Modeling the Surface Temperature Changes Due to Increased Atmospheric CO₂

Abstract

Although there are considerable differences in atmospheric models’ structure and sensitivity, there is a consensus that increasing the concentration of CO₂ in the atmosphere will lead to an overall warming. The more simplified models (which treat the atmosphere as a vertical column in which the temperature is determined by the net absorption of radiation and the redistribution of heat by convection) generally yield a surface temperature rise of several degree C for a doubling of atmospheric CO₂. Other simplified models built around the meridional variation of the surface heat balance (in which the surface radiative, sensible and latent heat fluxes are parameterized in terms of the surface temperature) also yield an average warming of several degree C for a CO₂ doubling. Such heat balance models are particularly sensitive to the feedback between the surface temperature and the albedo of snow- or ice-covered surfaces, and generally show a poleward amplification of the CO₂-induced warming when this feedback is included.

General circulation models or GCMs (in which the three-dimensional distribution of global climate is simulated in response to a wide variety of interacting time-dependent processes) generally confirm the simpler models’ warming, and serve to emphasize the climate’s sensitivity to the treatment of the oceans. Using idealized geography and a motionless swamp-like ocean of zero heat capacity, GCMs yield a global average warming from about 3°C in the tropics to about 6°C in high latitudes for a doubling of CO₂ under annual average insolation. The importance of the swamp ocean in these calculations is shown by a model with realistic geography and an ocean whose surface temperature is constrained to follow the normal climatological seasonal variation. Such a GCM yields only about 0.2°C overall warming for doubled CO₂, with significantly larger warming over the higher latitude continents in summer. When an atmospheric GCM and a shallow mixed-layer ocean are used with realistic geography and seasonal insolation, a more realistic average warming of about 2°C is inferred for doubled CO₂, with ice albedo-temperature feedback responsible for a several-fold amplification in higher latitudes in winter. This value is regarded as the best estimate of the average surface warming due to doubled CO₂ now available.

To determine the seasonal and regional distribution of a CO₂-induced warming as required for the reliable estimation of energy-related climate impacts, further model simulations are required in which the natural interannual variability of the atmosphere and surface ocean are more adequately represented. These factors, together with the time-dependent response of the surface biomass and of the deeper ocean, will determine the course of climate as a result of future CO₂ increases.