the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Nov 2025
Atmospheric 10Be from Talos Dome (East Antarctic) ice core records geomagnetic dipole intensity from 170 to 270 ka BP
Abstract. We present high-resolution 10Be concentration and flux records from the Talos Dome ice core (East Antarctica), covering the period from 170 to 270 ka BP, to assess the capacity of Antarctic ice cores to capture the dipole moment reductions triggered by geomagnetic excursions of different amplitudes. Three distinct geomagnetic events are identified in the 10Be flux. The dipole collapse linked to the Iceland Basin Excursion (IBE) is clearly recorded as a 10Be peak flux 1.59 to 2.08 times above background between (192.0 ± 1.4) ka BP and (185.6 ± 1.4) ka BP. A clear asymmetric structure is observed, with a rapid decline of the geomagnetic dipole, followed by a three-step recovery. Two dipole decreases of lower amplitude are also resolved in relation with the Pringle Falls Excursion (PFE), lasting from (218.5 ± 1.90) to (206.0 ± 0.8) ka BP, and the Mamaku Excursion (ME), identified at (242.0 ± 0.3) ka BP, both showing an increase of the 10Be flux by a factor of 1.24 to 1.63. A total of 52 short-term 10Be concentration minima were also identified and are consistently associated with peaks in major ion concentrations, indicating post-depositional effects that affect concentration but not the longer-term flux signal. Comparison with Dome Fuji ice core and oceanic authigenic 10Be/9Be records reveals strong agreement in the timing and structure of the dipole moment collapses linked with these excursions. These results further support the use of 10Be for synchronizing ice and marine archives as well as to reconstruct past geomagnetic dipole moment variations and refining age models over the Pleistocene.
- Preprint
(1118 KB) - Metadata XML
-
Supplement
(333 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-5707', Anonymous Referee #1, 30 Dec 2025
-
RC2: 'Comment on egusphere-2025-5707', Kazuho Horiuchi, 05 Jan 2026
Atmospheric 10Be from Talos Dome (East Antarctic) ice core records geomagnetic dipole intensity from 170 to 270 ka BP by Lamothe et al.
This study presents a long-term 10Be record from the Talos Dome ice core (TALDICE) in East Antarctica, spanning a period from 170 to 270 thousand years before present (ka BP). After removing some potential post-depositional "noise", the 3-kyr smoothed 10Be flux profile appears to faithfully represent the broad variations in the atmospheric 10Be production signal. These variations are likely sufficient to serve as a proxy for relative paleointensity (RPI). The smoothed profile shows the 10Be maxima caused by three geomagnetic dipole moment lows: the Iceland Basin excursion (IBE), the Pringles Fall excursion (PFE), and the Mamaku excursion (ME). The characteristic 10Be variations of the largest low, the IBE, are also evident. These maxima and variations could serve as useful tie-points for paleo-archives. Ice core data is especially valuable for the 200–270 ka interval, for which no data has been published yet. Furthermore, I generally agree with their interpretations and conclusions. However, I have the following concerns, which, in my opinion, should be properly addressed before this work is published in GChron.
1. Earlier studies' findings are disregarded
Comprehensive research on the IBE period (170–200 ka BP) has already been conducted by Horiuchi et al. (2016). They presented unprecedented, high-resolution 10Be data from the Antarctic Dome Fuji (DF) ice core and western equatorial Pacific sediments. They also discovered the following: (i) a 7-kyr plateau of the 10Be maximum at the IBE, (ii) a twofold enhancement in 10Be production (i.e. cosmic ray intensity), (iii) an asymmetric pattern of the 10Be peak that is opposite to that of geomagnetic reversals, and (iv) an apparent age offset of several kyr between the ice core and the marine sediments, mainly due to uncertainty in the chronology of the sediments. I found that all of these findings are confirmed using independent data sets by Lamothe et al. in this preprint. This is truly wonderful. However, this preprint does not refer to the earlier findings. It should properly indicate what is known from the earlier research and what new findings were obtained in this study.2. Ocean 10Be records on the IBE
In addition to the aforementioned concern, it is also strange that only the ice core record was picked up from Horuichi et al. (2016). The 10Be records from the two marine sediment cores (KR0515-PC4 and KR0515-PC2) remain the highest-resolution ocean records available for the IBE period. I recommend using these records for comparison, at least to discuss the fine features of the 10Be variations in the IBE period.3. Data processing
The descriptions of the statistical analyses and criteria are sometimes lacking or insufficient. A careless mistake was made when correcting the 10Be flux data of the DF ice core. See the specific comments and questions below for details.Specific comments and questions are as follows:
Lines 162–163. Why is there a sampling gap between 1499 and 1505 m? Please clarify.
