Ocean heat uptake in transient climate change: Mechanisms and uncertainty due to subgrid-scale eddy mixing

Joint Program Reprint • Journal Article
Ocean heat uptake in transient climate change: Mechanisms and uncertainty due to subgrid-scale eddy mixing
Huang, B., P.H. Stone, A.P. Sokolov and I.V. Kamenkovich (2003)
J. of Climate, 16(20): 3344-3356

Reprint 2003-6 [Download]

Abstract/Summary:

The ocean heat uptake (OHU) is studied using the Massachusetts Institute of Technology (MIT) ocean general circulation model (OGCM) with idealized ocean geometry. The OGCM is coupled with a statistical-dynamic atmospheric model. The simulation of OHU in the coupled model is consistent with other coupled ocean-atmosphere GCMs in a transient climate change when CO2 concentration increases by 1% per year. The global average surface air temperature increases by 1.78C at the time of CO2 concentration doubling (year 70). The ocean temperature increases by about 1.08C near the surface, 0.18C at 1000 m in the Pacific, and 0.38C in the Atlantic. The maximum overturning circulation (MOTC) in the Atlantic at 1350 m decreases by about 4.5 Sv (1 Sv [10^6 m^3 per second). The center of MOTC drifts upward about 300 m, and therefore a large OTC anomaly (14 Sv) is found at 2700 m. The MOTC recovers gradually, but the OTC anomaly at 2700 m does not seem to recover after CO2 concentration is kept constant during 400-yr simulation period.
        The diagnosis of heat flux convergence anomaly indicates that the warming in the lower latitudes of the Atlantic is associated with large-scale advection. But, the warming in the higher latitudes is associated with the heat brought down from the surface by convection and eddy mixing. In global average, the treatments of convection and eddy mixing are the two main factors affecting the OHU.
        The uncertainty of OHU due to subgrid-scale eddy mixing is studied. In the MIT OGCM this mixing is a combination of Gent-McWilliams bolus advection and Redi isopycnal diffusion (GMR), with a single diffusivity being used to calculate the isopycnal and thickness diffusion. Experiments are carried out with values of the diffusivity of 500, 1000, and 2000 m^2/sec. The total OHU is insensitive to these changes. The insensitivity is mainly due to the changes in the vertical heat flux by GMR mixing being compensated by changes in the other vertical heat flux components.
        In the Atlantic when the diffusivity is reduced from 1000 to 500 m^2/sec, the surface warming can penetrate deeper. Therefore, the warming decreases by about 0.158C above 2000 m but increases by about 0.158C below 2500 m. Similarly, when the diffusivity is increased from 1000 to 2000 m2 s21, the surface warming becomes shallower; the warming increases by about 0.28C above 1000 m but decreases by about 0.28C below 1000 m. These changes in the vertical distribution of the OHU also contribute to the insensitivity of the total OHU to changes in the GMR mixing. The analysis of heat flux convergence indicates that the difference of OHU seems to be associated with the MOTC circulation.

© 2003 American Meterological Society

Citation:

Huang, B., P.H. Stone, A.P. Sokolov and I.V. Kamenkovich (2003): Ocean heat uptake in transient climate change: Mechanisms and uncertainty due to subgrid-scale eddy mixing. J. of Climate, 16(20): 3344-3356 (http://globalchange.mit.edu/publication/14275)
  • Joint Program Reprint
  • Journal Article
Ocean heat uptake in transient climate change: Mechanisms and uncertainty due to subgrid-scale eddy mixing

Huang, B., P.H. Stone, A.P. Sokolov and I.V. Kamenkovich

2003-6
16(20): 3344-3356

Abstract/Summary: 

The ocean heat uptake (OHU) is studied using the Massachusetts Institute of Technology (MIT) ocean general circulation model (OGCM) with idealized ocean geometry. The OGCM is coupled with a statistical-dynamic atmospheric model. The simulation of OHU in the coupled model is consistent with other coupled ocean-atmosphere GCMs in a transient climate change when CO2 concentration increases by 1% per year. The global average surface air temperature increases by 1.78C at the time of CO2 concentration doubling (year 70). The ocean temperature increases by about 1.08C near the surface, 0.18C at 1000 m in the Pacific, and 0.38C in the Atlantic. The maximum overturning circulation (MOTC) in the Atlantic at 1350 m decreases by about 4.5 Sv (1 Sv [10^6 m^3 per second). The center of MOTC drifts upward about 300 m, and therefore a large OTC anomaly (14 Sv) is found at 2700 m. The MOTC recovers gradually, but the OTC anomaly at 2700 m does not seem to recover after CO2 concentration is kept constant during 400-yr simulation period.
        The diagnosis of heat flux convergence anomaly indicates that the warming in the lower latitudes of the Atlantic is associated with large-scale advection. But, the warming in the higher latitudes is associated with the heat brought down from the surface by convection and eddy mixing. In global average, the treatments of convection and eddy mixing are the two main factors affecting the OHU.
        The uncertainty of OHU due to subgrid-scale eddy mixing is studied. In the MIT OGCM this mixing is a combination of Gent-McWilliams bolus advection and Redi isopycnal diffusion (GMR), with a single diffusivity being used to calculate the isopycnal and thickness diffusion. Experiments are carried out with values of the diffusivity of 500, 1000, and 2000 m^2/sec. The total OHU is insensitive to these changes. The insensitivity is mainly due to the changes in the vertical heat flux by GMR mixing being compensated by changes in the other vertical heat flux components.
        In the Atlantic when the diffusivity is reduced from 1000 to 500 m^2/sec, the surface warming can penetrate deeper. Therefore, the warming decreases by about 0.158C above 2000 m but increases by about 0.158C below 2500 m. Similarly, when the diffusivity is increased from 1000 to 2000 m2 s21, the surface warming becomes shallower; the warming increases by about 0.28C above 1000 m but decreases by about 0.28C below 1000 m. These changes in the vertical distribution of the OHU also contribute to the insensitivity of the total OHU to changes in the GMR mixing. The analysis of heat flux convergence indicates that the difference of OHU seems to be associated with the MOTC circulation.

© 2003 American Meterological Society