the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 17 Feb 2026
An ice core record of volcanic eruptions for the past 4000 years from Dome A, Antarctica
Abstract. Improving the spatial and temporal coverage of volcanic records is essential to accurately quantify volcanic forcing and to provide reliable references for climate models validation. In this study, we present a new volcanic record derived from a 133 m ice core (DA2009) drilled at Dome A, Antarctica. Based on measurements of non-sea-salt sulfate concentrations, 95 volcanic events are identified. Using 15 volcanic age markers aligned with the West Antarctic Ice Sheet (WAIS) Divide ice core (WDC) record, the DA2009 core is dated to cover the past 3951 years, from 1951 BCE to 2000 CE. By comparing the DA2009 record with three Antarctic ice cores from WAIS Divide, Dome C and South Pole, 12 prominent volcanic events are recognized. The period between 1000 and 2000 CE exhibits the most intense volcanic activity of the past 4000 years. The mean snow accumulation rates calculated between adjacent age markers indicate a marked decline in accumulation at Dome A since the 13th century CE. This low-accumulation interval coincides with a pronounced cold phase on the East Antarctic Plateau, suggesting a potential connection between regional climate variability and local accumulation rates at Dome A.
- Preprint
(936 KB) - Metadata XML
-
Supplement
(185 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- CC1: 'Comment on egusphere-2026-769', Alexey Ekaykin, 22 Feb 2026
-
RC1: 'Comment on egusphere-2026-769', Anonymous Referee #1, 06 Apr 2026
An et al. provide a volcanic record for the past 4ka from Dome A. The record is based on sulfate measurements and volcanic matching to the WAIS Divide ice core. The identification of volcanic events follows established methods and seems robust. The dating also appears reasonable. The record is largely similar to other Antarctic ice core records, while raising some intriguing questions about the relative magnitude of deposition around Antarctica.
The main question I had while reading this manuscript is why more effort was not put into matching the volcanic events with Dome C and South Pole (SPICEcore) , in addition to WAIS Divide. Yes, WAIS Divide has the benchmark chronology, but it has been transferred to the volcanic events at both Dome C and South Pole, which are both considerably closer to Dome and more similar in climate characteristics. That, the volcanic matching to Dome C and South Pole should be more robust. Both cores have publicly available sulfate data (https://www.usap-dc.org/view/dataset/601850 and https://www.usap-dc.org/view/dataset/601759).
I would also be interested in seeing the patterns of volcanic events in the sulfate records matched instead of just the inferred event fluxes. Matching of individual events is less compelling, especially when very large events like Tambora can be missing the Dome A record and the 1809 Unknown vent is larger. This raises the possibility of mis-identification, and additional figures and support for the matching of patterns of events would be help establish the specific reliability of the Dome A timescale.
In conclusion, I recommend publication after some revisions that more fully incorporate at least the Dome C and SPICEcore sulfate records.
General Comments
-Provide density data used to convert to an accumulation rate.
Specific comments
L82 – Cole-Dai, 2010 seems like a strange reference for this given how many papers were written before this
L105 – fewer significant figures given measurements precision
L117 – Why use WAIS Divide and not EDC, SPICEcore?
L130 – I don’t understand how this is used to determine the dating uncertainty.
L161 – Table 1 – how do you get the timing to a fraction of the year? Is this from WAIS Dvide? If so, isn’t there a good chance the transport times are enough different to affect the precise dating? Need to define timescale
L175 – what about local volcanic events in the more active West Antarctica? And I don’t think it’s that volcanic deposition is different. I think it’s that the snow accumulation at Dome A can be disrupted by wind scour/redposition
L204 – This is a really weird table. I think it would be a lot easier to list the events and put their ranking. This doesn’t make much sense to me in table form. Is there a way to show it graphically?
L209 – This sentence needs to be reworded
L225 – I’m concerned that the volcanic frequency and flux cannot be used to reliably indicate an increase in the past millennium. For the two higher resolution cores, WDC and SP, the frequency of events does not stand out as being statistically significant. For the flux, there Is evidence of rapid diffusion in the firn, which would lead to older (deeper) events being broader in width, and thus having less sulfate cross the threshold, making them appear smaller. I’m concerned that in the low accumulation cores, DA and EDC, this is a preservation issue rather than a reliable signal. At the least, these conclusions should be tested statistically and issues discussed.
L236 and Figure 5 – the accumulation history is interesting. First, it would be good to show and post the density data and determine the uncertainty. I wonder if the recent decrease is because of poor core recovery leading to densities that are too low. This should be considered and discussed.
Data Availability – I appreciate the data being publicly available. I was able to download it without problem.
Citation: https://doi.org/10.5194/egusphere-2026-769-RC1 -
RC2: 'Comment on egusphere-2026-769', Anonymous Referee #2, 14 May 2026
Summary:
In their manuscript An et al. present a sulfate record for the last ~4000 from Dome A ice core DA2009, Antarctica. They present their methods for discrete sampling and analysis of major ions and volcanic signal detection, identifying 95 volcanic sulfate peaks above the nssS baseline, and use key volcanic markers as tie points to construct the age model for DA2009. The authors then go on to evaluate the volcanic record, classifying detected eruptions as Large, Moderate and Small events based on their flux magnitudes, and comparing the sulfate flux at DA2009 to other ice core volcanic sulfate records across this period from EDC, WDC and SP. From this evaluation they show the majority of Large and Moderate eruptions are tropical, stratospheric eruptions, and therefore likely to have global climate impact.
