REVIEW 42 cited by
AIFS -- ECMWF's data-driven forecasting system
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Signed reviews
read the original abstract
Machine learning-based weather forecasting models have quickly emerged as a promising methodology for accurate medium-range global weather forecasting. Here, we introduce the Artificial Intelligence Forecasting System (AIFS), a data driven forecast model developed by the European Centre for Medium-Range Weather Forecasts (ECMWF). AIFS is based on a graph neural network (GNN) encoder and decoder, and a sliding window transformer processor, and is trained on ECMWF's ERA5 re-analysis and ECMWF's operational numerical weather prediction (NWP) analyses. It has a flexible and modular design and supports several levels of parallelism to enable training on high-resolution input data. AIFS forecast skill is assessed by comparing its forecasts to NWP analyses and direct observational data. We show that AIFS produces highly skilled forecasts for upper-air variables, surface weather parameters and tropical cyclone tracks. AIFS is run four times daily alongside ECMWF's physics-based NWP model and forecasts are available to the public under ECMWF's open data policy.
Forward citations
Cited by 42 Pith papers
-
Do AI Forecast Ensembles Sample the Correct Conditional Distribution?
A diffusion model for coastal sea level has skillful single-station forecasts but joint spatial structure worse than a climatological draw, and the gap persists with training volume in Lorenz-96 experiments.
-
Timestep-Conditioned Transformers for Global Weather Forecasting
A single weather transformer with inference-time timestep conditioning matches specialist models across 1 to 24 hour steps and improves rollout stability under mixed-timestep training.
-
Toward Mechanistic Interpretability of an AI Foundation Model Fine-Tuned for Atmospheric Chemistry
Aurora, a foundation model fine-tuned for air quality, captures rough NOx–ozone coupling but lacks the chemical consistency and emission-plume fidelity of process-based models.
-
Global reanalysis from observations alone with machine learning
Observation-only machine learning can generate multi-decade global atmospheric reanalyses with large-scale skill near ERA5 and surface errors between ERA-Interim and ERA5, in a single day of compute.
-
Accurate Mediterranean Sea forecasting via graph-based deep learning
SeaCast, a graph neural network, makes 15-day Mediterranean Sea forecasts that outperform the operational MedFS system over the evaluated period, while producing a forecast in 20 seconds on one GPU.
-
Fixing the Double Penalty in Data-Driven Weather Forecasting Through a Modified Spherical Harmonic Loss Function
A parameter-free spectral loss that separates amplitude from coherence errors, applied as a fine-tuning loss to GraphCast, produces sharper deterministic weather forecasts and better ensemble spread.
-
Advancing global aerosol forecasting with artificial intelligence
AI-GAMFS, a deep learning model trained on 42 years of MERRA-2 reanalysis, produces 5-day global forecasts of aerosol components and meteorology faster and more accurately than CAMS, GEOS-FP, and several regional dust...
-
DeepMedcast: A Deep Learning Method for Generating Intermediate Weather Forecasts among Multiple NWP Models
A U-Net trained to interpolate a single weather model in time can also blend two different models' forecasts at the same time into a realistic intermediate field.
-
HourGlass: A probabilistic data-driven temporal downscaler for global and regional weather forecasting
HourGlass probabilistically reconstructs hourly weather evolution between 6-hourly forecast states using CRPS training on NWP trajectories, preserving skill and small-scale variability better than deterministic downscalers.
-
MarsCast: Transfer Learning of AI Weather Foundation Models to Planetary Atmospheres
Fine-tuning GraphCast on Mars Climate Database data enables the Earth-trained model to capture Martian diurnal, seasonal, and vertical temperature structure in as few as 10 epochs.
-
Prithvi-Precip: Integrating Satellite Observations into an Atmospheric AI Foundation Model for Precipitation Forecasting
Training on IMERG satellite precipitation and directly ingesting satellite observations improves medium-range AI precipitation forecasts, with the largest gains at lead times under 40 hours.
