Abstract
We analyze the publicly released outputs of the simulations performedby
climate models (CMs) in preindustrial (PI) and Special Reporton Emissions
Scenarios A1B (SRESA1B) conditions. In the PI simulations,most CMs
feature biases of the order of 1 W m−2 for the net globaland the
net atmospheric, oceanic, and land energy balances. Thisdoes not
result from transient effects but depends on the imperfectclosure
of the energy cycle in the fluid components and on inconsistenciesover
land. Thus, the planetary emission temperature is underestimated,which
may explain the CMs' cold bias. In the PI scenario, CMs agreeon the
meridional atmospheric enthalpy transport's peak location(around
40\,^∘N/S), while discrepancies of ∼20% exist on the intensity.Disagreements
on the oceanic transport peaks' location and intensityamount to ∼10\,^∘
and ∼50%, respectively. In the SRESA1B runs, the atmospherictransport's
peak shifts poleward, and its intensity increases upto ∼10% in both
hemispheres. In most CMs, the Northern Hemisphericoceanic transport
decreases, and the peaks shift equatorward in bothhemispheres. The
Bjerknes compensation mechanism is active both onclimatological and
interannual time scales. The total meridionaltransport peaks around
35\,^∘ in both hemispheres and scenarios, whereasdisagreements
on the intensity reach ∼20%. With increased CO2 concentration,the
total transport increases up to ∼10%, thus contributing to polaramplification
of global warming. Advances are needed for achievinga self-consistent
representation of climate as a nonequilibrium thermodynamicalsystem.
This is crucial for improving the CMs' skill in representingpast
and future climate changes.
