Learning skillful medium-range global weather forecasting
Open accessThe DOI matches its Crossref record; 65 of the 66 references it cites were found in a registry or an index, and 1 are not checked yet; 12 of the 38 sources its claims cite were checked for support.
Abstract
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.
Verified 2026-11-22 · 1 reference not checked yet
Overview
CoChat Summary
TL;DR
GraphCast is a graph neural network trained directly on ERA5 reanalysis that produces 10-day global weather forecasts at 0.25° resolution in under one minute. It outperforms the operational HRES deterministic system on most evaluated targets and supports better prediction of tropical cyclones, atmospheric rivers, and extreme temperatures.
Key findings
- GraphCast significantly outperformed HRES on 90.3% of 1380 verification targets, and significantly on 89.9% of targets; excluding 50 hPa it significantly outperformed HRES on 96.9% of 1280 targets, and excluding 50 and 100 hPa on 99.7% of 1180 targets.
- On z500, GraphCast had better skill across all lead times, with skill score improvement around 7%–14%; it also outperformed Pangu-Weather on 99.2% of the 252 targets Pangu-Weather presented.
- For severe events, GraphCast had lower median cyclone track error and was significantly better than HRES for lead times 18 hours to 4.75 days; it improved atmospheric river ivt prediction from 25% at short lead time to 10% at longer horizons; extreme heat precision-recall curves were above HRES at 5- and 10-day lead times, while HRES was better at 12 hours.
- Training on more recent data improved GraphCast’s skill scores, with variants trained up to before 2021 generally outperforming earlier-trained variants on 2021 test data.
Methods
- GraphCast is a 36.7-million-parameter graph neural network in an encode-process-decode configuration, using a multi-mesh icosahedral representation with 40,962 nodes and 327,660 mesh edges; it takes two most recent weather states and predicts the next 6-hour state autoregressively.
- It was trained on 39 years (1979–2017) of ERA5 reanalysis, using a weighted MSE objective over 12 autoregressive steps, with training taking roughly four weeks on 32 Cloud TPU v4 devices; evaluation used held-out 2018 data, with some analyses on 2018–2021 and 2021.
- Forecast skill was evaluated against HRES using RMSE and ACC, with HRES-fc0 as HRES ground truth and ERA5 as GraphCast ground truth; comparisons used 06z/18z initializations and 12-hour target intervals to match data assimilation lookahead.
Limitations
- The paper states that GraphCast is deterministic and does not explicitly model uncertainty like ECMWF’s ensemble system ENS; its MSE training objective encourages spatial blurring, which may limit value for some applications.
- The paper states GraphCast operates at 0.25° resolution, 37 vertical levels, and 6-hour time steps, versus HRES at 0.1°, 137 levels, and up to 1-hour time steps, and that it depends critically on large quantities of high-quality data assimilated via NWP, so it should not be regarded as a replacement for traditional weather forecasting.
- The paper reports reduced skill in the stratosphere, especially at 50 hPa, and at high surface elevations; it also notes possible ERA5 versus HRES-fc0 domain differences and spurious spectral artifacts around 100 km wavelengths.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures



























Verification
Claims the paper makes
9 Valid · 0 Not valid · 18 Not checked · 150 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Generally, verifying a model against its own analysis gives the best skill estimates [45].”
§ 5. Verification methods > 5.2. Comparing GraphCast to HRES
What the sources say
- Partly supportedcites [45] [45] Richard Swinbank, Masayuki Kyouda, Piers Buchanan, Lizzie Froude, Thomas M Hamill, Tim D Hewson, Julia H Keller, Mio Matsueda, John Methven, Florian Pappenberger, et al. The TIGGE project and its achievements. Bulletin of the American Meteorological Society, 97(1):49–67, 2016. sourceExperimental — accuracy not yet measured
The Verification section reports an example in low-level tropical regions where the model verified against its own analysis appeared most accurate. It does not establish this as a general rule; instead, it warns that skill rankings depend strongly on the analysis used and recommends verifying against more than one analysis. (LLM, conf=0.86)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“For the HRES baseline, we used the TIGGE archive, which provides cyclone tracks estimated with the operational tracker, from HRES's forecasts at 0.1° resolution [8, 46].”
§ 1. Datasets > 1.3. Tropical cyclone datasets
What the sources say
- Partly supportedcites [8] [8] Philippe Bougeault, Zoltan Toth, Craig Bishop, Barbara Brown, David Burridge, De Hui Chen, Beth Ebert, Manuel Fuentes, Thomas M Hamill, Ken Mylne, et al. The THORPEX interactive grand global ensemble. Bulletin of the American Meteorological Society, 91(8):1059–1072, 2010. sourceExperimental — accuracy not yet measured
The Applications of TIGGE section says TIGGE providers supplied tropical cyclone tracks and that the archive centers would distribute them. It does not establish that those tracks came from HRES forecasts at 0.1° resolution or were estimated with the operational tracker; the examples instead describe ECMWF and UKMO ensemble tracks. (LLM, conf=0.85)
- Partly supportedcites [46] [46] Richard Swinbank, Masayuki Kyouda, Piers Buchanan, Lizzie Froude, Thomas M. Hamill, Tim D. Hewson, Julia H. Keller, Mio Matsueda, John Methven, Florian Pappenberger, Michael Scheuerer, Helen A. Titley, Laurence Wilson, and Munehiko Yamaguchi. The TIGGE project and its achievements. Bulletin of the American Meteorological Society, 97(1):49 – 67, 2016. sourceExperimental — accuracy not yet measured
The TIGGE Project and Its Achievements describes TIGGE as an archive of ensemble forecasts, and its Dynamics and Predictability section reports cyclone tracks derived from TIGGE data using a tracking method employed operationally by ECMWF. It does not establish that the archive provides tracks from HRES forecasts at 0.1° resolution; the tracking passage concerns global ensembles, not HRES. (LLM, conf=0.85)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Because extreme prediction is by definition an imbalanced classification problem, we base our analysis on precision-recall plots which are well-suited for this case [42].”
