Environmental Engineering Reference
In-Depth Information
of very high equilibrium climate sensitivities cannot be quantified at this
time. We recognize the importance to policy makers of quantitative state-
ments concerning high sensitivities but do not attempt to address this issue
further in this report.
We consider it to be more difficult to provide useful estimates of uncer-
tainty in equilibrium climate sensitivity than in the transient climate response
(Section 3.3), due to the stronger observational constraints on the latter aris-
ing from observations on multidecadal-to-century time scales. In Table 3-1,
we suggest values for the likely (66%) and very likely (90%) ranges for the
transient climate response based on the discussion is Section 3.3
We are cognizant that current climate models have limitations, espe-
cially resulting from deficiencies in cloud simulations and in simulations of
tropical convection. We do not try to delineate these deficiencies here or
relate them to uncertainties in sensitivity. We rely on a general consistency
between the range of sensitivities in the AR4 models and various observa-
tional constraints. The latter are discussed in Section 3.3 as they relate to the
transient climate response and in Section 6.1 on longer time scales.
The equilibrium global mean temperature change results shown in
Table 3.1 were computed by logarithmic extrapolation from the equilibrium
climate sensitivity values reported for the 17 general circulation models
in Table 8.2 of IPCC, Working Group I, The Physical Science Basis (IPCC,
2007a).
3
Over the range of CO
2
covered in the table, logarithmic extrapola-
tion is equivalent to assuming temperature change to be linear in radiative
forcing.
On time scales longer than a few years, the oceanic mixed layer and
the atmosphere warm as a unit. However, heat loss out of the bottom of
the mixed layer keeps the warming below the equilibrium value until the
deep ocean has warmed up and equilibrated. It takes a great deal of energy
to warm the deep ocean, and consequently it takes many centuries for the
temperature to relax to the equilibrium warming. The idealized situation in
which CO
2
is held fixed for many centuries through carefully defined and
ever decreasing emissions is unrealistic, and this might seem to make the
equilibrium sensitivity a concept of mostly academic interest. Nonetheless,
if used properly, the equilibrium climate sensitivity yields important informa-
3
These values were, in turn, obtained from atmosphere/land models coupled to “slab”
ocean models. Slab ocean models equilibrate quickly but assume that there are no changes in
horizontal ocean heat transports as climate changes. The results are usually considered only
rough approximations to the true equilibrium of the coupled system. Danabasogulu and Gent
(2009) have recently provided encouraging results on the accuracy of this approximation in
one particular model.
