GraphCast
PulseAugur coverage of GraphCast — every cluster mentioning GraphCast across labs, papers, and developer communities, ranked by signal.
- developed by Google DeepMind 100%
- used by ERA5 90%
- used by European Centre for Medium-Range Weather Forecasts 90%
- used by Sparse Autoencoders 90%
- competes with Fuxi 80%
- developed by European Centre for Medium-Range Weather Forecasts 70%
- competes with aurora 70%
- competes with Hresk 70%
- other Pangu-Weather 60%
- used by aurora 60%
- affiliated with European Centre for Medium-Range Weather Forecasts 50%
- competes with Pangu-Weather 50%
- 2026-09-03 product_launch Google launched its new AI weather model, GraphCast, developed by Google DeepMind. source
- 2026-09-03 research_milestone Google's AI weather model, GraphCast, demonstrates improved forecasting capabilities. source
- 2026-06-02 research_milestone A new paper demonstrates machine learning models can achieve skillful deterministic weather forecasts beyond 30 days. source
6 day(s) with sentiment data
-
Machine learning weather models: distinguishing initial-condition and model error
A new research paper explores the distinction between initial-condition errors and model errors in machine learning weather models. The study evaluates gated and ungated optimizations in GraphCast and Pangu-Weather, mea…
-
ECMWF integrates AI weather forecasts alongside traditional models
The European Centre for Medium-Range Weather Forecasts (ECMWF) is now displaying machine learning-based weather forecasts from models like FourCastNet, Pangu-Weather, and GraphCast alongside their traditional physics-ba…
-
AI weather model GraphCast's internal cyclone data organization revealed
Researchers have explored how AI weather models internally represent complex phenomena like tropical cyclones. Using sparse autoencoders (SAEs) with the GraphCast model, they analyzed features related to cyclone positio…
-
GraphCast-Mamba enhances weather forecasts with Mamba correction branch · 2 sources tracked
Researchers have developed GraphCast-Mamba, a novel approach to weather forecasting that enhances existing models by incorporating a trainable Mamba correction branch. This new branch augments a pretrained GraphCast bac…
-
New probabilistic model KNOU-Flood enhances global streamflow forecasting
Researchers have developed KNOU-Flood, a novel probabilistic machine learning model designed to improve global streamflow forecasting. This model builds upon existing transfer-learning techniques by incorporating a long…
-
AI weather models improve power outage prediction for crew deployment
A new research paper explores the use of AI weather models, specifically GraphCast and FourCastNet, to predict power outages caused by severe storms. The study analyzed how the lead time of these forecasts impacts the e…
-
New framework improves precipitation forecasts using diffusion models
Researchers have developed a new framework called Dynamics-Precipitation Decoupling to improve daily precipitation forecasts. This plug-and-play method uses a separate generator to map existing weather model outputs to …
-
Google's GraphCast weather model adopted by European forecasters without job losses
Google's DeepMind has developed a machine learning model called GraphCast that can generate a ten-day global weather forecast in under a minute, significantly outperforming traditional physics-based models. The company …
-
New ScaleCast framework aligns global and regional models for kilometer-scale weather prediction
Researchers have developed ScaleCast, a new framework for kilometer-scale regional weather prediction that aligns global and regional weather models. This framework uses a Global-Regional Conversion module to align repr…
-
AI weather model GraphCast's internal variables revealed by new interpretability method
Researchers have developed a method to interpret the internal workings of AI weather models like GraphCast by training sparse autoencoders (SAEs). These SAEs were used to uncover learned concepts within GraphCast, focus…
-
AI weather models' internal concepts revealed by new interpretability methods
Researchers are using sparse autoencoders (SAEs) to investigate the internal workings of AI weather models like GraphCast, focusing on how they represent phenomena such as atmospheric rivers. By training SAEs on GraphCa…
-
Machine learning weather models show promise but need refinement for complex terrain
A new study evaluated the performance of three machine learning weather prediction (MLWP) models—FourCastNet3 (FCN3), GraphCast, and ECMWF High Resolution Forecast (HRES)—for wind speed forecasting in Northern Norway. T…
-
New method extracts weather model uncertainty without retraining
Researchers have developed a new method called stochastically perturbed weights (SPW) to extract uncertainty estimates from deterministic machine-learning weather models without retraining. This technique involves pertu…
-
Google DeepMind unveils more accurate AI weather model · 5 sources tracked
Google has enhanced its AI weather forecasting capabilities with a new model that leverages live satellite data for improved resolution. This advanced model, developed by Google DeepMind, is reportedly more accurate tha…
-
New AI Model TC-Next Enhances Tropical Cyclone Forecasting Accuracy
Researchers have developed TC-Next, a multimodal deep learning model for forecasting tropical cyclone track and intensity. The model leverages foundation model forecasts of atmospheric fields and satellite imagery. When…
-
VeinCast framework enhances weather forecasting with physics-guided graphs
Researchers have developed VeinCast, a novel framework for global medium-range weather forecasting that integrates physics-guided dynamic field graphs with graph-conditioned fusion. This approach combines predefined atm…
-
AI weather model adapted for Mars atmospheric prediction
Researchers have successfully adapted an AI weather model, GraphCast, originally trained on Earth's atmosphere, to predict conditions on Mars. By fine-tuning the model using data from the Mars Climate Database, they wer…
-
AI weather emulators show promise but struggle with extreme heat prediction accuracy
A new arXiv paper evaluates six deep learning weather emulators, including Pangu-Weather, FuXi, ArchesWeather, AIFS, GraphCast, and Aurora, against traditional physics-based models for predicting extreme heat events. Wh…
-
AI excels at physics prediction but lags in theory-building, study finds
AI is making significant strides in physics discovery, but current trends suggest a reverse progression compared to human scientific advancement. While early AI efforts focused on discovering explicit equations, recent …
-
AI weather models lag physics-based approaches for extreme events
AI weather models like GraphCast, PanguWeather, and Fuxi are showing promise in predicting daily weather conditions, surpassing traditional physics-based models in some instances. However, a recent study indicates these…