Learning skillful medium-range global weather forecasting

By Remi Lam, Alvaro Sanchez-Gonzalez, Matthew Willson, Peter Wirnsberger, Meire Fortunato, Ferran Alet, Suman V. Ravuri, Timo Ewalds, Z. Eaton-Rosen, Weihua Hu and 8 more. Science, 2023

1,707 citations recorded by the source corpus.

Global medium-range weather forecasting is critical to decision-making across many social and economic domains. Traditional numerical weather prediction uses increased compute resources to improve forecast accuracy but does not directly use historical weather data to improve the underlying model. Here, we introduce GraphCast, a machine learning–based method trained directly from reanalysis data. It predicts hundreds of weather variables for the next 10 days at 0.25° resolution globally in under 1 minute. GraphCast significantly outperforms the most accurate operational deterministic systems on 90% of 1380 verification targets, and its forecasts support better severe event prediction, including tropical cyclone tracking, atmospheric rivers, and extreme temperatures. GraphCast is a key advance in accurate and efficient weather forecasting and helps realize the promise of machine learning for modeling complex dynamical systems. Editor’s summary The numerical models used to predict weather are large, complex, and computationally demanding and do not learn from past weather patterns. Lam et al. introduced a machine learning–based method that has been trained directly from reanalysis data of past atmospheric conditions. In this way, the authors were able to quickly predict hundreds of weather variables globally up to 10 days in advance and at high resolution. Their predictions were more accurate than those of traditional weather models in 90% of tested cases and displayed better severe event prediction for tropical cyclones, atmospheric rivers, and extreme temperatures. —H. Jesse Smith Machine learning leads to better, faster, and cheaper weather forecasting.

Subfields: Atmospheric Science, Global and Planetary Change

View Paper: https://www.semanticscholar.org/paper/33ed3d9a505131daf631ca5fa1307e3aafdd3343

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Atmospheric Science

1. Stephan Rasp
2. Stephan Hoyer
3. I. Pisso

In Global and Planetary Change

1. Stephan Rasp
2. Stephan Hoyer
3. I. Pisso

Citations in the corpus

• Probabilistic weather forecasting with machine learning (2024, neutral): “To account for the 6-h offsets between GenCast and ENS initializations, we estimate the position error of ENS at the same 06/18 UTC initializations as GenCast by averaging the two position errors on either side of that initialization with the same lead time 2 .”
• A foundation model for the Earth system (2025, neutral): “Recent advances in artificial intelligence (AI) have shown promise in improving both predictive performance and efficiency 2,3 , yet their potential remains underexplored in many Earth system domains.”
• The operational medium-range deterministic weather forecasting can be extended beyond a 10-day lead time (2023, neutral): “It is noteworthy that, while GraphCast_operational exhibits similar overall prediction skill of tropical cyclones to FengWu, FengWu demonstrates superior prediction accuracy speci fi cally for forecasts at the 4-day lead time.”
• AlphaEvolve: A coding agent for scientific and algorithmic discovery (2025, neutral): “Over the last decade, AI systems have been applied to a wide range of scientific disciplines and tasks, from protein structure prediction [46] to quantum physics [6, 84] to climate sciences [53].”
• WeatherBench 2: A Benchmark for the Next Generation of Data‐Driven Global Weather Models (2023, methodology): “Here, we evaluate a version of GraphCast that was trained on ERA5 data from 1979 to 2019 (incl.).”
• Neural general circulation models for weather and climate (2023, methodology): “Specifically, we diagnose precipitation minus evaporation (Supplementary Information section H.5) rather than directly predicting these as in machine-learning-based approaches 3 .”
• TRACE: Grounding Time Series in Context for Multimodal Embedding and Retrieval (2025, unknown): “Time-series data is prevalent across critical domains such as healthcare, weather, and energy [1, 2, 3, 4].”