Peltier & Liu reply

Replying to: P. F. Hoffman, J. W. Crowley, D. T. Johnston, D. S. Jones & D. P. Schrag Nature 456, 10.1038/nature07655 (2008) ; Y. Goddéris & Y. Donnadieu Nature 456, 10.1038/nature07653 (2008) The first of several questionable assertions in the Comment of Hoffman et al. https://www.nature.com/articles/nature07653 1 on our paper2 is that our results are “sensitive to initial conditions and model parameters”. This references an undergraduate report3 in which this issue was not addressed. Figure 1 illustrates a sequence of trajectories of the nonlinear system. These are ‘attracted to’ the set of steady state solutions such that, following transient adjustment, the system locks-on to the oscillatory ‘slushball’ solution unless the initial CO2 concentration is extremely low or F21 , the control variable of the model that determines the sensitivity of the rate of remineralization to temperature changes, is supercritical. This first assertion is therefore misleading. Figure 1 Trajectories of the solution for the coupled carbon cycle and climate model started from different initial conditions, indicated by filled circles of different colours. The dotted line represents the set of steady state solutions of the climate model; this set includes the hysteresis loop on the cold branch of which the ‘slushball’ forms. The parameter drad denotes the increase in infrared radiative forcing at the surface of the Earth due to the increasing concentration of carbon dioxide in the overlying atmosphere. The variable Tsurf is the annual mean surface temperature of the planet. PowerPoint slide