Lines 203–204. When the data was merged, did the authors account for the systematic differences (if any) between laboratories? What are the criteria for removing outliers? Please clarify.
Lines 214–234. I agree with the authors’ statement that the uncertainty of the raw 10Be flux is overestimated due to the feature of Paleochrono for SAR estimation, which reflects the large prerequisite uncertainty of SAR. However, I have my doubts about the validity of the uncertainty correction for SAR performed here. According to equation 1, if a given SAR value is equal to the mean of all SAR values, then it has no uncertainty. Is that reasonable? All measurements and estimations are subject to some degree of uncertainty.
Lines 236–238. Correcting the previously published 10Be flux data based on the most recent DF chronology (DF2021) is an excellent attempt. However, this preprint incorrectly uses the DF2 depth for SAR estimation (I verified this by recalculating the updated 10Be flux myself.). Since the DF2021 chronology is associated with the DF1 ice core, the equivalent DF1 depth (see Horiuchi et al.'s (2016) supplementary data file) must be used instead of the DF2 depth. Additionally, it appears that the previous chronology (DFO-2006) is still being used for the age model in this preprint (Figs. 4 and 5). To maintain consistency, I recommend that the authors use the DF2021 chronology for the age model of the corrected 10Be flux of DF. As a result, the r-squared values shown in lines 355–356 (and the relevant discussion?) will change.
Lines 239–241, and 280. If the authors used objective statistical criteria, why did they focus only on low values and not high ones? If the reason is related to concurrent peaks of certain ions, why were low 10Be values unrelated to those peaks excluded from the final rolling average? Please clarify.
Line 263. Does 52 refer to the number of minima or the number of data points included? In Fig. 3, I see 24 shaded lines that I think represent the minima. So, isn’t the number of minima 24? Please specify.
Lines 264–265. Please describe the details about the statistical analysis in either the Materials and Methods or the Supplemental Information.
Figs. 4 and 5. In these figures, the 10Be flux profile of the DF ice core appears to have been smoothed using a 5-point rolling average or a similar procedure. I recommend showing the data at its original resolution. In any case, the authors should clearly indicate in the captions and text that the profile is smoothed.
Lines 333–335. I understand that the extraction protocol is essentially the same for both EDC in Raisbeck et al. (2006) and TALDICE in the present preprint. However, the appearance of 10Be anomalies in the deep parts of the two cores is opposite: increasing spikes in the EDC core and minima in the TALDICE core. The authors argue that the minima may be caused by the incorporation of 10Be into large aggregates that are not released by the extraction protocol. I don't understand how the mechanism keeps only 10Be and releases other major ions for TALDICE only. What is it? I agree with the authors that future studies are needed to clarify this issue. Nevertheless, some trials using a strong leaching method could support the authors' hypothesis and are not difficult to attempt.
Lines 363–365. The 10Be record from the DF ice core over the last millennium (Horiuchi et al., 2008) was normalized using the previous nominal value of the ICN 10Be standard. Additionally, the 10Be flux was calculated using an earlier SAR estimation based on the simple empirical relationship between SAR and d18O in surface snowpacks (Satow et al., 1999) (for more details, see Horiuchi et al., 2008). Then, the average of the last millennium's 10Be flux was updated in Horiuchi et al. (2016) using a revised standard value and the formulation of Parrenin et al. (2007) (i.e. using the same methodology as the published 10Be record for the IBE) (see the Supplementary Material of Horiuchi et al. (2016)). Although it is still just about 1.3 times higher, the updated value of 2.07 ×105 at cm–2 a –1 should be compared to the EDC value.
Lines 368–370. As mentioned above, the DF 10Be flux is not twice as high as the EDC ones, but rather, just 1.3 times higher for the last millennium. Therefore, the difference of two times between the DF and TALDICE is not persistent, but has been observed (so far) only during the IBE. Although this seems enigmatic, I agree with the authors that data from other cores is necessary to resolve this issue.
Lines 383–393. Horiuchi et al. (2016) clearly described the twofold enhancement of 10Be production for IBE by comparing the high-resolution 10Be flux record from DF with the updated 10Be flux for the past millennium (see above). This earlier work should be referenced appropriately here.