§ 1 2 3 4 5 6 7 8 9 10 > 8.3. Extreme heat and cold
What the sources say
- Supportedcites [42] [42] Takaya Saito and Marc Rehmsmeier. The precision-recall plot is more informative than the ROC plot when evaluating binary classifiers on imbalanced datasets. PloS one, 10(3):e0118432, 2015. 101 https://confluence.ecmwf.int/display/FCST/202208+-+Heatwave+-+UK https://confluence.ecmwf.int/display/FCST/202208+-+Heatwave+-+UK https://www.ecmwf.int/node/17262 https://www.ecmwf.int/node/17262 GraphCast: Learning skillful medium-range global weather forecasting source
The Summary of the re-evaluation says precision-recall plots show classifiers’ susceptibility to imbalanced datasets with clear visual cues. This supports using them to evaluate an imbalanced classification problem; the citing authors’ characterization of extreme prediction is their own framing. (LLM, conf=0.93)
boostedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“However, the HRES cyclone tracks in the TIGGE archive [8] are only initialized at 00z and 12z.”
§ 8. Additional severe event forecasting results > 8.1. Tropical cyclone track forecasting
What the sources say
- Partly supportedcites [8] [8] Philippe Bougeault, Zoltan Toth, Craig Bishop, Barbara Brown, David Burridge, De Hui Chen, Beth Ebert, Manuel Fuentes, Thomas M Hamill, Ken Mylne, et al. The THORPEX interactive grand global ensemble. Bulletin of the American Meteorological Society, 91(8):1059–1072, 2010. sourceExperimental — accuracy not yet measured
The Applications of TIGGE section says tropical cyclone tracks were to be distributed by TIGGE archive centers. Table 3 lists two daily forecast cycles for ECMWF’s ensemble, but the paper does not identify HRES cyclone tracks or state that their only initialization times are 00z and 12z. (LLM, conf=0.79)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“we can derive this quantity from the predicted atmospheric variables specific humidity, q, and horizontal wind, (u, v), via the relation [38]:”
§ 1 2 3 4 5 > 8.2. Atmospheric rivers
What the sources say
- Supportedcites [38] [38] Paul J Neiman, F Martin Ralph, Gary A Wick, Jessica D Lundquist, and Michael D Dettinger. Meteorological characteristics and overland precipitation impacts of atmospheric rivers affecting the West Coast of North America based on eight years of ssm/i satellite observations. Journal of Hydrometeorology, 9(1):22–47, 2008. source
In section c, “The link between SSM/I IWV plumes and atmospheric rivers” (page not provided), the paper defines IVT as vertically integrated horizontal water-vapor flux calculated from q and the horizontal wind components. It describes summing the flux components through pressure layers, which supports the cited relation; the paper uses reanalysis variables rather than predictions. (LLM, conf=0.92)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
65 more claims
Create a free account to see all 177 claims and to verify the 16 references still open.
- Probably not a substantial claim. 54
- It may state two separate claims as one. 70
- The quote only partly states this claim. 2
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Existing machine-learning models have focused on deterministic prediction, and surpass deterministic numerical weather prediction in terms of the aggregate metrics for which they are trained3,4.”
From Neural general circulation models for weather and climate (2024) · cites this paper as [3]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Most state-of-the-art machine learning models utilized in medium-range weather forecasting are built upon encoder-decoder84 architectures27–29,85.”
From A machine learning model that outperforms conventional global subseasonal forecast models (2024) · cites this paper as [27]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“However, these advances have focused primarily on single, deterministic forecasts that fail to represent uncertainty and estimate risk.”
From Probabilistic weather forecasting with machine learning (2025) · cites this paper as [2]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“They are typically trained to minimize the mean squared error (MSE) of their predictions and as a result tend to produce blurry forecasts, especially at longer lead times, rather than a specific realization of a possible weather state2.”
From Probabilistic weather forecasting with machine learning (2025) · cites this paper as [2]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Even so, these techniques are still fundamentally reliant on precise meteorological observations [1-3].”
From Advancing Data Quality Assurance with Machine Learning: A Case Study on Wind Vane Stalling Detection (2025) · cites this paper as [3]
10 more claims about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ 7. Additional forecast verification results > 7.1. Detailed results for additional variables
“The ACC skill score is the normalized ACC difference between model 𝐴 and baseline 𝐵 as (ACC𝐴 − ACC𝐵)/(1 − RMSE𝐵).”
