In de Chalendar, Garing & Benson (Reference de Chalendar, Garing and Benson2018), an incorrect value (expressed in kPa rather than Pa) was used for Henry’s constant when generating the simulations for figures 7 and 8. Using an appropriate value for Henry’s constant increases the times shown in those figures by three orders of magnitude ( $10^{8}$  s rather than $10^{5}$  s in figure 7, for example). In consequence, the time scale for equilibrium is now estimated as needing years rather than days.