Lines 406–417. Horiuchi et al. (2016) clearly pointed out both the asymmetric pattern and its opposite sense to reversals based on the DF record and twin, unprecedented, high-resolution 10Be ocean records from the western equatorial Pacific (This preprint makes no mention of the ocean records, for reasons that are not clear). While it is notable that the authors provide a more comprehensive discussion, they should properly reference the earlier work here.
Lines 418–426. The ≈7-kyr plateau of the IBE cosmogenic anomaly is also an important finding by Horiuchi et al. (2016), follows Knudsen et al.'s (2007) suggestion based on their low-resolution data. Both of these earlier works should be properly cited in this paragraph, along with their original age estimations.
Lines 427–430. When discussing fine-scale features, the high-resolution 10Be ocean records published by Horiuchi et al. (2016) should not be overlooked, as they remain the highest-resolution 10Be records from marine sediments for the IBE interval.
Fig. 5. In the above discussion, what's important is not the value itself, but rather the variations in each index. Therefore, MD05-2930 should have a y-axis value of around 5 to 6 ×10⁸ at the top, meaning a different y-axis should be used for MD05-2920 and MD05-2930. Otherwise, readers will not be able to effectively compare the 10Be profiles. I also recommend adding data points using symbols for lower-resolution data, such as MD05-2920 and MD05-2930, to clarify the time resolution issue mentioned by the authors on lines 485–486.
Line 430 (and Fig. 5) . The criteria for adding the black star in Figure 5 are unclear. A more detailed explanation is necessary to convince audiences of the validity of the tie-point selection.
Lines 481–484. This type of study has already been conducted by Horiuchi et al. (2016) for IBE. At the very least, this earlier work should be referenced here. Furthermore, their ocean records should also be compared with the ice core records.
Lines 485–486. What is the authors' opinion on the clear maximum observed around 232 ka in the MD05-2930 record? Please clarify.
Lines 487–490. Horiuchi et al. (2016) found apparent lags of 3 and 4.5 kyr in the 10Be record of an ice core (DF) compared to two 10Be records from western Pacific sediment cores (KR0515-PC4 and KR0515-PC2) during the IBE period (170–200 ka). They attributed this discrepancy primarily to the uncertainty of the age model for marine sediments. This should be referenced as an earlier study here. Interestingly, an independent examination of this preprint, which used the TALDICE ice core and the MD05-2930 core, yielded a similar result: a lag of 3 kyr. This implies a systematic difference between ice and sediment chronologies (the latter of which essentially relies on marine isotope chronostratigraphy) for this period.
The following are suggestions for technical corrections:
Line 49. Replace "concomittent" with "concomitant."
Lines 140–141. Shouldn't it be 210 and 220 ka BP instead of 200 and 215 ka BP? Around 205 ka, there is a local maximum (minimum) of PRI (10Be) in PISO (MD05-2920/2930).
Line 192. Add ×104 to the mean concentration value.
Line 195. What is “the latter”? Please clarify.
Line 271. I see no data below 1570 m in this preprint. Is it 1520 m instead?
Citation: https://doi.org/10.5194/egusphere-2025-5707-RC2
Data sets
TALDICE 10Be and major ions 1470-1531 m Alexis Lamothe et al. https://doi.org/10.5281/zenodo.17433638
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 369 | 341 | 18 | 728 | 37 | 24 | 29 |
- HTML: 369
- PDF: 341
- XML: 18
- Total: 728
- Supplement: 37
- BibTeX: 24
- EndNote: 29
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The manuscript presents a new ice core record from Talos Dome, Antarctica, covering the period from 170 to 270 ka BP. This 10Be record is highly significant and provides valuable insights into geomagnetic field variations, including one well-defined global excursion and several others of moderate intensity. The study is thorough, containing extensive information on 10Be measurements, as well as climate proxies and ion concentrations, which are used to assess postdepositional effects and accurately estimate the 10Be flux. The resulting signal is consistent with data from the Dome Fuji ice core, sediment archives, and global production reconstructions based on the VADM stack.
It is important to clarify the reference background to which the results are reported. Discuss the possible options in a single section, and rather than reporting multiple values for one event, decide on one reference value. For example, the background, 1.36, 1.44 or 1.56 x10^5 at cm^-2 a^-1, is over which period, mean, or estimated over the running mean, before or after excluding the identified 10Be minima? (Lines 280 to 293, 349 to 350)
Line 80: Several geomagnetic excursions are reported (Figure 1), but the caption reads three. It would be beneficial to provide a table, either in the main text or in the Supplementary Material, listing the records plotted, including their references. Additional columns could indicate which excursion is recorded and, potentially, the corresponding age. I assume the authors already have this information, as the studies are discussed in Section 2.