§ 1 2 3 4 > 5.3. Evaluation metrics
“and the normalized ACC difference as (ACC𝐴 − ACC𝐵)/(1 − ACC𝐵).”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ 1 2 3 4 5 6 7
“The x-axis represents lead time, at 6-hour steps over 10 days.”
§ 6. Comparison with previous machine learning baselines
“Pangu-Weather[7] reports its 7-day forecast accuracy (RMSE and ACC) on: z500, t500, t850,”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Conclusions
“because of the ERA5 training data's native 0.25° resolution”
§ 102 103 104
“the native resolution of ERA5 is 0.28125◦ corresponding to a shortest wavelength of 62km”
Citations
Cites
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
Exploiting Generative AI to Scale up Intelligent Tutoring Systems
Jakubův, J., Chvalovský, K., Goertzel, Z., Kaliszyk, C., Olšák, M., Piotrowski, B., Schulz, S., Suda, M., & Urban, J. (2023) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.itp.2023.19 · Cited by 76,658
Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi, A., Muñoz‐Sabater, J., Nicolas, J., Peubey, C., Radu, R., Schepers, D., et al. (2020) · Quarterly Journal of the Royal Meteorological Society · doi:10.1002/qj.3803 · Cited by 31,882
Decoupled Weight Decay Regularization
Loshchilov, I., & Hutter, F. (2017) · arXiv · doi:10.48550/arxiv.1711.05101 · Cited by 12,522
Veličković, P., Cucurull, G., Casanova, A., Romero, A., Liò, P., & Bengio, Y. (2017) · arXiv · doi:10.48550/arxiv.1710.10903 · Cited by 12,107
Saito, T., & Rehmsmeier, M. (2015) · PLOS ONE · doi:10.1371/journal.pone.0118432 · Cited by 5,309
SGDR: Stochastic Gradient Descent with Warm Restarts
Loshchilov, I., & Hutter, F. (2016) · arXiv · doi:10.48550/arxiv.1608.03983 · Cited by 5,309
Backpropagation through time: what it does and how to do it
Werbos, P. J. (1990) · Proceedings of the IEEE · doi:10.1109/5.58337 · Cited by 4,879
Knapp, K. R., Kruk, M. C., Levinson, D. H., Diamond, H. J., & Neumann, C. J. (2010) · Bulletin of the American Meteorological Society · doi:10.1175/2009bams2755.1 · Cited by 3,237
Research works printed but not linked: 3
- [46] Richard Swinbank, Masayuki Kyouda, Piers Buchanan, Lizzie Froude, Thomas M. Hamill, Tim D. Hewson, Julia H. Keller, Mio Matsueda, John Methven, Florian Pappenberger, Michael Scheuerer, Helen A. Titley, Laurence Wilson, and Munehiko Yamaguchi. The TIGGE project and its achievements. Bulletin of the American Meteorological Society, 97(1):49 – 67, 2016.
- [31] David H Levinson, Howard J Diamond, Kenneth R Knapp, Michael C Kruk, and Ethan J Gibney. Toward a homogenous global tropical cyclone best-track dataset. Bulletin of the American Meteorological Society, 91(3):377–380, 2010.
- ECMWF IFS Documentation CY45R1 – Part II: Data assimilation (European Centre for Medium-Range Weather Forecasts 2018); .10.21957/a3ri44ig4 · doi:10.21957/a3ri44ig4
Cited by
Items 2,001–2,100 of 2,424
Threshold-Consistent Ordinal Evidence for Probabilistic Multi-Step Ocean Wind–Wave Risk WarningPreprint
Ji, S., Lyu, H., Law, S., Yin, Y., & Gao, X. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6996457
Event-Aware Loss Design for Forecasting of Convective Precipitation and LightningPreprint
Lee, C., Yang, H., & Kim, B. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2606.25937
Machine learning is revolutionizing weather forecasting - the next step is a change in how we work
Dueben, P. D., Bauer, P., Fuhrer, O., Koldunov, N., & Kristiansen, J. (2026) · Journal of the European Meteorological Society. · doi:10.1016/j.jemets.2026.100050
Does Aurora Encode Atmospheric Structure? Latent Regime Analysis and AttributionPreprint
Kasteleyn, E., & Lučić, A. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2606.26361
Yang, Y., Xie, Q., Hu, X., Pant, R., & Huang, J. (2026) · Journal of Industrial Ecology · doi:10.1007/s44498-026-00107-y
Combining Domain and Tensor Parallelism to Train Multi-billion-parameter AI Weather Models
To, D., Götz, M., Heyen, L. H., Streit, A., & Debus, C. (2026) · Platform for Advanced Scientific Computing Conference (PASC) · doi:10.1145/3815572.3815765 · Cited by 1
Predicting Atomistic Transitions with Transformers
Tischler, H., Li, W., Tang, Q., Pérez, D., & Vogel, T. (2026) · Data Science in Science · doi:10.1080/26941899.2026.2685344
Mahajan, S., Kelleher, M. E., & Fan, M. Y. (2026) · Platform for Advanced Scientific Computing Conference (PASC) · doi:10.1145/3815572.3815761
“This behavior is qualitatively consistent with concerns that deterministic AI forecast systems, when naively ensemble-ized, tend to produce overly tight ensembles unless additional mechanisms for sampling model or initial-condition uncertainty are introduced [5, 19, 36].”