-
Weather Emulators at the Frontier of Heat Extremes Predictability
At 10–15 day leads, AI weather emulators can match or beat dynamical models on global temperature skill but under-represent heat-extreme intensity and lose to IFS on recall.
-
AIFS-SUBS: Extending Data-Driven Forecasting to Sub-Seasonal Timescales
AIFS-SUBS matches IFS probabilistic skill at weeks 2–6 with reduced biases, extends skilful MJO OLR forecasts by eight days, and runs at ~200× lower energy cost.
-
Automated Data Readiness for Scientific AI
REDI automates raw-to-AI-ready scientific data prep via a five-stage IPTSO pipeline with provenance, readiness scoring, and multi-domain validation matching expert references.
-
Scaling Laws of Global Weather Models
Across five global weather models, validation loss follows power-law scaling, with wider architectures and larger training datasets outperforming deeper or smaller-data configurations.
-
Learning from nature: insights into GraphDOP's representations of the Earth System
GraphDOP, trained purely on meteorological observations, spontaneously forms unified latent representations of Earth system state, viewing effects, and synoptic dynamics.
-
A comparison of stretched-grid and limited-area modelling for data-driven regional weather forecasting
Stretched-grid and limited-area machine-learning weather models are competitive for Europe, with stretched-grid models showing better generalization to unseen forecast times and limited-area models benefiting from ext...
-
Jigsaw: Training Multi-Billion-Parameter AI Weather Models with Optimized Model Parallelism
An MLP-based weather model called WeatherMixer trains efficiently with a new Jigsaw parallelization scheme that shards data and model across GPUs, reaching 11 PFLOPs on 256 GPUs with 72% scaling efficiency.
-
Forecast error diagnostics in neural weather models
Correcting the stratosphere or boundary layer toward reanalysis truth substantially improves mid-latitude medium-range forecasts in a neural weather model; correcting the tropics does not.
-
PEAR: Equal Area Weather Forecasting on the Sphere
A transformer weather model operating natively on the equal-area HEALPix grid beats an equiangular-grid counterpart at longer lead times with 2.6x fewer parameters.
-
Diffusion-LAM: Probabilistic Limited Area Weather Forecasting with Diffusion
A diffusion-based limited area weather model that uses future boundary conditions from a global model improves short-lead forecast accuracy and boundary consistency.
-
AIFS-CRPS: Ensemble forecasting using a model trained with a loss function based on the Continuous Ranked Probability Score
A stochastic transformer weather model trained with the almost fair CRPS loss produces ensemble forecasts that beat ECMWF's physics-based IFS ensemble on most medium-range variables and match it at subseasonal timescales.
-
GraphDOP: Towards skilful data-driven medium-range weather forecasts learnt and initialised directly from observations
A graph-neural-network weather model trained only on raw observations produces skillful global forecasts out to five days, with tropical 2-meter temperature forecasts competitive with the operational IFS.
-
Global spatio-temporal downscaling of ERA5 precipitation through generative AI
A conditional GAN downscales ERA5 precipitation from 24 km/hourly to 2 km/10-minute resolution, with realistic small-scale patterns and extremes in Germany, the US, and Australia.
-
FengWu-W2S: A deep learning model for seamless weather-to-subseasonal forecast of global atmosphere
An AI model trained on ERA5 generates skillful 6-hourly global forecasts out to 42 days, matching or exceeding ECMWF on subseasonal indices like MJO and NAO.
-
A robust and stable hybrid neural network/finite element method for 2D flows that generalizes to different geometries
Replay training on the DNN-MG's own perturbed trajectories removes the long-time instability of the hybrid Navier-Stokes solver, while Transformers or larger patches improve accuracy on unseen geometries.
-
Evaluating Extreme Precipitation Forecasts: A Threshold-Weighted, Spatial Verification Approach for Comparing an AI Weather Prediction Model Against a High-Resolution NWP Model
Combining HiRA neighborhood verification with threshold-weighted CRPS shows that AI-vs-NWP rankings for extreme precipitation depend strongly on neighborhood size.
