Model description

Variables, training data, architecture, and evaluation of AuroraChem v1

Overview

AuroraChem is a fine-tuned version of the Aurora atmospheric foundation model (Bodnar et al., 2025). Aurora is a 1.3 billion-parameter model pretrained on more than 1 PB of reanalysis, analysis, forecast, and climate model output. AuroraChem adds seven chemical species to the pretrained model and keeps its meteorological variables.

AuroraChem takes the atmospheric state at two times, t − 6 h and t, and predicts the state at t + 6 h. Longer simulations, or rollouts, feed each prediction back in as the next input.

Variables and grid

All fields are on a 4°×5° global grid. Three-dimensional fields are on 13 pressure levels: 1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, and 50 hPa.

CategoryFieldsLevelsRole
ChemicalOH, Cl, NO, NO2, O3, CO, 14COPressurePredicted
MeteorologicalTemperature, specific humidity, zonal and meridional wind, geopotential heightPressurePredicted
Surface meteorology2 m temperature, 10 m zonal and meridional wind, mean sea level pressureSurfacePredicted
EmissionsMonthly anthropogenic CO, NOx, NH3, SO2SurfaceInput
StaticLand-sea mask, soil type, topography, time of day, day of week, day of yearSurfaceInput

Training data

The chemical fields come from a GEOS-Chem v14.5.3 simulation performed for the FETCH4 project, which adds 14CO to the chemical mechanism. GEOS-Chem is driven by MERRA-2 meteorology. The meteorological fields come from ERA5. Emissions are 2019 monthly mean anthropogenic emissions from the CAMS global inventory, used for both the training and test years. The model is trained on 2018 and evaluated on 2019.

ERA5 fields are bilinearly interpolated from 0.25°×0.25° to 4°×5°. GEOS-Chem fields are conservatively regridded on each model level, then linearly interpolated from the hybrid sigma-pressure levels to the 13 pressure levels. Meteorological fields are normalized by their mean and standard deviation, and chemical fields by half of their time-averaged maximum.

Architecture

Schematic of the AuroraChem encoder, backbone, and decoder
The encoder maps the fields at t − 6 h and t to a latent state with four vertical levels. The backbone advances the latent state by 6 h. The decoder maps it back to the output fields at t + 6 h.

AuroraChem keeps the encoder, Swin Transformer U-Net backbone, and decoder of the base Aurora model, with three changes. The patch size is 1×1 instead of 4×4, which is affordable at 4°×5°. Each pressure level has its own patch embedding, because chemical species vary more with altitude than meteorological fields do. The chemical fields have their own decoder, and the meteorological fields use the pretrained decoder.

Training

All weights are updated during fine-tuning. The new encoder and decoder layers are trained from scratch at a learning rate of 0.01. The pretrained weights, which make up 98% of the model, are updated at a learning rate of 0.001. The loss is a latitude-weighted mean absolute error, with each chemical field weighted so that all fields contribute about equally.

Evaluation

AuroraChem is evaluated against GEOS-Chem for chemistry and ERA5 for meteorology using the spatial correlation Rs at each pressure level. After 3 days, Rs at 500 hPa is 0.97 for temperature, 0.96 for OH, 0.91 for O3, and 0.96 for 14CO. After six months, it is 0.90 for temperature, 0.86 for OH, 0.59 for O3, and 0.61 for 14CO.

Maps of true and AuroraChem 3-day forecasts of temperature, OH, ozone, and 14CO at 500 hPa
3-day forecast at 500 hPa, valid 12 UTC on 4 January 2019. Left: ERA5 and GEOS-Chem. Right: AuroraChem.

Replacing fields with the true values every 5 days keeps the rollout close to the truth, including fields that are never replaced. Replacing only the meteorological fields keeps 500 hPa OH at an annual mean Rs of 0.93, compared with 0.89 without replacement. Replacing only the chemical fields keeps 500 hPa temperature at 0.92, compared with 0.89. Inserting only NO2 columns at TROPOMI pixels keeps 500 hPa OH at 0.93.

Time series of spatial correlation for temperature, OH, and ozone in four rollout configurations
Rs at 500 hPa in 2019 for a free rollout and for rollouts where all fields, only meteorological fields, or only chemical fields are replaced every 5 days.

Known limitations

  • Free rollouts have a weak seasonal cycle. A rollout started in January stays close to Northern Hemisphere winter.
  • O3 and 14CO drift after a few months of free running, with artifacts over the Amazon, the Congo basin, and the Maritime Continent.
  • The model receives no biomass burning or biogenic emissions.
  • The same emissions were used in the training and test years, so the response to emission changes is untested.