25 papers in the archive · tier-1 watchlist.
Systematic study of design choices for DL weather models finds fixed-grid architectures beat grid-invariant ones, proposes a hybrid, and shows multi-step fine-tuning is essential.
Software, scaling & compressionNVIDIAUniversity of Cambridge (Aardvark)
Nature Climate Change perspective on machine learning in climate modelling and analysis; abstract absent, judged from title and venue.
Climate emulationAllen Institute for AI (Ai2, ACE)California Institute of Technology (Caltech)Colorado State University (incl. CIRA)Columbia University (LEAP, Lamont-Doherty)Google ResearchInriaLawrence Berkeley National Laboratory / NERSCNSF NCAR / UCARNVIDIANew York University (M2LInES)University of California, BerkeleyUniversity of Oxford
StormCast, a generative diffusion emulator of the 3 km HRRR convection-allowing model, gives competitive 1-6 hour radar reflectivity skill with realistic cold pools and power spectra.
A 1.1-billion-parameter SFNO ensemble with bred vectors and perturbed parameters passes spectral tests and is calibrated, setting up ML huge ensembles of 1,000+ members.
Probabilistic & ensemble forecastingLawrence Berkeley National Laboratory / NERSCNVIDIAUniversity of California, Berkeley
A diffusion model (CPMGEM) emulates 2.2 km UK convection-permitting precipitation from 60 km GCM inputs, reproducing extremes up to about 100-year return times at much lower cost.
A U-Net emulator of embedded cloud-resolving models gives a stable 5-year hybrid climate run with realistic cloud condensate and a 0.96 mm/day precipitation RMSE.
Score-based diffusion data assimilation at km scale assimilates 40 sparse weather stations, lowering held-out station RMSE by 10% versus HRRR.
Lagged ensembles give a parameter-free probabilistic comparison showing GraphCast and Pangu tie on CRPS, and that multi-step losses hurt probabilistic skill.
Probabilistic & ensemble forecastingLawrence Berkeley National Laboratory / NERSCNVIDIAUniversity of California, Berkeley
ACE is a 200M-parameter autoregressive emulator that stays stable for 100 years, reproduces a 100 km reference model's climate, and runs about 100x faster.
Climate emulationAllen Institute for AI (Ai2, ACE)NVIDIAUniversity of California, Berkeley
Two-step UNet plus diffusion residual model downscales 25 km reanalysis to 2 km over Taiwan, recovering spectra and sharpening fronts and typhoon structures.
Position paper on Earth Virtualization Engines that pair high-resolution physics-based climate models with machine learning and federated access to exabyte-scale climate data.
Software, scaling & compressionCSCS - Swiss National Supercomputing CentreETH ZurichMeteoSwissMassachusetts Institute of Technology (MIT)Max Planck Institute for MeteorologyNSF NCAR / UCARNVIDIAUniversity of Reading
A deep learning atmospheric river tracker is applied to a high-resolution large-ensemble climate dataset and checked for consistency against established algorithms over the US West Coast.
Extremes & hazardsNVIDIANSF NCAR / UCARETH ZurichUniversity of OxfordUniversity of California, Berkeley
WR-Net combines TV-L1 optical flow warping with a refinement network to interpolate and predict GK2A geostationary imagery from 4- to 2-minute intervals.
A one-class SVM novelty detector disables ML corrections on out-of-sample inputs, stabilizing coarse climate runs and improving temperature and precipitation over standard ML corrections.
Hybrid physics-ML & parameterizationNVIDIAAllen Institute for AI (Ai2, ACE)Columbia University (LEAP, Lamont-Doherty)
Unsupervised clustering of storm-scale climate model output shows extreme precipitation changes driven mainly by shifts in storm regimes rather than precipitation efficiency.
FourCastNet uses adaptive Fourier neural operators to produce global medium-range forecasts about 80,000 times faster than NWP, scaling to thousands of GPUs.
Software, scaling & compressionLawrence Berkeley National Laboratory / NERSCNVIDIA
Provides a recipe for building atmospheric-science ML benchmark datasets and lists domain-specific challenges and needed benchmarks for weather, climate and air quality.
Benchmarks, datasets & evaluationECMWFForschungszentrum Jülich / Jülich Supercomputing CentreNSF NCAR / UCARNVIDIAUniversity of Oklahoma (AI2ES)
Convolutional networks learn IFS model bias from radio occultation temperatures, pretrained on ERA5 and transfer-learned on one year of IFS data, as an alternative to weak-constraint 4D-Var.
A convolutional VAE pretrained on imperfect quasi-geostrophic simulations and transfer-learned on noisy observations gives stable stochastic long-term rollouts with accurate climatology.
Climate emulationLawrence Berkeley National Laboratory / NERSCNVIDIA
FourCastNet uses adaptive Fourier neural operators to forecast global 0.25 degree weather in under 2 seconds, matching IFS at short leads and beating it for precipitation.
Global medium-range NWPCalifornia Institute of Technology (Caltech)Lawrence Berkeley National Laboratory / NERSCNVIDIA
Climate-invariant ML embeds physical knowledge into parameterization emulators so they generalize across climate regimes, maintaining offline accuracy in three atmospheric models.
Explainability, uncertainty & physicsColumbia University (LEAP, Lamont-Doherty)Google ResearchMassachusetts Institute of Technology (MIT)NVIDIAUniversity of California, Los Angeles (UCLA)
An autoencoder latent-space ensemble Kalman filter with a neural surrogate reduces data assimilation cost and beats ETKF-Q on an augmented Lorenz 96 system.
Surveys physics-informed ML approaches with ten case studies for emulating, downscaling and forecasting weather and climate, reporting greater consistency and data efficiency.
Explainability, uncertainty & physicsCalifornia Institute of Technology (Caltech)Lawrence Berkeley National Laboratory / NERSCNVIDIAStanford UniversityUniversity of California, Berkeley
ClimateNet releases an expert-labeled dataset of tropical cyclones and atmospheric rivers in CAM5.1 output, and a deep learning segmentation model that enables pixel-level extreme event analysis.
Extremes & hazardsColorado State University (incl. CIRA)ETH ZurichLawrence Berkeley National Laboratory / NERSCNSF NCAR / UCARNVIDIAUniversity of California, Berkeley
Tiramisu and DeepLabv3+ segment extreme weather patterns and scale to 27,360 V100 GPUs at 325.8 PF/s sustained throughput.
Software, scaling & compressionLawrence Berkeley National Laboratory / NERSCNVIDIAOak Ridge National Laboratory
