FASST: A Compact Stateless Atmosphere Model for Long-Term Ocean Coupling

Pratham Aggarwal; Duncan Watson-Parris; William E. Chapman

physics.ao-phOct 26, 2026

FASST: A Compact Stateless Atmosphere Model for Long-Term Ocean Coupling

Authors: Pratham Aggarwal, Duncan Watson-Parris, William E. Chapman

Organizations: University of California San Diego · University of Colorado Boulder

Abstract

Coupled climate simulations are computationally expensive, largely because of the atmosphere. We introduce FASST, a coupled climate model that retains the prognostic POP2 ocean while replacing the atmosphere and sea ice with lightweight U-Net emulators. The atmospheric emulator is diagnostic, predicting surface fluxes from the evolving ocean-ice state every $6$ simulated hours, while the sea-ice emulator advances the ice state monthly. For $140$ years of unforced integration, FASST maintains a stable global-mean climate without SST restoring, with post-transient global-mean SST of $291.30 \pm 0.22\mathrm{K}$ compared with $291.25 \pm 0.10\mathrm{K}$ in CESM $2$ piControl. FASST also captures interannual Niño- $3.4$ variability, although with weaker amplitude than CESM ( $0.62\mathrm{K}$ versus $1.02\mathrm{K}$ ). These results demonstrate that a stateless learned atmosphere can sustain stable, free-running coupled climate simulations over century-scale integrations.

Figures & tables

Figure 1: Long-term behavior of FASST (orange, solid) and CESM

2

piControl (grey, dashed) over

140

years. The shaded region denotes the initial spin-up, and the vertical dashed line marks its end at year

31

. (a) Global-mean SST. Following spin-up, FASST has a mean of

291.30\pm 0.22~\mathrm{K}

compared with

291.25\pm 0.10~\mathrm{K}

for piControl. (b) Niño-

3.4

SST anomaly. FASST captures interannual variability but with lower amplitude, with standard deviations of

0.62

and

1.02~\mathrm{K}

, respectively.

Figure 2: Annual-mean global SST anomaly following a

+2~\mathrm{K}

perturbation to the top

100~\mathrm{m}

of the initial ocean state, relative to an identical unperturbed control. The anomaly decays from

+1.90~\mathrm{K}

at model year 0 to

+0.98~\mathrm{K}

at model year

4

, a

48

% reduction, with no runaway amplification.

Appendix figures & tables8 assets

Supplementary material from the paper’s appendix.

Appendix

Hyperparameter	Atmospheric	Sea-ice
Loss	MSE	MSE
Optimizer	AdamW	AdamW
Learning rate	$3\times 10^{-4}$	$2\times 10^{-4}$
Weight decay	$10^{-4}$	$10^{-4}$
LR schedule	Cosine	Cosine
Epochs	80	80

Appendix

Table 1: Training hyperparameters and computational cost for the atmospheric and sea-ice emulators.

Atmosphere emulator ( $6$ -hourly, diagnostic)
Role	Variable	Description	Units
Input	SST	Sea-surface temperature (from POP2)	K
Input	ICEFRAC	Sea-ice fraction (from sea-ice emulator)	0–1
Input	SOLIN	Incoming solar radiation, top of atmosphere	W m -2
Input	SST_prev	Sea-surface temperature, $24$ h lag	K
Input	ICEFRAC_prev	Sea-ice fraction, $24$ h lag	0–1

Appendix

Table 2: Input and target variables for the two emulators. The atmosphere emulator is diagnostic and operates every

6

hours; the sea-ice emulator is autoregressive and operates monthly.

Atmospheric emulator
Variable	$R^{2}$	RMSE
PS	0.9996	$5.9646\times 10^{2}$
Tbot	0.9994	1.8736
FLDS_J	0.9840	$3.2478\times 10^{5}$
Qbot	0.9678	$8.6057\times 10^{-4}$
QFLX	0.9243	$1.6252\times 10^{-4}$

Appendix

Table 3: Test-set performance of the atmospheric emulator against CREDIT data and the sea-ice emulator against CESM2 Large Ensemble member 1231.002.

Figure 3: Test-period spatial maps of the atmospheric emulator’s boundary-layer state outputs (Ubot, Vbot, Tbot, Qbot, PS; $R^{2}=0.72$ – $1.00$ ): reference (left), FASST prediction (middle), and error (right), averaged over the held-out test period.

Figure 4: Test-period spatial maps of the atmospheric emulator’s turbulent and moisture flux outputs (SHFLX, LHFLX, QFLX; $R^{2}=0.87$ – $0.92$ ): reference (left), FASST prediction (middle), and error (right), averaged over the held-out test period.

