Abstract. In the following, we test the validity of a one-box climate model as an
emulator for atmosphere–ocean general circulation models (AOGCMs). The one-box
climate model is currently employed in the integrated assessment models FUND,
MIND, and PAGE, widely used in policy making. Our findings are twofold.
Firstly, when directly prescribing AOGCMs' respective equilibrium climate
sensitivities (ECSs) and transient climate responses (TCRs) to the one-box
model, global mean temperature (GMT) projections are generically too high by
0.5 K at peak temperature for peak-and-decline forcing scenarios, resulting
in a maximum global warming of approximately 2 K. Accordingly, corresponding
integrated assessment studies might tend to overestimate mitigation needs and
costs. We semi-analytically explain this discrepancy as resulting from the
information loss resulting from the reduction of complexity. Secondly, the
one-box model offers a good emulator of these AOGCMs (accurate to within 0.1 K
for Representative Concentration Pathways, RCPs, namely RCP2.6, RCP4.5, and
RCP6.0), provided the AOGCM's ECS and TCR values are universally mapped onto
effective one-box counterparts and a certain time horizon (on the order of the
time to peak radiative forcing) is not exceeded. Results that are based on
the one-box model and have already been published are still just as informative
as intended by their respective authors; however, they should be
reinterpreted as being influenced by a larger climate response to forcing
than intended.