-
CRPS-LAM: Probabilistic Regional Weather Forecasting with Continuous Ranked Probability Score
CRPS-LAM produces 57-hour probabilistic limited-area forecasts on MEPS at diffusion-comparable accuracy with single-forward-pass sampling, roughly 39x faster than Diffusion-LAM.
-
Contribution of expert aggregation to temperature prediction part II: Second order bounds with sleeping experts
A sleeping-expert framework, activated by gradient boosted trees, makes an online forecast aggregation more reactive without raising its average error, slightly lowering the 95th percentile of absolute temperature errors.
-
A multi-scale loss formulation for learning a probabilistic model with proper score optimisation
Adding a multi-scale loss to proper-score-trained AIFS-CRPS reduces small-scale variability in forecasts without changing skill scores.
-
Deep Learning Weather Models for Subregional Ocean Forecasting: A Case Study on the Canary Current Upwelling System
An adapted GraphCast graph neural network trained on satellite sea surface temperature outperforms ConvLSTM and the GLORYS reanalysis for medium-range forecasts in the Canary Current upwelling system.
-
Utilizing Strategic Pre-training to Reduce Overfitting: Baguan -- A Pre-trained Weather Forecasting Model
Baguan shows that a two-frame Siamese masked autoencoder pre-training step reduces overfitting in a plain ViT weather model, improving medium-range forecasts over Pangu-Weather and IFS on the 2018 ERA5 test set.
-
Machine Learning (ML)-Physics Fusion Model Outperforms Both Physics-Only and ML-Only Models in Typhoon Predictions
A hybrid model that nudges machine-learning forecasts into a physics-based typhoon model reduces track and intensity errors compared to either approach alone in a full-season evaluation.
-
Leadsee-Precip: A Deep Learning Diagnostic Model for Precipitation
Leadsee-Precip is a deep learning diagnostic model that converts circulation fields into 6-hour precipitation with a loss weighted toward rare heavy rain, and reports strong hit scores for heavy rain in China.
-
Data-driven Surface Solar Irradiance Estimation using Neural Operators at Global Scale
An AFNO-based neural network estimates global 6-hourly solar irradiance from weather model inputs, with satellite fine-tuning improving accuracy against ground stations.
-
On the Genealogy of Machine Learning Weather Prediction
ML weather prediction inherited NWP’s IVP framing, so model choice should explicitly align with either physical system structure or data statistical structure rather than defaulting to learned time-steppers.
-
The Rise of AI in Weather and Climate Information and its Impact on Global Inequality
AI weather and climate tools inherit Northern-controlled data and compute, risking worse forecasts and maladaptation for the Global South rather than democratizing climate information.
-
Post-processing of ensemble photovoltaic power forecasts with distributional and quantile regression methods
On Hungarian PV data, every tested post-processing method improves raw ensemble forecasts, and nonlinear quantile regression performs best.
-
PyG 2.0: Scalable Learning on Real World Graphs
PyG 2.0 is presented as a modular, scalable graph-learning framework with heterogeneous and temporal graph support, compilation-based speedups, and explainability.
-
Modernizing CNN-based Weather Forecast Model towards Higher Computational Efficiency
A 7-million-parameter convolutional weather model trains in 12 hours on one GPU and is reported to match or beat much larger AI and numerical weather models in medium-range forecasts.
-
EPT-2 Technical Report
EPT-2 is a 9 km global AI weather model reported to beat Aurora, IFS HRES, and the ECMWF ENS mean over 0 to 240 hours on energy-relevant variables.
-
Contribution of expert aggregation to temperature prediction part i
Expert aggregation algorithms, especially BOA and MLpol with a gradient trick, improved deterministic 2m temperature forecasts at 33 French stations and matched the best fixed convex combination in hindsight.
Discussion (0). Continue with ORCID to comment.