Figure 5: Test-period spatial maps of the atmospheric emulator’s radiation and precipitation outputs (FSDS, FLDS, PRECT): reference (left), FASST prediction (middle), and error (right), averaged over the held-out test period. Radiative fluxes are well reproduced ( $R^{2}\geq 0.93$ ); PRECT is the weakest-performing variable in the atmospheric emulator ( $R^{2}=0.48$ ).

Figure 6: Test-period spatial maps of the atmospheric emulator’s wind-stress outputs (TAUX, TAUY; $R^{2}=0.63$ – $0.73$ ): reference (left), FASST prediction (middle), and error (right), averaged over the held-out test period. These are the weakest-performing forced fields, of interest given their direct role in ocean momentum forcing.

Figure 7: FASST versus CESM $2$ piControl sea-surface temperature climatology. Columns show CESM $2$ piControl (truth), FASST (predicted), and the corresponding error (FASST minus truth). Rows show the global domain, tropical Pacific, North Atlantic, Southern Ocean, and Arctic. Climatologies are computed over the post-transient period (model years 31–140) using elapsed-year-matched intervals.

Climate Modeling

Climate modeling uses physical simulations and data-driven methods to understand, predict, and project Earth’s climate and its response to external forces.

See all 69 papers

FloeNet: A mass-conserving global sea ice emulator that generalizes across climates
Mar 12, 2026William Gregory, Mitchell Bushuk, James Duncan +8Climate Modeling Climate Emulation
Atmospheric and Oceanic Sciences Program, Princeton University, NJ, USA · Geophysical Fluid Dynamics Laboratory, NOAA, Princeton, NJ, USA · Allen Institute for Artificial Intelligence (Ai2), Seattle, USA +1
Stochastic Emulation of a Fully Coupled Preindustrial E3SMv3 Simulation
Aug 10, 2026Elynn Wu, James P. C. Duncan, Troy Arcomano +11Surrogate Modeling Climate Science
Allen Institute for Artificial Intelligence (Ai2), Seattle, WA, USA · Pacific Northwest National Laboratory, Richland, WA, USA · Los Alamos National Laboratory, Los Alamos, NM, USA +2
HClimRep-Ocean: A Global Ocean Emulator on an Unstructured Mesh
Sep 23, 2026Kacper Nowak, Aleksei Koldunov, Nikolay Koldunov +7Neural Surrogate Modeling Climate Modeling
Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research, Bremerhaven, Germany · Forschungszentrum Jülich GmbH, Jülich Supercomputing Center, Jülich, Germany · Karlsruhe Institute of Technology, Karlsruhe, Germany +3

Mar 12, 2026J/K move · Enter open · S save

FloeNet: A mass-conserving global sea ice emulator that generalizes across climates

William Gregory, Mitchell Bushuk, James Duncan, Elynn Wu+7

Atmospheric and Oceanic Sciences Program, Princeton University, NJ, USA · Geophysical Fluid Dynamics Laboratory, NOAA, Princeton, NJ, USA · Allen Institute for Artificial Intelligence (Ai2), Seattle, USA+1

Climate Modeling Climate Emulation Neural Surrogate Modeling

Open paper AskPDF arXiv

We introduce FloeNet, a machine-learning emulator trained on the Geophysical Fluid Dynamics Laboratory global sea ice model, SIS2. FloeNet is a mass-conserving model, emulating 6-hour mass and area budget tendencies related to sea ice and snow-on-sea-ice growth, melt, and advection. We train FloeNet using simulated data from a reanalysis-forced ice-ocean simulation and test its ability to generalize to pre-industrial control and 1% CO2 climates. FloeNet outperforms a non-conservative model at reproducing sea ice and snow-on-sea-ice mean state, trends, and inter-annual variability, with volume anomaly correlations above 0.96 in the Antarctic and 0.76 in the Arctic, across all forcings. FloeNet also produces the correct thermodynamic vs dynamic response to forcing, enabling physical interpretability of emulator output. Finally, we show that FloeNet outputs high-fidelity coupling-related variables, including ice-surface skin temperature, ice-to-ocean salt flux, and melting energy fluxes. We hypothesize that FloeNet will improve polar climate processes within existing atmosphere and ocean emulators.

Abstract

Climate Modeling

Explore similar work

FloeNet: A mass-conserving global sea ice emulator that generalizes across climates

Stochastic Emulation of a Fully Coupled Preindustrial E3SMv3 Simulation

HClimRep-Ocean: A Global Ocean Emulator on an Unstructured Mesh