| 26 Mar 2026
Investigating the relationship between Total Air Content (TAC) variations in polar ice cores and local surface climate conditions
Abstract. While air bubbles in polar ice cores are well known for archiving past atmospheric composition, the quantity of air trapped in these bubbles also carries significant paleoclimatic information. This parameter, known as Total Air Content (TAC), was first used to reconstruct past ice sheet elevation and later became an orbital dating tool after its temporal variations revealed a strong correlation with local summer insolation. However, limited understanding of relationships between TAC, pore volume at close-off depth, and surface parameters has restricted its widespread use. In this study, we investigate the link between surface parameters (e.g. temperature and accumulation) and changes in TAC at both spatial and temporal scales in order to better understand TAC as an environmental proxy and as an orbital dating tool. To do so, we first present a new dataset extending the TAC record from the EPICA Dome C (EDC) ice core back to 800 ka, as well as new TAC records from the TALDICE and EDML ice cores covering the last glacial-interglacial cycle. Second, we combine these new datasets with a compilation of published TAC data from ice cores across Antarctica and Greenland to explore the influence of surface climate parameters controlling the changes in TAC both at spatial and temporal scales. Our spatial-scale analysis first examines how recent TAC values relate to atmospheric pressure and elevation. We then investigate pore volume at close-off (i.e. TAC values corrected for ideal gas law effects) to assess the influence of mean annual temperature, accumulation rate, and local summer insolation. We evidence a strong correlation between the pore volume at close-off and local half year summer insolation for East Antarctic sites, suggesting a direct control of local insolation on firn densification in this region. Temporal-scale analyses on TAC records covering at least 40 ka confirm that TAC records contain an orbital-scale signature of local insolation but also show that local summer insolation alone cannot capture the full TAC variability. Multiple regression analysis incorporating local insolation and reconstructed surface temperatures or accumulation better predicts TAC temporal changes, particularly during glacial terminations. Our new EDC high-resolution record also revealed significant millennial-scale TAC changes during these glacial terminations. Hence, our results highlight that in addition to the orbital-scale impact of local summer insolation, millennial-to-multi-millennial-scale changes in surface climate parameters also influence the temporal-scale TAC changes. Our findings suggest that orbital tuning between TAC and local insolation that neglects surface climate controls could introduce age uncertainties of 1–5 ka, calling for surface climate-related corrections prior to TAC-based orbital dating.
The authors present an extended total air content record for EPICA Dome C (EDC) as well as investigate other previously published TAC datasets to understand different climate parameters’ influence on TAC. The authors indicate that the records can be used for orbital tuning but only after considering local climatic influences on the TAC record. Overall, very well written and thorough manuscript that does an excellent job summarizing not only the current state of understanding of TAC but also explores new ways to consider TAC in different climate instances. I recommend the manuscript be published after considering the following minor comments:
Page 4, line 96: “...(Clausius-Clapeyron)” is a reference that makes sense here, but should be written out. A quick sentence about how the CC relationship relates here or even adding the relationship would add to the manuscript.
Pg 4, ln 99: Use of parenthesis back to back made the sentence difficult to read. Recommend rewording it.
Pg 8, ln 160 and pg 20 ln 399: Typo - Remove period inside parenthesis of “(Equation 1.).”. Actually, lots of extra periods inside parenthesis when referring to figures and tables.
Pg 11, ln 253: Remove quotes from “Densification age scale”. Quotations are unnecessary. Also related, on page 12, ln 273, you then refer to the same concept as a ‘densification integrated’ age-scale. Recommend to pick one label for the concept. However, I’m glad to see an in depth discussion of this issue, and appreciate the need for a densification age scale to be developed.
Page 14, figure 2 – The newly measured TAC in figure 2a seems to be much higher resolution than older data. In section 3.1 you compare the resolution of the Bern and IGE datasets. What is the resolution comparison between the older and newer IGE datasets? I don’t suspect it does, but does this change the analysis in any way?
Pg 16, figure 4b – the ‘HYSI+1 and HYSI +2’ labels are a bit confusing. I spent a while thinking that the HYSI +2 was a multiple regression using HYSI, temperature AND accumulation, while I see now that it is HYSI and accumulation. Consider different labels, or being more clear in the caption.
Pg 16, figure 4b – why is there no data for DF HYSI +2? There is data in the single regression for both temperature and accumulation, so I would expect the mulitple regression to have some solution.
Pg 18, ln 378: No need for a new paragraph after first sentence here.
Pg 18, paragraph 3: This paragraph places an emphasis on the multiple regressions of HYSI + temperature. However, I also see in figure 4b you did multiple regressions with both temperature and accumulation (though not in the same regression). Did you do multiple regressions with both temperature and accumulation? If not, why? Also, some discussion on how accumulation changes the strength of the regressions would be helpful.
Page 21, Figure 7 caption: confusing wording after “b. Vc values and Present-day temperature c. Present-day accumulation”. Maybe move the reference to Vc before the ‘b.” Since it makes the following insets (c-e) seem like they are not compared to anything. I had to read the caption a couple times to understand it.
Page 22, line 428: “In line with the observations from Fig 7.e…” was a bit confusing because you then go on to describe how there is not correlation at the EAIS sites between HYSI and Vcr, when fig 7e definitely shows a correlation (though its between HYSI and Vc). Maybe something like “different from the observations in Fig 7.e” would be more clear.
Page 23, figure 8: A bit of confusion on labels with this figure. Change the WAIS ice cores to a different shape to more easily see them, like you used squares for Greenland ice cores. The use of different shades of greens for three ice cores on two different regional correlations makes it difficult to see which ice core is which. On a different note, the WAIS ice core should be called WD (to avoid confusion with the West Antarctic Ice Sheet). What are the dark grey diamond shapes? They don’t seem to appear in figure 8a, but do appear in figure 8b.
Related to Figure 8 - Did you look at a correlation between accumulation and HYSI? It may also highlight the regional difference in TAC behavior. The R2 for temperature/HYSI in Greenland cores is 0.00, but might be different for accumulation, further strengthening your conclusion that climate-regime needs to be considered when parameterizing TAC.
Pg 26, line 518: Don’t use ‘both’ and then describe three parameters. Use the word ‘all’ or restructure the sentence.