the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Apr 2026
Exploring the Cenozoic Earth system with extraterrestrial 3He
Abstract. A key method for contextualizing climate today and into the future is to draw upon the past. Marine sediments accumulating at the bottom of the ocean serve as the only continuous archive of Earth's climate history spanning tens of millions of years. From the earliest studies of marine sediments, reconstructed changes in the inputs of sedimentary constituents such as volcanic glass, ice-rafted debris, particulates carried by rivers, desert dust, organic and inorganic remnants of organisms and biological process, and even extraterrestrial material, have been used to discern past environmental conditions. Specifically, it is the rate at which these different components of sediments from the bottom of the ocean accrue over time that can provide unique and important insight into how Earth system dynamics operated in the near and distant past. Traditional techniques for quantifying sediment mass accumulation rates (MARs) contain numerous complexities that can lead to erroneous MAR determinations. Constant flux proxies (CFPs), particularly extraterrestrial 3He (3HeET) delivered in interplanetary dust particles from space, are underutilized geochemical tools that alleviate these issues as long as key assumptions can be constrained. In this perspective, we review CFP 3HeET as a CFP and outline its benefit for understanding Cenozoic climate beyond the Quaternary. Ultimately, we suggest that the application of 3HeET is valuable for providing the broader paleoclimate and paleoceanography communities with accurate records of Cenozoic Earth system change from a MAR perspective.
- Preprint
(1415 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- CC1: 'Comment on egusphere-2026-1583', Giacomo Medici, 16 Apr 2026
-
RC1: 'Comment on egusphere-2026-1583', David McGee, 13 May 2026
This manuscript presents a helpful and clear review of the use of extraterrestrial helium-3 (3HeET) to calculate vertical rain rates of marine sediments, and it argues that applications of this tool have been limited relative to its potential for providing insight into past climate and marine productivity changes.
The writing is thoughtful and well-organized, and the manuscript presents two helpful case studies documenting how mass accumulation rates (MARs) estimated from 3HeET lead to very different results than previous work relying on age model-based MARs. The authors make a good argument that 3HeET should be more widely used in the field, and the manuscript is worthy of publication with a few revisions.
Much of the manuscript is essentially an updated version of section 4.2 of McGee and Mukhopadhyay (2013); that work also noted the advantages of 3HeET-based MARs over age model-based MARs and highlighted examples where use of 3HeET substantially changes conclusions, and it ended with a similar argument: “the use of 3HeET as a constant flux proxy has been relatively limited compared to its potential applications.” That is not to say that this overlap makes the current manuscript unnecessary: reviews need updating every decade or so, and it is worthwhile to have such a review appear in a journal rather than a (rather obscure) book. Moreover, this manuscript is much more specifically addressed to the paleoceanography community.
I have a few suggestions for improvements:
Line Comment
67-69 This sentence is confusing.
76 “loss”->”lost”
78 I suggest changing “at least from a MAR perspective” to “at least those based upon MARs”
93 “Ee”->”We”
104 I suggest changing “from an age model perspective” to “based on age models”
Fig 1 I was unclear about the meaning of this figure. The text doesn’t explain how the higher-resolution age model was determined, beyond mentioning a “statistical tuning approach” in the caption. I guess the point is that there can be substantial MAR changes between low-resolution age model tie points, but a) this could be said more clearly, and b) the high-resolution MAR doesn’t appear to average to the same value as the low-resolution MAR over the same intervals (e.g. see the interval from the coretop to ~15 ka), which is confusing.
134 Equations 3 and 4 seem unnecessary and misleading – Equation 3 states an assumption, and Equation 4 states something that is usually but not always true. I think equations should be saved for definitions or relationships that are always true.
185 Change “thought of as the ‘excess’ or ‘missing’ sediment” to “thought of as the ratio of the total sediment”
225 “works”->”work”
226 “heavy particle scavenging” is confusing to me. Are you talking about scavenging by heavy particles, or about the extent to which vertical fluxes are impacted by scavenging?
226 I don’t find the age model-dependence argument to be a reason for favoring 3HeET over 230Th. My sense is that in most sediment core records of the last ~400-500 ka (i.e. where 230Th is an option), the age model-based uncertainty is only a small fraction of the total uncertainty in 230Th-based MARs; and if the age model is so uncertain that it is a dominant part of the 230Th uncertainty, then your flux time series is not going to be very useful for paleoceanographic purposes anyway.
Fig 4 Why is Fig 4A the same as Fig 3C?
