US Edition
Study Finds AI Weather Models Use 21 Times Less Energy Than Traditional Systems
A study found AI weather models use at least 21 times less energy annually than traditional models, reducing the carbon footprint of forecasting, researchers said.
- Researchers publishing in the journal Weather reported on March 10, 2026 that AI data-driven models are estimated to consume at least 21 times less energy than traditional forecasting models despite substantial training energy use.
- To test practical needs, researchers chose meteorology because it combines large, time-varying datasets and a need to reason in plain language, aiming to build AI agents for natural-language weather queries and broader climate gains.
- In experiments, researchers found Zephyrus translates language into code to query models and performs well on simple forecasts but struggles with extreme-weather detection and report generation.
- The study's authors say the results open the door to significantly reduce forecasting emissions and urge future studies so developers of future weather models set energy reduction targets alongside performance goals, while Wiley could help disseminate these findings.
- Researchers argue this matters for sectors including agriculture, disaster preparedness, transportation, and energy management, aiming to democratize earth science and lower entry barriers, as Rose Yu and Duncan Watson‑Parris emphasized.
12 Articles
12 Articles
AI agent could transform how scientists study weather and climate
Computer scientists and weather scientists have taken the first steps toward creating an AI agent capable of analyzing and answering questions in natural language, such as English, about data from AI-driven weather and climate forecasting models. The research team from the University of California San Diego will present the first AI weather agent they developed, named Zephyrus, at the 14th International Conference on Learning Representations (IC…
Energy and carbon footprint considerations for data-driven weather forecasting models
Data-driven models for weather forecasting display impressive computational speed-up. However, these gains do not include the training phase of the models. This study gathers information from the literature about training and inference to estimate their energy and carbon footprints. Despite being considerably more costly than a physics-based forecast, the training is rapidly compensated by the savings made during inferences. For a use-case corres
AI Agent Zephyrus Aims to Revolutionize Weather and Climate Analysis - San Diego Today
Computer scientists and weather researchers at the University of California San Diego have created an AI agent named Zephyrus that can analyze weather and climate data, and answer questions about it in natural language. The goal is to make it easier for students, scientists, and the public to access and understand critical weather and climate information.
Coverage Details
Bias Distribution
- 100% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
