We have become aware of a calculation error in Aylmer et al. (2020; hereafter A20). For the mean ice thickness, ⟨Hi⟩, and mean heat transport convergences, ho and ha, the area-weighting factor was omitted. This has not had a substantial impact, and the main results and the conclusions of A20 are unaffected.
Figure 2b shows the corrected time series of ⟨Hi⟩ (cf. Fig. 2b of A20). The annual-mean ice thickness was stated to be 1.44 m in the energy balance model (EBM); this should be 1.21 m, which remains a reasonable value. Using the unweighted average in Eq. (13) of A20 amounts to a 1% difference in the estimate of so/sa compared to using the correct average. Testing of Eq. (13) of A20 with different values of BOLR and Bdn (appendix B of A20) still yields estimates of so/sa accurate to within 5% of the (corrected) experimentally derived values.
(b) Area-weighted mean sea ice thickness in the EBM (black, solid), compared to observations (PIOMAS; black, dashed) and the erroneously calculated series without area weighting in A20 (gray, dashed).
Citation: Journal of Climate 34, 13; 10.1175/JCLI-D-20-0914.1
Because the heat transport convergences are roughly independent of latitude at high latitudes, the impacts on the Ka and Fbp sensitivity experiments (Figs. 4 and 5 of A20) are negligible. The sensitivities are affected by a few percent (Table 2).
Updated summary of results, with significantly impacted values (more than 10% error in A20) in bold. For the seasonal case, values obtained when the ice-edge latitude is calculated as a mean only when ice is present (rather than the annual mean) are indicated with an asterisk (*).
The Ko sensitivity experiment (Fig. 3) is moderately affected because increases in ho due to varying Ko are concentrated near the ice edge where the area weighting is greater. The range of variation of ho is about 3 times larger than given in A20, and the seasonal sensitivities are about a factor of 3 smaller. The reduction in Δϕi/Δho between the seasonal and perennial ice cover cases is about a factor of 40 (not 20 as given in A20). We stated that the value of ho required to give a seasonally ice-free solution when varying Fbp was about the same as that when varying Ko—actually, it is about half, which is consistent with our discussion in section 4c paragraph 3. Overall, our qualitative description of the Ko sensitivity analysis holds with the corrected numerical results. Particularly, we concluded that in a seasonally ice-free climate, enhanced ocean heat transport convergence (OHTC) near the ice edge plays a less dramatic role than in a perennial-ice climate, which is (more) consistent with the numbers given here.
Updated results of the Ko sensitivity experiment. Fits are made to the same subset of simulations as in A20. Note that (a) is unaffected.
Citation: Journal of Climate 34, 13; 10.1175/JCLI-D-20-0914.1
REFERENCE
Aylmer, J., D. Ferreira, and D. Feltham, 2020: Impacts of oceanic and atmospheric heat transports on sea ice extent. J. Climate, 33, 7197–7215, https://doi.org/10.1175/JCLI-D-19-0761.1.
