2012

GFDL’s ESM2 Global Coupled Climate-Carbon Earth System Models: Physical Formulation and Baseline Simulation Characteristics

Observation and Context

Accurate climate modeling requires understanding how heat, carbon, and water cycle through the atmosphere, land, and oceans. Traditional ocean models divide the sea into horizontal layers based on depth (z-coordinate system). While this works well near the surface, it often causes artificial numerical mixing and struggles to model deep, gravity-driven currents down underwater slopes. Alternatively, models can divide the ocean along layers of equal water density (isopycnal coordinates), which naturally follow deep ocean flows but present challenges in weakly layered upper waters. Scientists needed to know how these two mathematical frameworks impact simulated global climate.

Hypothesis

If identical atmospheric, land, sea ice, and biogeochemical models are coupled to two different ocean dynamical cores—one based on depth levels and the other based on density layers—both models will achieve stable climate states, but they will produce distinct differences in deep ocean circulation, internal mixing, and climate variability patterns such as the El Niño–Southern Oscillation (ENSO).

Experiment and Methodology

Researchers at the Geophysical Fluid Dynamics Laboratory (GFDL) developed two coupled Earth System Models:

  • ESM2M: Uses the MOM4p1 ocean core with vertical depth/pressure coordinates (z∗).
  • ESM2G: Uses the GOLD ocean core with vertical density (isopycnal) layers and an active bulk surface mixed layer.
  • Shared Components: Both shared the AM2 atmosphere, LM3.0 land and vegetation model, sea ice model, dynamic iceberg tracking, and carbon cycle representations.
  • Both models were initialized with modern ocean temperature and salinity data and run for over 1,000 model years under fixed pre-industrial (1860) conditions to reach a stable equilibrium before evaluation against modern observations.

Results and Data

  • Thermal Drift and Stratification: Over 1,000 years, ESM2M experienced a warm deep-ocean drift of +0.60∘C due to spurious interior numerical mixing, resulting in an overly deep thermocline (173×1015 m3). ESM2G remained much closer to equilibrium, cooling by only 0.18∘C with a shallower thermocline (131×1015 m3).
  • Ocean Circulation: ESM2G sustained more realistic, deeper-reaching North Atlantic Deep Water (NADW) flows and stronger Antarctic bottom water entry into the deep Pacific (16 Sv vs. 6 Sv in ESM2M). In contrast, ESM2M better captured the Antarctic Circumpolar Current transport through the Drake Passage (164 Sv) compared to ESM2G (108 Sv vs. observed 140 Sv).
  • Climate Variability: Ocean physics heavily altered atmospheric climate cycles. ENSO was overly active and frequent in ESM2M, but overly muted in ESM2G. ESM2M produced slightly better patterns for sea surface temperature (r2=0.99), salinity, and sea surface height.

Conclusion and Climate Impact

The hypothesis was supported: changing only the ocean’s vertical coordinate system profoundly impacts ocean heat storage, deep-water ventilation, and large-scale climate variability like ENSO. Neither ocean formulation was fundamentally superior. ESM2G is better suited for studies on total ocean heat storage and deep circulation due to its minimal drift, while ESM2M excels at surface climate, sea level height, and tropical Pacific variability. Providing both models to the CMIP5 project allows scientists to evaluate structural model uncertainties, leading to more reliable predictions of long-term human-driven climate change.

Full Citation

Dunne, J. P., John, J. G., Adcroft, A. J., Griffies, S. M., Hallberg, R. W., Shevliakova, E., Stouffer, R. J., Cooke, W., Dunne, K. A., Harrison, M. J., Krasting, J. P., Malyshev, S. L., Milly, P. C. D., Phillipps, P. J., Sentman, L. T., Samuels, B. L., Spelman, M. J., Winton, M., Wittenberg, A. T., & Zadeh, N. (2012). GFDL’s ESM2 Global Coupled Climate-Carbon Earth System Models. Part I: Physical Formulation and Baseline Simulation Characteristics. Journal of Climate, 25(19), 6646–6665. https://doi.org/10.1175/JCLI-D-11-00560.1