasts,” significantly expanding the range of possible outcomes considered in weather forecasts.

The system’s effectiveness is demonstrated by modeling the 2022 European heatwave using historical weather data. Seven days before the heatwave, U.S. operational ensemble prediction data failed to indicate its occurrence. SEEDS, with its increased prediction ensembles, proved more capable of anticipating such events.

Additionally, SEEDS offers significant cost savings and computational efficiency compared to traditional methods. The computing costs associated with SEEDS’ calculations are described as “negligible,” with the AI system boasting a throughput of 256 ensembles for every three minutes of processing time in a sample Google Cloud architecture.

Photo by Johannes Plenio

Related posts

LinkedIn Empowers Recruiters with Innovative AI Features

Chloe Taylor

AI Firms Restrict, Not Ban Election ‘Deepfakes’

Henry Clarke

Microsoft Teams With U.S. Labor For Worker-Centric AI Development

Henry Clarke

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More

Please enter CoinGecko Free Api Key to get this plugin works.