From a point a view of the geomagnetic field, the authors discussed variations derived from the PISO-1500 RPI stack. A global model of IBE exists, though outdated, and built with limited, hemispherically-constrained sediment records (Lanci et al. 2008, doi: 10.1016/j.pepi.2008.06.004), the dynamics of IBE can be compared with the TALDICE record. As well as with the global stacks of Frank et al., 1997 (https://doi.org/10.1016/S0012-821X(97)00070-8)
General comment on using brackets when reporting ages and standard deviation on the age? Why? If it is the journal rule ok, otherwise the text will be cleaner without, and removing them improves readability in my opinion (throughout the whole text). The same comment applies to the values of the 10Be/9Be ratio, 10Be half-life (e.g., line 187), and all other numbers in brackets.
Minor comments and questions:
Line 30: Add a reference for the first sentence. E.g., Bono et al., 2022 (doi: 10.1029/2022GL100898)
Line 34: the geodynamo
Line 38: I wouldn't say ‘occasionally conflicting’, rather ‘with high uncertainties’
Line 46: Add ‘e.g.’, in front of references, as these are examples.
Line 47: decreases in plural and (recovery) in the brackets in singular. I understood what the sentence says, but it can be clearer.
Line 48: (ref)?
Line 49: another word for migration?
Line 51: We don’t know if the IBE is probably the strongest. By which criteria? It is probably among the strongest.
Line 53: reduction in % with respect to what? Present-day values?
Line 59: Better make two sentences. It is produced in the atmosphere … 10Be mainly results …
Line 59: too many brackets: .. in the stratosphere, from 60% to 66% (refs)
Line 64: remove the comma before that
Line 65: studies = records
Line 73: avoid double brackets, e.g., … events such as the Laschamps, ~41 ka BP (ref) and
Line 74: The Pringle Falls also named Mamaku or Jamaica? As written so far, Pringle Falls and Mamaku are two separate events?
Line 77: add reference for the sedimentary, as it is given for the other record used for comparison from Dome Fuji.
Line 111: Stoner et al., 1998 is not in the References
Line 113: remove ‘on’
Line 125: an RPI
Line 338: PISO-1500
Line 141/142: Reference only once, as it is the same.
Line 155: Section name: ‘Material, methods and chronology’, as the chronology is a Subsection in this Section
Line 165/175: Only suggestion: Can these long links be removed from the main text and added in the Acknowledgment section, where the institutes and facilities are listed anyway again?
Line 167/168: repetition with Line 178 about the cutting/melting.
Line 173: … developed by Raisbeck et al., 2006 and Baroni et al., 2011. Cite without the brackets, and avoid repetition.
Line 177: remove ‘the’ in front of the number of samples. The number is introduced for the first time here.
Line 177: Why is the standard deviation reported like this? If it is important to report that this is one standard deviation, then: 121+-10 g (1 \sigma).
Line 184: in a crucible
Line 187: The ratio is unitless, right? On this line, and the rest of the text
Line 198: then shared for analysis ..., Please clarify whether ‘shared’ means that different measurements were performed in different laboratories, or that the same measurements were carried out on the same samples for inter-laboratory comparison.
Line 200: What is MSA?
Line 209: The AICC2023 chronology (Bouchet et al., 2023) was used to date
Line 210: Make a new sentence ‘The mean snow accumulation is 5.5 cm a-1’, and move it to Line 212 after the range of accumulation rates is listed, and before the sentence starting with ‘On average, …’
Line 233: Maybe not refer to Figure 5 here because the figure comes much later.
Line 240: moving is more commonly used than rolling
Line 234: Why 0.2 in Eq. 1?
Line 235: The paragraph starting here should be moved later when the comparison is discussed
Line 240: the sentence to be clearer, e.g., ‘3 ka rolling average of the standard deviation’. Suggestion: These 10Be minima were identified when the concentration fell below the mean minus one standard deviation, both calculated using a 3 ka moving window. (If this correctly explains how the minima were identified)
Line 280: rolling again? If you decide to change it in the first place, change it accordingly
Line 245: Add the MIS periods mentioned here on top of Fig. 2?
Line 246: Refer to Fig. 2D
Line 281: Testing 1 ka, 3 ka …
Line 284: To avoid figure reference and citation together. Suggestion: ‘The resulting 3-ka averaged 10Be flux record can be compared to geomagnetic reconstructions, including the Dome Fuji ice core data (Horiuchi et al., 2016) in Figure 4, and authigenic 10Be/9Be records from marine sediment cores (Simon et al., 2016), and the 10Be production (Poluianov et al., 2016) calculated from RPI-based VADM (Channell et al., 2009) in Figure 5.