Sampling Parallelism for Fast and Efficient Bayesian Learning
Özdemir, A. K., Heyen, L. H., Weyrauch, A., Streit, A., Götz, M., & Debus, C. (2026) · Platform for Advanced Scientific Computing Conference (PASC) · doi:10.1145/3815572.3815745 · Cited by 1
Geng, Y., Zhao, N., & Du, J. (2026) · Swarm and Evolutionary Computation · doi:10.1016/j.swevo.2026.102451
Ahmed, S. M. S., & Güneyli, H. (2026) · Intelligent Climate and Eco-Environment · doi:10.1016/j.icee.2026.100009
Aprendizaje híbrido de datos y física para pronósticos atmosféricos usando PIML
Lopretto, F. (2025) · Repositorio Digital Institucional de la Universidad de Buenos Aires (Universidad de Buenos Aires)
Latent Attention on Masked Patches for Flow Reconstruction
Eze, B., Magri, L., & Nóvoa, A. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-29915-4_15
“Reconstructing high-dimensional fields from sparse, noisy, and masked measurements is a central challenge in scientific machine learning and fluid mechanics [8, 19, 10].”
Comparative Study of Linear Attention Architectures for Geopotential Height Forecasting
Sastry, V. K., & Rao, V. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-29915-4_13
“Recent advances in deep learning have demonstrated that data-driven approaches can achieve competitive forecasting skill while significantly reducing computational costs at inference time [15,2,9].”
Xie, L. (2026) · Climate · doi:10.3390/cli14070137
“Recent progress in data-driven weather prediction, including models such as FourCast-Net and GraphCast, demonstrates that deep neural networks can learn complex mappings from gridded atmospheric and oceanic states to future weather fields [25,26].”
Accurate and Realistic Nowcasting Geo-Kompsat-2A Level-1 Infrared Observation With SimVP-GAN
Kim, Y., Jeong, J., & Hong, S. (2026) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2026.3708145
“Furthermore, whilerecent large-scaleDL forecastingsystems,suchasGraphCast[45]andPangu–Weather [46],havedemonstratedstrongskill,theyareprimarilytrained andevaluatedonreconstructedNWPdatasetsratherthanon real-time,rawsatelliteradianceobservations.”
Li, N., Zhang, J., Zhang, J., Yang, H., Zhao, B., & Li, S. (2026) · Atmosphere · doi:10.3390/atmos17070644
Adaptive hierarchical representation of marine environmental variables based on mixture of experts
Guo, B., Guo, Y., You, Q., & Tao, W. (2026) · Ocean Engineering · doi:10.1016/j.oceaneng.2026.126454
Km-scale solar irradiance ensemble forecasting with generative deep learning
Liu, Z., Pan, B., Chao, J., Xu, J., Du, H., Wang, J., Li, W., Gao, S., Nai, C., Cheng, Y., et al. (2026) · npj Artificial Intelligence · doi:10.1038/s44387-026-00133-y
“Data-driven weather forecasting models, such as Pangu-Weather 25 and GraphCast 26 , reduce computational cost while maintaining prediction accuracy 27 .”
Explainable deep learning methods for medium- and long-term storm surge forecast
Su, C., Sahoo, B., Mao, M., & Xia, M. (2026) · Ocean Modelling · doi:10.1016/j.ocemod.2026.102789
Exact waveguide backbones for sparse precursors to tropical-cyclone rapid intensification
Li, Z. (2026) · Chaos Solitons & Fractals · doi:10.1016/j.chaos.2026.118721
ResistNet: Quantifying solar energy reduction during hurricanes with deep learning
Zeng, H., Xu, L., Lin, N., & Poor, H. V. (2026) · Applied Energy · doi:10.1016/j.apenergy.2026.128287
AI-Driven Framework for Early Weather Prediction and Disaster Detection Using Deep Learning
Kumar, K. S., Siddhanth, & Vishal (2026) · International Conference on Connected Intelligence for Industrial Applications (CI2A) · doi:10.1109/ci2a69097.2026.11576788
“Lam et al. (2023) [4] have illustrated the promise of these newer architectures in continuing the evolution of deep learning application in meteorology.”
Mori, R., Harada, K., Mizuno, S., Onishi, T., Horiguchi, I., Sakamoto, K., Horibe, K., Kojima, T., Matsuo, Y., & Iwasawa, Y. (2026) · Transactions of the Japanese Society for Artificial Intelligence · doi:10.1527/tjsai.41-4_an40-g
Time-parallel incremental 4D-Var for two-dimensional rotating shallow water system
Bhatt, R. R., Debreu, L., & Vidard, A. (2026) · Journal of Computational Physics · doi:10.1016/j.jcp.2026.115164
Optimizing GPU-Accelerated Data Processing Frameworks for Scalable Deep Learning
Al-Mehdhar, H. (2025) · Research Publication Repository of King Fahd University of Petroleum and Minerals (King Fahd University of Petroleum and Minerals)
Statistical Downscaling of Daily Temperature over MexicoPreprint
Ramírez-Gómez, G. A., Romero-Padilla, J. M., Ramírez-Guzmán, M. E., Suárez-Espinosa, J., Macedo-Cruz, A., & Solorio, C. A. O. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7051904
TropDS: downscaling to enhance tropospheric delay grid precision in space geodesy
Ding, J., Chen, W., Chen, J., Wang, J., & Zhang, Y. (2026) · Journal of Geodesy · doi:10.1007/s00190-026-02086-1
“Finally, we validate TropDS on VMF3 forecast data and three weather models: Pangu-Weather (Bi et al. 2023), GraphCast (Lam et al. 2023), and FengWu (Chen et al. 2025).”
