B. Timbal et al., SENSITIVITY TO PRESCRIBED CHANGES IN SEA-SURFACE TEMPERATURE AND SEA-ICE IN DOUBLED CARBON-DIOXIDE EXPERIMENTS, Climate dynamics, 12(1), 1995, pp. 1-20
Time sclice experiments are performed with the atmospheric GCM ARPEGE,
developed at Meteo-France, to study the impact to increases in the at
mospheric carbon dioxide. This spectral model runs at T42 horizontal r
esolution with 30 vertical layers including a comprehensive tropospher
ic and stratospheric resolution and a prognostic parameterization of t
he ozone mixing ratio. The model is forced in a 5-year control run by
climatological SSTs and sea-ice extents in order to obtain an accurate
simulation of the present-day climate. Two perturbed runs are perform
ed using SSTs and sea-ice extents for doubled CO2 concentration, obtai
ned from transient runs performed by two coupled atmospheric-oceanic m
odels run at the Max Planck Institute (MPI) in Hamburg and the Hadley
Centre (HC). A global surface temperature warming of 1.6 K is obtained
with the MPI SST anomalies and 1.9 K with the HC SST anomalies. The p
recipitation rate increases by 4.2% (and 4.7%). The features obtained
in the stratosphere (a cooling increasing with the altitude and an inc
rease in the ozone mixing ratio) are not sensitive to the oceanic forc
ing. On the contrary, the ano malies in the troposphere such as a warm
ing increasing with altitude, an acceleration of westerly jets and a r
aised cloud height, depend on the oceanic forcing imposed in the two p
erturbed runs. Special attention is given to continental areas where t
he impact of the oceanic forcing is studied over eight regions around
the globe. Regions sensitive to oceanic forcing such as Europe are ide
ntified in contrast with areas where the patterns are driven by land-s
urface physical processes, such as over continental Asia. Finally, the
Koppen classification is applied to the climate simulated in the thre
e experiments. Both doubled CO2 runs show the same predominance of glo
bal warming over precipitation changes in the Koppen analyses.