Stability and Variability in a Coupled Ocean–Atmosphere Climate Model: Results of 100-year Simulations

View More View Less
  • 1 Department of Meteorology, University of Wisconsin-Madison, Madison, Wisconsin
© Get Permissions
Full access

Abstract

Two 100-year seasonal simulators, one performed with a low resolution atmospheric general circulation model (GCM) coupled to a mixed-layer ocean formulation and the other made with the GCM forced by prescribed ocean conditions, are compared to assess the effects of an interactive ocean and sea-ice component on the stability and interannual variability of a climate system. Characteristics of the time variation of surface temperature, 700 mb temperature and sea-ice coverage are analyzed for selected land and ocean areas. Both simulations showed stable seasonal cycles of basic variables, although small trends were found. These trends were roughly linear in nature and quite distinct from all other components of variability. Detrended time series were used to describe the other aspects of variability.

There was pronounced interannual variability in the simulations from both models as seen in the time series for temperature and sea ice over the entire 100-year time period. Consistent with observations, variations tended to be larger in polar areas and over land. The inclusion of the interactive ocean and sea-ice component produced a red spectrum for surface temperature but not for 700-mb temperature. Using a linearized air-sea model patterned after the coupled models, this result is shown to be linked to the combined effects of the model longwave cooling and ocean-atmosphere energy exchange. The shift towards lower frequency in surface temperature was most evident in polar regions and occurred in conjunction with very low frequency (even decadal-scale) variability in the computed sea-ice coverage. The simulated mean and variability characteristics of sea ice corresponded fairly well with observations. This suggests that the low resolution model is able to represent some relevant aspects of the physics of climate fluctuations and thus provide useful simulations for studies of interannual variability.

Abstract

Two 100-year seasonal simulators, one performed with a low resolution atmospheric general circulation model (GCM) coupled to a mixed-layer ocean formulation and the other made with the GCM forced by prescribed ocean conditions, are compared to assess the effects of an interactive ocean and sea-ice component on the stability and interannual variability of a climate system. Characteristics of the time variation of surface temperature, 700 mb temperature and sea-ice coverage are analyzed for selected land and ocean areas. Both simulations showed stable seasonal cycles of basic variables, although small trends were found. These trends were roughly linear in nature and quite distinct from all other components of variability. Detrended time series were used to describe the other aspects of variability.

There was pronounced interannual variability in the simulations from both models as seen in the time series for temperature and sea ice over the entire 100-year time period. Consistent with observations, variations tended to be larger in polar areas and over land. The inclusion of the interactive ocean and sea-ice component produced a red spectrum for surface temperature but not for 700-mb temperature. Using a linearized air-sea model patterned after the coupled models, this result is shown to be linked to the combined effects of the model longwave cooling and ocean-atmosphere energy exchange. The shift towards lower frequency in surface temperature was most evident in polar regions and occurred in conjunction with very low frequency (even decadal-scale) variability in the computed sea-ice coverage. The simulated mean and variability characteristics of sea ice corresponded fairly well with observations. This suggests that the low resolution model is able to represent some relevant aspects of the physics of climate fluctuations and thus provide useful simulations for studies of interannual variability.

Save