InfluentialSadak, M. S. (2026) · International Journal of Data Science and Analytics · doi:10.1007/s41060-026-01189-4 · Cited by 1
Zhang, P., Xiang, D., Huo, C., Ding, K., & Xiang, S. (2026) · Remote Sensing · doi:10.3390/rs18132202
Liu, J., Zhou, T., Li, J., Zhong, X., Zhang, P., Feng, J., Chen, L., & Li, H. (2026) · Journal of Geophysical Research Atmospheres · doi:10.1029/2025jd046208
“Models such as FourCastNet (Pathak et al., 2022), Pangu‐Weather (Bi et al., 2022), GraphCast (Lam et al., 2023), FengWu (Chen et al., 2025), and FuXi (Chen et al., 2023) achieve deterministic forecasts with remarkable computational efficiency and skill.”
Predictability of Polar-Low Environments Using Deep Learning Sensitivity Analysis
Lloveras, D. J., & Doyle, J. D. (2026) · Journal of the Atmospheric Sciences · doi:10.1175/jas-d-25-0237.1
Cuadros, D. F., Chen, X., & Tang, M. (2026) · Environments · doi:10.3390/environments13070382
Trentini, L., Lehmann, F., Crocetti, L., & Soja, B. S. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.05658 · Cited by 1
“Data-driven models such as FourCastNet (Pathak et al., 2022), Pangu-Weather (Bi et al., 2023), GraphCast (Lam et al., 2023) and the probabilistic GenCast (Price et al., 2025), among others, have demonstrated global medium-range forecast skill competitive with operational Numerical Weather Prediction (NWP) at a fractio…”
InfluentialA chemistry-informed deep learning network for mitigating the stratospheric OH data gap
Yin, W., Zhou, C., Feng, W., Dhomse, S. S., Pope, R. J., Millan, L. F., & Chipperfield, M. P. (2026) · Science Advances · doi:10.1126/sciadv.aee2319
“Notably, temporal and spatial variables (time, latitude, longitude, and SZA) are excluded from the DRCAT inputs, which is different from conventional ML practice in atmospheric science ( 17 , 33 , 34 ).”
GeoXplain: On-the-Fly Visual Explanations for Weather Foundation ModelsPreprint
Koprolin, C. W., Trentini, L., Soja, B. S., El‐Assady, M., & Humer, C. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.05655
“…many standard metrics, surpass traditional numerical weather prediction: deterministic systems such as Pangu-Weather, GraphCast, and FourCastNet [5,17,23], and probabilistic ones such as GenCast [25], operate on atmospheric grids and are evaluated with benchmark suites such as WeatherBench 2 [26].”
On the importance of historical context and bias correction in deep learning MJO prediction models
Virji, K., Geruo, A., & Velicogna, I. (2026) · npj Climate and Atmospheric Science · doi:10.1038/s41612-026-01459-9
“Thesebiasesmayalsore fl ect more general characteristics of deep learning atmospheric models, where MSE-based training objectives combined with autoregressive forecast rollouts tend to produce smooth, weak-amplitude predictions at longer leads 37 .”
Cui, Y. J., Xu, H., Wei, Q., Li, K., Feng, K., Song, Y., & Hou, J. (2026) · Sustainability · doi:10.3390/su18147020
Vine Copula-Based Probabilistic Weather Forecasting: Review, Challenges, and Future Work
Möller, A., Jobst, D., & Buchner, F. (2026) · Springer eBooks · doi:10.1007/978-3-032-14252-8_12
Lu, Y.-C., Chuang, L. L.-H., & Lai, J. (2026) · Remote Sensing · doi:10.3390/rs18142303
Evaluation of the Aurora foundation model for multi-pollutant air quality forecasting over China
Li, C. G., Chen, Z., & Yao, J. (2026) · Frontiers in Environmental Science · doi:10.3389/fenvs.2026.1882322 · Cited by 1
Short‐Term Hourly Weather Forecasting Using PredRNN With Image Preprocessing
Tran, H., Li, H., Tran, V. N., Nguyen, V. T., Le, M., Dang, T. D., Xuan, H., Pham, H. T., Bui–Thanh, T., & Leung, L.-Y. R. (2026) · Journal of Geophysical Research Machine Learning and Computation · doi:10.1029/2026jh001291
Prasad, B., Lad, D., Rafiuddin, M., Bhatia, U., & Batra, N. (2026) · doi:10.5194/egusphere-2026-4377
Kanchanaart, C., Arunkunarax, C., Saetang, K., & Pattiyanon, C. (2026) · International Joint Conference on Computer Science and Software Engineering · doi:10.1109/jcsse68839.2026.11597063
“5 [9], weather [10], prescribed-fire PM 2 .”