The record volcanic sulfate events are then used to make inferences about how the frequency and sulfate flux have changed over the last 4000 years, and how this compares at DA2009 to the other considered records. The authors conclude from this that volcanic activity has increased in the last millennia, compared to the previous 3000 years, accompanied by a reduction in snow accumulation rate since 1258 CE, attributed to shifts in regional climate and the onset of the Little Ice Age.
This study provides a very useful additional record of volcanic sulfur flux to Antarctica which is valuable to the community, and fits well with the journal scope. Comparison to other ice core records across the period is a useful addition to the discussion and demonstrates how variable volcanic ice core records of volcanism can be.
I believe that the manuscript would be suitable for publication after some revisions and elaboration of discussion points.
General Comments:
- I’d like to understand why the selected cores were chosen for comparison, as there are more cores which extend beyond the last 2000 years. The paper presenting the previous DA2005 core, which mentioned in the text (Jiang et al., 2012), compares to 6 other Antarctic cores including DML. There are others that extend beyond the CE too - Vostok, Dome Fuji, B53/B54.
- There could be more discussion on the possible differences in signals between the cores. I’ve added suggestions below, but how susceptible to scouring/erosion the site is, and some thought to how this might affect the confidence of volcanic tie-points for age models as well if the ties are made using the size of individual events.
- In terms of the signal attenuation being lower at DA2009 than other cores, how much could the background nssS threshold contribute to the estimated flux of volcanic sulfate to Antarctica? Could the sampling resolution of 3-4 cm give an elevated S “background” concentration, especially if samples are ~1 year? Could the method in which you calculate this be contributing to the overall lower fluxes compared to other cores, as your threshold for detection could be higher?
Line by line comments:
Line 40-41: the authors say an ideal site with “exceptionally flat surface” and “extremely low ice-flow velocity” but then in a later section (line 101) say the “extremely low snow accumulation rate“ prevents annual layer counting. Can it be an ideal paleoclimate site if annual layer counting isn’t possible? Does the very low accumulation rate affect the quality potential of paleoclimate record? Perhaps the language is quite superlative in this case.
Line 143: a little elaboration on why the impacts of stratospheric eruptions greater would be useful (i.e. longer residence time)
Line 147: Again a short elaboration on why tropical stratospheric eruptions have global impact rather than impact being concentrated in one hemisphere may be relevant here.
Line 150-151: These are assumed tropical and southern hemisphere based on their appearance as peaks in both Greenland and Antarctica. Aside from the events you attribute to known eruptions (e.g. 1258 = Samalas is known tropical) we cannot rule out large NH and SH eruptions happening within uncertainty of each other and appearing as tropical signals.
Line 155: is there a need for decimal in 1533.6 BCE?
Line 159: What effect does diffusion of the sulfate in ice have in this core/at this depth? Could this make shorter deposition from a local, tropospheric volcano appear to last 2 years?
Line 170: The authors use the “relative flux metric” to compare the flux of each eruption to Tambora (f/fTambora), as has been performed in previous publications (Cole-Dai et al., 2021). However, in lines 195-200 the authors discuss how Tambora 1815 sulfate in particular greatly varies between ice cores in Antarctica, and how at DA2009 it appears very modest. It doesn’t feature in the top 10 ranked volcanic events at DA2009 but does in all other ice cores compared to, so is comparing each eruption to Tambora and using the relative flux metric appropriate?
Line 196-197: The variability of the Tambora signal at Dome C is discussed in Gautier et al., (2016) Clim. Past., https://doi.org/10.5194/cp-12-103-2016 , where it is sometimes entirely missing from records. It would be good to bring these results into discussion here as well if focussing on 1815/1816 signal variability. In this paper there is also discussion of post-depositional processes that may remove the signal, such as erosion and snow drifting. It would be good to see these post-depositional processes acknowledged in the manuscript too when considering the possibly lower fluxes at DA2009. Comparing the signals at DA2005 and DA2009 may aid this as they are closer in location.
Line 230: Query that you state a mean of 2.37 cm w.e. yr-1but in previous sections you use 2.3. Should this be rounded to 2.4?
Figure 3:
It would be useful to point out the individual volcanic peaks used to tie the DA2009 core the WDC age model. Also showing the S records from other ice cores compared in the study would be a useful addition to robustly demonstrate the age model holds across other sites as well. It would be interesting to see the overlapping period between DA2009, DA2005 (also DA2005 in Fig 2.?) and EDC especially due to similarities in low accumulation rate and the number of signals preserved.
Figure 4:
From visuals of Fig 4., it seems the frequency of eruptions is only significantly higher for 2000-1000 CE in the DA and EDC records. Is there a reason this might be the case? Could it be related to their geographical position and deposition conditions being more similar at EDC and Dome A? Are there uncertainties to include?
Citation: https://doi.org/10.5194/egusphere-2026-769-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,209 | 658 | 90 | 1,957 | 180 | 92 | 209 |
- HTML: 1,209
- PDF: 658
- XML: 90
- Total: 1,957
- Supplement: 180
- BibTeX: 92
- EndNote: 209
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Dear authors, thank you for the manuscript! You might be interested in comparing your results with Vostok data as well: . Sincerely, Alexey