| Abstract: The transient responses of two versions of the Geophysical Fluid Dynamics Laboratory (GFDL) coupled climate model to a climate change forcing scenario are examined. The same computer codes were used to construct the atmosphere, ocean, sea ice and land surface components of the two models, and they employ the same types of sub-grid-scale parameterization schemes. The two model versions differ primarily, but not solely, in their spatial resolution. Comparisons are made of results from six coarse-resolution R15 climate change experiments and three medium-resolution R30 experiments in which levels of greenhouse gases (GHGs) and sulfate aerosols are specified to change over time. The two model versions yield similar global mean surface air temperature responses until the second half of the 21st
century, after which the R15 model exhibits a somewhat larger response.
Polar amplification of the Northern Hemisphere's warming signal is more
pronounced in the R15 model, in part due to the R15's cooler control
climate, which allows for larger snow and ice albedo positive feedbacks.
Both models project a substantial weakening of the North Atlantic
overturning circulation and a large reduction in the volume of Arctic
sea ice to occur in the 21st century. Relative to their respective
control integrations, there is a greater reduction of Arctic sea ice in
the R15 experiments than in the R30 simulations as the climate system
warms. The globally averaged annual mean precipitation rate is simulated
to increase over time, with both model versions projecting an increase
of about 8% to occur by the decade of the 2080s. While the global mean
precipitation response is quite similar in the two models, regional
differences exist, with the R30 model displaying larger increases in
equatorial regions. |