Multivariate time series forecasting with a numerical spiking neural P system-inspired network
Xu, J., Pan, J., Su, S. B., Li, H., Zhao, R., & Ding, F. (2026) · Neurocomputing · doi:10.1016/j.neucom.2026.134476
Spatially Aware Calibration of NWP and AI Precipitation Forecasts
Trotta, B., & Abellán, E. (2026) · Journal of Geophysical Research Machine Learning and Computation · doi:10.1029/2026jh001327
Multi-modal dual-target spatio-temporal graph learning for regional evapotranspiration estimation
Gyamfi, K. A., & Chung, . (2026) · Journal of Hydrology · doi:10.1016/j.jhydrol.2026.136014
Data-Driven Fire-Zone Segmentation for Improved Short-Term Wildfire Prediction
Caron, N., Noura, H., Guyeux, C., & Aynes, B. (2026) · IFIP advances in information and communication technology · doi:10.1007/978-3-032-30809-2_5
FRAKTALGA ASOSLANGAN MODELLASHTIRISH VA VIZUALIZATSIYA TIZIMLARIDA GENERATIV AI QO‘LLANILISHI
Elmurodovich, B. B., & Asror, B. (2026) · doi:10.65164/1yfjca59
Zheng, P., Liang, Z., Xie, F., & Yin, M. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-3438-7_35
Zhang, H., He, Y., & Sheng, Z. (2026) · Geophysical Research Letters · doi:10.1029/2026gl123655
“…research is highly concentrated on forecasting (Allen et al., 2025; Bodnar et al., 2025; Bi et al., 2023; K. Chen et al., 2025; L. Chen et al., 2023; Lam et al., 2023; Kurth et al., 2023; NOAA, n.d.), while research on reconstructing high‐fidelity, high‐dimensional atmospheric states from sparse…”
Sikora, P., & Kontogiannis, S. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.10208
Aitnazarova, A., Mamatova, U., & Zhoroeva, M. (2026) · Bulletin of Science and Practice · doi:10.33619/2414-2948/128/08
Yan, Z., Zhu, G., Tang, Y., Pan, H., Παππάς, Ν., & Chan, T.-T. (2026) · International Conference on Communications · doi:10.1109/icc59461.2026.11587081
“Notable examples include FourCast-Net [17] and GraphCast [18], which are built on large-scale neural network architectures and trained in a self-supervised autoregressive manner to predict future atmospheric states from past observations.”
Luitel, S., Mukhopadhyay, P., Singh, M., Juneja, S., & Dhanuka, S. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.11905
“GraphCast has demonstrated competitive or superior performance relative to the ECMWF operational analysis on standard verification metrics such as the anomaly correlation coefficient (ACC) and root-mean-square error (RMSE) for mid-latitude geopotential height and wind fields (Lam et al, 2023).”
InfluentialYan, Y., Li, J., Wang, Y., Yu, Z., Zheng, Y., Sun, S., Xu, T., & Feng, J. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3713650
“Recently, the advent of large-scale global foundation models, such as Pangu-Weather, GraphCast, and FourCastNet [17], [18], [19], marked a significant breakthrough in deep learning for weather forecasting.”
Subseasonal 2-m Temperature Prediction over East Asia Based on SwinUNet-AR
Ding, J., Wu, H., & Zhang, Q. (2026) · Applied Sciences · doi:10.3390/app16147162
Neural advection-diffusion equation for long-term climate dynamics
Cho, N., Cho, S. W., Hong, Y., Hwang, H. J., Lee, J. Y., & Son, H. (2026) · Neurocomputing · doi:10.1016/j.neucom.2026.134500
A Survey of Transformer-Based Time Series Forecasting Methods
Niu, H., Li, W., Chen, D., & Huang, M. (2026) · International Conference on Artificial Intelligence and Materials (ICAIM) · doi:10.1109/icaim69488.2026.11601445
NeuralGCM-En4DVar: A Differentiable Hybrid Ensemble 4D-Var Data Assimilation Framework
Guo, Y. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-3410-3_17
Unsupervised Smoke Cloud Forecasting from Multi-view Videos Using 3D Gaussians
Denis, L., Janny, S., Simeon, T., & Bronz, M. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-37679-4_3
Rehman, H. U., & Teferle, F. N. (2026) · Geoscientific model development · doi:10.5194/gmd-19-9831-2026
“Thus, even where the aggregated skill of AIFS or GraphCast matches or exceeds WRF, the physics-based model retains a clear advantage in representing the fine-scale structure and amplitude of extreme convective rainfall.”
InfluentialWang, Z., Lü, B., Xin, Y., Ito, T., Zhou, L., Cheng, L., Li, Y., Wang, X., & Jin, M. (2026) · Earth system science data · doi:10.5194/essd-18-5069-2026
From Prediction to Discovery: A Review of Scientific Discovery with Foundation ModelsPreprint
Wang, T. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7008878
Huang, W., Xu, R., Peng, B., & Zhang, S. (2026) · International Conference on Communication, Remote Sensing and Information Technology (CRSIT) · doi:10.1109/crsit69981.2026.11605208
“…adaptive Fourier neural operators [2], Pangu-Weather showed that carefully designed neural architectures can achieve strong medium-range global weather forecasting performance [3], and GraphCast further confirmed the potential of learned weather models for skillful medium-range forecasting [4].”
