Delay due to oceanic heat capacity

Numerical experiment with an ocean-atmosphere coupled climate model in idealized condition [Spelman and Manabe, 1984]

Idealized geography, symmetric w.r.t equator. No seasonal cycle.

• A: Standard exp. Present CO₂ concentration (300 ppm) is kept. Look at the quasi-steady state.
• B: Steady-state response exp. 4-times the present CO₂ concentration (1200 ppm) is kept. Look at the quasi-steady state. Difference "B - A" is considered as "steady-state response to CO₂ quadrupling". (From experience of previous numerical experiments, response to 4xCO2 is nearly twice of response to 2xCO2. The authors used larger forcing in order to make signals more visible.)
• C: Transient response exp. to step-like forcing. At one time in the quasi-steady-state of A, CO₂ concentration is quadrupled suddenly, and kept at the new level.

The result of C generally looks like approaching the state of B according to an exponential function of time (with a negative coefficient).

Result: Meridional (South-North) - vertical section of temperature difference

B (steady state exp.) - A (standard exp.)

• Stratosphere cools.
• Troposphere and ocean warm.
• In troposphere, the lower part in high latitudes warms strongly. Because ...
  • In cold climates, "near-surface temperature inversion" (dT/dz > 0) is often observed. In warmer climates, this situation occurs less often.
  • There are less snow and sea ice in warmer climates. (Ice-albedo feedback)
  • Warmer climates have more condensation of water vapor, resulting in smaller lapse rate (-dT/dz).
  • (These facts already appear in Manabe & Wetherald 1975, where just the heat capacity of the mixed layer (～100 m) was considered as the ocean's role.)
• The upper ocean warms toghether with the troposphere at the same latitude.
• The deep ocean warms together with the lower troposphere in the high latitude, where the sinking water forms.

Result: Meridional (South-North) - vertical section of temperature difference

C (25 years after CO₂ quadrupling) - A (standard exp.)

• Temperature of deep ocean hardly changes yet in 25 years.
• Temperature changes in the atmosphere and the upper ocean is similar to the steady-state response, and the magnitude is approx. 70 % of s.s.r.

The delay of global warming due to ocean heat capacity can be approximated by a linear model with a time constant corresponding to the heat capacity of the upper ocean (the part of ocean where wind-driven circulation dominates). The time series can be approximated by an exponential function tending to the new steady state.

A rough estimate of future temperature can be made by giving a realistic scenario of CO₂ concentration to the linear model. (More detailed estimate will be done by giving such a CO₂ concentration scenario to coupled 3-dimensional climate models.