Sources of low-frequency variability in observed Antarctic sea ice
Abstract. Antarctic sea ice gradually increased from the late 1970s until 2016, when it experienced an abrupt decline. A number of mechanisms have been proposed for both the gradual increase and abrupt decline of Antarctic sea ice, but how each mechanism manifests spatially and temporally remains poorly understood. Here, we use a statistical method called low-frequency component analysis to analyze the spatial-temporal structure of observed Antarctic sea-ice concentration variability. The identified patterns reveal distinct modes of low-frequency sea ice variability. The leading mode, which accounts for the large-scale, gradual expansion of sea ice, is associated with the Interdecadal Pacific Oscillation and resembles the observed sea-surface temperature trend pattern that climate models have trouble reproducing. The second mode is associated with the central Pacific El Niño–Southern Oscillation (ENSO) and the Southern Annular Mode, and accounts for most of the sea ice variability in the Ross Sea. The third mode is associated with the eastern Pacific ENSO and Amundsen Sea Low, and accounts for most of the pan-Antarctic sea-ice variability and almost all of the sea ice variability in the Weddell Sea. This mode is associated with periods of abrupt Antarctic sea-ice decline and is related to a weakening of the circumpolar westerlies, which favors surface warming through a shoaling of the ocean mixed layer and decreased northward Ekman heat convergence. Broadly, these results suggest that climate model biases in long-term Antarctic sea-ice and global sea-surface temperature trends are related to each other and that eastern Pacific ENSO variability causes abrupt sea ice changes.
David B. Bonan et al.
Status: open (until 07 Jul 2023)
- RC1: 'Comment on egusphere-2023-750', Anonymous Referee #1, 03 May 2023 reply
David B. Bonan et al.
David B. Bonan et al.
Viewed (geographical distribution)
This is an interesting manuscript that applies a novel methodology (LFCA) in the context of Antarctic sea-ice analysis. It provides results that confirm earlier ones, such as the influence by IPO and ENSO, and new results, such as the co-variability of LFC modes and identifies a mode that climate models struggle to reproduce, a finding that is particularly useful to climate modellers.
One potentially major issue is the robustness of the findings. The current analysis is based on one sea-ice concentration data from NSIDC, one atmospheric
reanalysis (ERA5), and one ocean reanalysis (ORAS5). It is known that the Antarctic sea-ice concentration data has significant uncertainties (https://doi.org/10.1029/2011GL047553), as do reanalyses (https://doi.org/10.1016/j.ocemod.2023.102183 and https://doi.org/10.1007/s00382-018-4242-z). It would be important to estimate how much these uncertainties affect the findings of this study.
Other than the aforementioned issue, some minor comments and suggestions could be considered as listed next
in no particular order if importance:
- line 30. The abrupt decline still persists as February 2023 SIA was record low.
- line 73. As LFCA is based on PCA it is a statistical method and can not determine physical and dynamical processes causing sea-ice variability.
- lines 116-121. Could be further clarified how the explained SIC variance byt LFPs compare with explained variance of LFP SIA time series. For example, why LFP1 has the highest variance ratio, while LFP3/LFC3 explain the highest proportion of pan-Arctic SIA variance.
- line 135. Did you try other than 10-year cutoff filter and 5 leading EOFs?
- line 149. LFC3 during the abrupt sea-ice decline coincides with low LFC1 values which could perhaps mentioned.
- line 189. '... contributions from each LFC to SIA separately ...'
- line 200. Some of these trends look rather small in Figure 5, e.g. in 5h. Are they statistically significant and therefore worth of discussing?
- In some figures, SIC anomaly colourbars are from blue to red (e.g. Fig. 2), in some others red to blue (e.g. Fig 1.). Would be reader friendly to have similar colourbars in all SIC anomaly plots.
- line 251. Stronger winds in the Weddell and Scotia Seas are not apparent in Fig. 6c.
- lines 253-254. Also LFC2 wind patterns would cause strong Ekman pumping and suctions but that is not discussed.
- line 255. Would be useful to discuss how LFC3 effects are transmitted from the Pacific sector to the Weddell and Scotia Seas.
- line 265. 'much higher' seems exaggeration, 'higher' is better.
- line 266. '... the Southern Ocean than LFC1-4.'
- line 284. The event2 positive sea-ice concentration anomaly in the Amundsen Bellingshausen Seas is almost as strong and somewhat more extensive than during event1.
- line 284. 'Amundensen' -> 'Amundsen'.
- line 300. Would be good to mention that the positive flux is directed downward.
- line 315. '... geostrophic velocities and mesoscale eddies.'
- line 326. Term 'anomalous southward Ekman transport' is misleading because the Ekman transport is northward directed. 'Weaker northward Ekman transport' would be better.
- line 345. A novel statistical method is not 22 year old. It is novel in the context of Antarctic sea-ice analysis.
- line 346. '...patterns and distinct modes of...'
- line 347. Delete sentence 'We identiﬁed distinct modes of low-frequency Antarctic sea-ice variability'.
- line 363. 'Southern ocean' -> 'Southern Ocean'.
- line 374. Fig. 6c shows rather modest looking surface warming, while Figure 2c shows remarkable sea-ice loss. Note that sea-ice loss may not necessarily be associated with strong SST increases as polar ocean surface temperatures tend to stay close to the freezing point.
- line 398. Mention here that ORAS5 was used.