Gao, W., Li, J., Liu, L., Le, T. D., Liu, J., Zhao, Y., & Chen, Y. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.17507
“Domain-specialized weather systems like GraphCast (Lam et al., 2023) and Pangu-Weather (Bi et al., 2023) achieve strong numerical weather prediction but are not designed as drop-in forecasters for regional drought indices and usually involve re-training or domain-specific pipelines.”
Apeliotes: A Diffusion-Based Modeling Framework for km-scale Multi-Level Atmospheric Fields
Frastali, E. R., Paudel, A., Golbazi, M., & Liu, F. (2026) · International Conference on Machine Learning and Applications · doi:10.1109/icmla69890.2026.00057
Tong, X., & Zhou, W. (2026) · Geophysical Research Letters · doi:10.1029/2025gl117592
Observing and monitoring global runoff
Brinkerhoff, C. (2026) · Elsevier eBooks · doi:10.1016/b978-0-443-15748-6.00047-2
Cui, X., Zhang, G., & Liu, C. (2026) · Ocean Modelling · doi:10.1016/j.ocemod.2026.102803
A unified deep learning framework for subseasonal wind forecast post-processing
Damiani, A., Miyazawa, Y., Nonaka, M., Ratnam, J. V., & Patil, K. R. (2026) · Scientific Reports · doi:10.1038/s41598-026-63299-0 · Cited by 1
Diffusion model-based ensemble generation for highly nonlinear systems with limited ensemble size
Anonymous (2026) · Physical review. E · doi:10.1103/xq7z-9cg5
Introduction to Scientific Machine Learning
Bhuyan, B. P., Saadi, A. A., Meraihi, Y., & Ramdane-Chérif, A. (2026) · Springer eBooks · doi:10.1007/978-981-95-7856-6_1
Annetta, M., Purohit, S., Vagle, V., & Deo, D. (2026) · International Journal of Information Technology and Computer Science Applications · doi:10.58776/ijitcsa.v4i2.256
“GraphCast showed that graph neural network-based forecasting can outperform a leading operational numerical system on many synoptic tasks [6].”
InfluentialJie, W., Ma, D., Junkai, W., Wang, Q. Y., & Zhang, J. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7190903
Authelet, V., Dandrifosse, S., Michaud, V., Huart, J. P., Planchón, V., & Rosillon, D. (2026) · Atmosphere · doi:10.3390/atmos17080728
Zhang, A., Zhang, S., Bian, J., Xing, J., Li, W., Gu, H., & Wang, X. (2026) · Sensors · doi:10.3390/s26154717
“Thus, online-deployment-oriented evaluation of greenhouse air temperature forecasting should simultaneously consider predicted-temperature feedback and uncertainty in future exogenous sensor variables [21–23].”
From climate data to regulatory decisions: integrating climate AI into marine EIAs
Gong, X., & Cheng, L. (2026) · Frontiers in Marine Science · doi:10.3389/fmars.2026.1870082 · Cited by 1
Dai, G., Li, H., Mu, M., Wang, G., Qin, B., Zhong, X., Zhang, F., & Liang, X. S. (2026) · Advances in Atmospheric Sciences · doi:10.1007/s00376-026-6149-7
“Subsequently, DeepMind developed GraphCast, which reformulates global weather prediction by representing the atmosphere as a graph on latitude–longitude grids (Lam et al., 2023).”
Jendersie, R., Margenberg, N., Lessig, C., & Richter, T. (2026) · Computer Methods in Applied Mechanics and Engineering · doi:10.1016/j.cma.2026.119254
“Finally, we summarize our findings in section 8.”
SCANet: Scene-Conditioned Attention Network for Long-Horizon Multimodal Vessel Trajectory PredictionPreprint
Zhang, Y., Wu, H., Xian, J., Mei, X., Liang, L., & Zhang, K. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7202165
Kotharkar, V., & Stein, M. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.27160
Dutel, M. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.26629
Pope, E., Dale, K. I., Steele, E. C. C., & Graham, P. H. R. (2026) · Artificial Intelligence for the Earth Systems · doi:10.1175/aies-d-25-0068.1
Assessing the real-world economic value of weather forecasts for compounding extremes
Olivetti, L., Messori, G., Avner, P., & Stéphane Hallegatte, A. (2026) · Environmental Research Letters · doi:10.1088/1748-9326/ae9327
“More recently, data-driven forecasts have seen an explosive development, and in only a few years have matched or exceeded the performance of numerical forecasts (Bi et al 2023, Lam et al 2023, Kochkov et al 2024, Price et al 2025, Lang et al 2026).”
