Abstract
Even though five different general circulation models are all currently producing about a 4° ± 1°C warming for doubled CO2, there is still substantial model disagreement about the degree of high latitude amplification of the surface temperature change. The consequences of this disagreement are investigated by comparing doubled CO2 climates with different latitudinal gradients of sea surface temperature. The GISS 4° × 5° general circulation model (GCM) was run with doubled CO2 and two sets of sea surface temperatures: one set derived from the equilibrium doubled CO2 run of the 8° × 10° GISS GCM, with minimal high latitude amplification, and the other set more closely resembling the GFDL results, with greater amplification. While the experiments differ in their latitudinal distribution of warming, they have the same global mean surface air temperature change. The differences in energy balance, atmospheric dynamics and regional climate simulations are discussed.
The results show that the two experiments often produce substantially different climate characteristics. With reduced high latitude amplification, and thus more equatorial warming, there is a greater increase in specific humidity and the greenhouse capacity (the concentration of infrared-absorbing gases) of the atmosphere, resulting in a warmer atmosphere in general. Features such as the low latitude precipitation, Hadley cell intensity, jet stream magnitude and atmospheric energy transports all increase compared to the control run. In contrast, these features all decrease in the experiment with greater high latitude amplification. There are also significant differences in the cloud cover and stationary eddy energy responses between the two experiments, as well as most regional climate changes; for example, there is greater drying of the midlatitude summer continents and greater polar ice melting when the high latitude amplification is greater. Predictions of the coming doubled CO2 climate and its societal consequences must be tempered by the current uncertainty in the degree of high latitude amplification.