Line 288: too many brackets
Line 289: After the discussion on the baseline, decide on one and report the enhancement factors with respect to the selected one. Maybe also add the enhancement with respect to the present-day values?
Line 340: running median.
Line 340: ‘similar to’ not ‘similar than’. rolling mean average? Suggestion: ‘During glacial periods, the median method results in 10Be fluxes similar to those obtained using a running mean (Figure S2).’
Line 341: I thought the running median and mean are both calculated on the minima identification results, or not? The sentence is confusing
Line 353: … age models, AICC2023 for TALDICE (Bouchet et al., 2023) and the Dome Fuji DFO-2006 chronology (Kawamura et al., 2007).
Line 359: over which period?
Line 363: .. over the last millennium, between 1000 and 1885 CE, where the …; and only 'CE' not 'a CE'
Line 366: remove brackets around Supplementary material, separate with a comma.
Line 367: remove brackets of the citation, separate with a comma.
Line 368: list the solar minima intervals
Line 384: … in an Antarctic ice core over the period from 170 to 270 ka BP.
Line 385-389: needs to make clear that this interpretation is based on 10Be,
Line 386: It is difficult to say ‘slightly higher’ because it depends a lot on the reference background, as the authors already discussed. I agree with ‘comparable’.
Line 407: (Figure 4) not (Figures 2, 4 and 5)
Line 410: mirrors? I would use another verb. ‘Mirrors’ as ‘reflects’ in terms of slow decrease during the M/B reversal and rapid decrease in excursions, or as ‘reproduces, resembles’, the same behavior? My understanding is the first one, but it could be misinterpreted
Line 415-417: Great point!
Line 418: remove the four brackets in this line, and in the paragraph
Line 422, 1) the ages come with large uncertainties, 2) do we expect excursions to occur at the same time globally, considering the geomagnetic field dynamics?
Line 449: in agreement with PISO-1500 variations
Line 450: Isn’t it the case that for excursions, i.e., enhancements, the only variations that should be interpreted are those in the flux, not in the concentration? What does it mean if the others are not expressed in the concentration?
Line 463: interstadial stages of MIS 7 (7.1 for IBE, ..), Line 400: interglacial MIS 7.1? I believe my confusion here comes because I don’t have expertise in paleo(climatology). Please explain: interstadial vs interglacial in terms of short/long, warm/cold, between/within glacial period. I think it is important to know the climate conditions when these excursions happened.
Line 509: add the age of the Mid-Pleistocene Transition
Line 511: Using the VADM reconstruction PISO-1500 (Channell et al., 2009)
Line 520: This study presents a high-resolution 10Be flux record from the East Antarctic TALDICE ice core, covering the interval from 170 to 270 ka BP, and evaluates the reliability of 10Be as a paleomagnetic proxy and as a tool for synchronizing the chronologies of different proxy archives. (or something along these lines, so the three following points are summarized, not only the first two)
Line 545: .. offer great possibility for cross-checking the chronologies ...
Feedback on the figures:
Figure 1: Capital letter in the caption. Expand the caption. The three excursions, which one?
Figure 2 caption: shaded in black? Are these visible? Explain \deltaD. I found the notations for the subfigures in the brackets confusing. It is much clearer to have: A) …. B) …. etc, as it is started, then continue for C), D), and add E). Write the name of the RPI stack. Y-axis is reversed. Note the MIS stages at the top
Figure 3: Capital letter in the caption. Shaded orange and grey, do these overlap? In total, they should be 52, or 52 orange only? Note down in the caption which ones are removed in the analysis? List the major ion concentration (…).
Figure 4: Capital letter in the caption. Which line is purple? .. the snow accumulation rate. Also, rather than the vertical lines showing the x-axis major ticks, add minor x-axis ticks only to the x-axis, and add vertical lines that distinguish the periods of rapid decline and the three-step recovery (and four arrows showing the decline and recoveries). I would suggest adding another panel showing the Laschamps excursion, as there is an extended discussion on it (with similar vertical lines).
Figure 5: Note the IBE, PFE and ME in the figure on top of the grey bars. Denote the subfigures A,B,C... and add these to the main text when referring to the subfigures. The timing of the stars in TALDICE is also different in MD05-2920, but the sediment record wasn’t considered for defining the shifts, right? Are the stars from TALDICE directly applied to the 10Be global production? I found the star around 210 ka very difficult to uniquely identify.