Dai, L., Yuan, H., Liu, Y., Li, X., Gu, X., Li, J., & Zhao, Y. (2026) · Journal of Geophysical Research Machine Learning and Computation · doi:10.1029/2026jh001347
Pengembangan Sistem Peringatan Dini Anomali Cuaca Berbasis Geospasial dan Asisten Virtual
Tarigan, Y. H., Zai, Y. G. H., & Elistia, C. (2026) · CESS (Journal of Computer Engineering System and Science) · doi:10.24114/cess.v11i2.73665
Topology-Preserving State Space Representation for Diagnosing Weather Forecast ModelsPreprint
Kim, H., & Cho, J. H. (2026) · doi:10.5194/egusphere-2026-3142
Wang, C., Li, Y., Ge, X., Zhang, H., & Lv, C. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-33865-5_30
Conformal Neural Operators with SWAN for Spatio-Temporal Forecasting Uncertainty Quantification
蔡, 铠. (2026) · Statistics and Applications · doi:10.12677/sa.2026.157167
Subseasonal Prediction of the Wintertime North Pacific Blocking in AI‐Based Weather Models
Hwang, J., Deng, Y., Yook, S., & You, Z. (2026) · Journal of Geophysical Research Atmospheres · doi:10.1029/2026jd046695
“With advances in artificial intelligence (AI) techniques and computing resources, AI‐based weather prediction models have recently emerged (e.g., Bi et al., 2023; Lam et al., 2023; Pathak et al., 2022).”
da Costa, J. V. F., Domingos, M. C., Júnior, V. A. D. S., & Gomes, J. L. (2026) · Anais do Workshop de Computação Aplicada à Gestão do Meio Ambiente e Recursos Naturais (WCAMA) · doi:10.5753/wcama.2026.22630
“Even with recent advances, accurately predicting rainfall remains challenging because of this complexity and the high computational cost associated with numerical simulations [Lam et al. 2023].”
Zhou, Z., Wu, Z., van Maren, D. S., Shen, X., Ding, J., & Ge, J. (2026) · Coastal Engineering · doi:10.1016/j.coastaleng.2026.105129
The Cascading Impact of Cyber on Climate
Chamikara, M. A. P., Grobler, M. M., Aamir, T., Ahmed, M. E., Chau, C.-K., Niemiec, M., & Romanach, L. (2026) · NATO science for peace and security series. C, Environmental security · doi:10.1007/978-3-032-13170-6_5
Dynamical-bridge fidelity as a pre-release standard for hybrid seasonal tropical-cyclone forecasts
Li, Z. (2026) · Environmental Research Communications · doi:10.1088/2515-7620/ae9468
“…to learned physical parameterizations; GenCast generates probabilistic global ensembles; and Aurora extends data-driven prediction across Earth-system domains, including short-range tropical-cyclone tracks (Bi et al 2023, Lam et al 2023, Kochkov et al 2024, Bodnar et al 2025, Price et al 2025).”
TerraNova: A Foundation Model for the AnthropocenePreprint
Rodriguez-Pardo, C., & Tavoni, M. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.29527
“Earth-system foundation models have improved the accuracy and speed of geophysical prediction, from global forecasting (Lam et al., 2023; Bi et al., 2023) to multi-variable and multi-domain adaptation (Nguyen et al., 2023; Bodnar et al., 2025), but they remain confined to the physical geometry.”
Exploring Marine Atmospheric Ducts: Current Sensing and Emerging Trends
Zhang, J., Li, Y.-J., Ji, H., Peng, H.-Y., Guo, X., & Yu-sheng, Z. (2026) · Remote Sensing · doi:10.3390/rs18152576
The Unfinished Map: Observing the Indian Ocean through Model-Corrected, Fleet-Based SensingPreprint
George, D. (2026) · doi:10.31223/x5j796
Local-Scale Wind Forecasting for Tropical Cyclone Early Warnings
Loridan, T., Bruneau, N., Mani, B., Sampson, C., & Knaff, J. A. (2026) · Bulletin of the American Meteorological Society · doi:10.1175/bams-d-25-0182.1
Metadata
- Authors
- Lam, R., Sanchez-Gonzalez, A., Willson, M., Wirnsberger, P., Fortunato, M., Alet, F., Ravuri, S., Ewalds, T., Eaton-Rosen, Z., Hu, W., Merose, A., Hoyer, S., Holland, G., Vinyals, O., Stott, J., Pritzel, A., Mohamed, S., & Battaglia, P.
- Affiliations
- Google DeepMind, London, UK.
- Google DeepMind (United Kingdom) (GB)
- Google (United States) (US)
- Corresponding author
- Willson, Matthew
- Wirnsberger, Peter
- Fortunato, Meire
- Alet, Ferran
- Ravuri, Suman
- Mohamed, Shakir
- Battaglia, Peter
- ORCID
- Venue
- Science
- Published
- 2023-12-22
- Volume
- 382
- Issue
- 6677
- Pages
- 1416-1421
- ISSN
- 0036-8075 (ISSN-L)
- 1095-9203
- OpenAlex
- W4388654737
- Cited by
- 1,466
- References
- 48
- Citations by year
- 2026: 599 · 2025: 522 · 2024: 319 · 2023: 25
- FWCI
- 161.54
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Science (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Topics
- Physical Sciences › Earth and Planetary Sciences › Atmospheric Science › Meteorological Phenomena and Simulations
- Physical Sciences › Environmental Science › Environmental Engineering › Hydrological Forecasting Using AI
- Social Sciences › Transportation › Traffic Prediction and Management Techniques
- Keywords
- medium-range weather forecasting, GraphCast, reanalysis data, machine learning weather prediction, deterministic forecasting, tropical cyclone tracking, atmospheric rivers, severe weather forecasting
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-11-08, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-11-08 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.