1991

Transient Responses of a Coupled Ocean-Atmosphere Model to Gradual Changes of Atmospheric Carbon Dioxide

S. Manabe, R. J. Stouffer, M. J. Spelman & K. Bryan

Observation and Context

Greenhouse warming studies traditionally look at “equilibrium responses,” which measure the permanent climate state achieved over infinite time after an abrupt jump in carbon dioxide (CO₂). However, real-world greenhouse gases increase gradually, meaning the actual climate layout over time depends heavily on how the oceans continuously circulate and absorb expanding thermal traps. Furthermore, simple linear models assume that reducing CO₂ generates an identical but perfectly mirrored cold response compared to a CO₂ increase. To understand how the global climate system actually reacts to real-world, time-dependent fluctuations, scientists must evaluate the specific physical and chemical changes that happen within the deep ocean branches.

Hypothesis

If a three-dimensional coupled ocean-atmosphere climate model is forced with gradual, asymmetric changes in atmospheric CO₂ (1% annual growth versus 1% annual reduction), then the resulting thermal penetration patterns will be non-mirrored (asymmetric) due to variations in deep-ocean static stability driven by salinity changes, but the net global surface air temperature change will remain surprisingly similar in magnitude due to compensating albedo and thermal inertia feedbacks.

Experiment and Methodology

The researchers used a global coupled general circulation model combining atmosphere, land surface, and ocean systems. The atmosphere utilized a spectral transform method with nine vertical levels, while the ocean utilized primitive equations with global geography and twelve vertical layers. To prevent the model from drifting into an unrealistic state, seasonal and geographical flux adjustments for heat and water were applied at the ocean-atmosphere boundary.

Starting from a stable quasi-equilibrium initial climate, the team ran three distinct 100-year numerical integrations:

1. G Integration: Atmospheric CO₂ increased by 1% per year compounded.
2. S Integration: Atmospheric CO₂ remained entirely constant as a control benchmark.
3. D Integration: Atmospheric CO₂ decreased by 1% per year compounded.

The transient climate changes were calculated by taking the differences between the growing/decreasing runs and the standard control run. They also compared these transient outcomes to an atmosphere-mixed-layer ocean model (AM model) to isolate the exact role of deep-sea heat storage.

Results and Data

• The Sinking Branch Delay: Surface temperature changes were remarkably slow over the northern North Atlantic and the Circumpolar Ocean of the Southern Hemisphere. In these zones, the deep downwelling branches of the ocean’s circulation mixed heat down across columns extending multiple kilometers deep, stalling sea surface warming.
• Land-Sea Contrast: In mid-latitudes, surface air warming was significantly stronger over continents than over oceans. This contrast matches the equilibrium tests because wet oceans expand their evaporative heat loss when warmed, whereas dry soils cannot.
• Asymmetric Penetration Depth: The penetration of cold anomalies in the CO₂ reduction experiment was significantly deeper than the penetration of warm anomalies in the growth experiment. In the CO₂ growth experiment, enhanced high-latitude precipitation capped the polar oceans with fresh, low-salinity water, which increased static stability and suppressed vertical mixing. In the reduction experiment, this freshwater capping reversed, causing deep convective overturning.
• Global Compensating Feedback: Despite the cold experiment having a much larger oceanic thermal inertia (slowing temperature changes down), the net change in global mean surface air temperature was nearly identical in magnitude to the warming experiment (global averages of −2.19 °C vs +2.31 °C at the time of halving/doubling). This occurred because the cold experiment triggered a massive expansion of high-albedo snow and sea ice, creating a powerful positive surface feedback that compensated for the ocean’s extra thermal drag.

Conclusion and Climate Impact

The hypothesis was supported. The experiment proved that the ocean’s dynamic structure responds non-linearly to climate forcings, showing that a cooling atmosphere triggers deeper thermal impacts than a warming one due to salinity and stability changes.

The climate impact highlighted by this model shows that deep-ocean sinking zones act as temporary shields against immediate global warming by burying heat thousands of meters below the surface. However, this comes at a severe environmental cost: the added moisture in a warming troposphere triggers heavy freshwater runoff into the Arctic and North Atlantic, which systematically weakens the ocean’s thermohaline circulation. Over time, this freshwater capping changes how the ocean redistributes heat globally, distorting historical weather trends and shifting climate impacts unpredictably across both hemispheres.