Fig 4 “Gubbio” is misspelled in two places.
285 Perhaps clearer to replace “LSRs” with “age models”?
288 can you be more specific about what is meant by “certain sedimentation patterns”? In general, this section could be more convincing in making the case that the 3HeET-based MARs at 885/886 are more accurate than age model-based MARs, not just that they’re different.
303 It’s unclear whether “contemporaneous” refers to the fact that the studies came out at around the same time, or whether they covered the same time interval.
318 Can you be clearer about what is meant by “core-specific characteristics”? I think it would be helpful to state specifically that the difference between the two MARs suggests that the sites are affected by substantial winnowing prior to 3.6 Ma, and that the large inferred increase in dust flux after 3.6 Ma instead reflects a decrease in winnowing/increase in focusing.
Fig 6A I’m not sure where it would go in the manuscript, but I would encourage the authors to emphasize more than they currently do that the age model-based MARs for specific sedimentary constituents are in reality limited to the resolution of the age model; age model-based dust and opal flux records that appear to be high resolution (like those shown in Fig 6A and 6C) reflect an assumption that all changes in dust and opal concentration are from changes in flux rather than changes in dilution.
340 The use of the word “driver” here makes it sound like you are talking about a driver of the PETM, not a driver of the recovery from the PETM.
368 “be”->”can be”
371 “less-than-optimal”->”low”
381 It would be helpful to point out that this sensitivity of the 3HeET concentration leads to large uncertainties in the 3HeET-based fluxes. In general, I think the authors can be clearer in pointing out that use of 3HeET might not be possible in quickly accumulating sediments with high detrital content.
422 Somewhere, the authors should point out that the inferred 3HeET fluxes for older time intervals may reflect site-specific differences in 3HeET loss (e.g. due to burial heating). For example, the data from Gubbio are from lithified, uplifted sediments that may have been heated.
438 I’m not clear that the Elderfield proxy curve applies to 3HeET…did 3HeET go through a “Pessimism” phase? If so, the authors should provide evidence of this phase – maybe it was during the time when people were suggesting cyclic 100-kyr changes in 3HeET flux to Earth? I feel like (as the authors later state), the real issues are lack of analytical facilities, the slow pace of data gathering (due to analytical procedures and the need for replicates), and the continued need for community education about the importance of CFPs (which this manuscript aims to address).
449 I am conflicted about writing this, but for #2 I guess feel that the McGee and Mukhopadhyay chapter also tried to “describ[e] 3HeET’s utility to the broader Earth science community”? Again, it’s clear this manuscript has an important role to play, but I think the authors could be a bit more specific in stating how this manuscript is different from previous work.
460 to me, “heavily stymied” is a bit of an awkward phrase.
Nice work, and I look forward to seeing this published.
-David McGee
Citation: https://doi.org/10.5194/egusphere-2026-1583-RC1 -
RC2: 'Comment on egusphere-2026-1583', Anonymous Referee #2, 05 Jul 2026
This perspective provides an overview of the potential of extraterrestrial helium-3 (3HeET) as a tool for estimating mass accumulation rates (MAR) in deep-sea sediments. The authors compare the 3HeET-based method with other approaches, including age-model-based techniques and 230Th-based estimates, highlighting both strengths and limitations. The paper makes a compelling case for broader adoption and further exploration of 3HeET in paleoceanographic studies. As someone working with conventional (i.e., age model-based) MAR reconstructions but lacks direct experience with 3HeET measurements, I offer the following comments aimed at improving clarity—particularly for readers with a similar background.
main points:
A key assumption underpinning the use of 3HeET as a constant-flux proxy (CFP) is the temporal constancy of its delivery to Earth. While the manuscript acknowledges this assumption, it would benefit from a more critical evaluation of its validity across different timescales. For example, it would be helpful if the authors expanded on: (i) whether the 3HeET flux can be considered constant over the last glacial cycle (e.g., since the LGM), (ii) potential long-term trends across tens to hundreds of thousands of years, or even millions of years, and (iii) any observational evidence that supports or challenges these assumptions.Figures 5-6 present compelling data showing divergent trends between vMAR derived from 3HeET and bulk MAR (bMAR) estimates. Notably, 3HeET-based vMAR is several times higher than bMAR during the late Miocene and early Pliocene. For readers unfamiliar with the site-specific context, this discrepancy raises important interpretive questions: does this discrepancy also reflect a fundamentally different age model in the 3HeET approach compared to the one used for bMAR? Alternatively, if the age models are consistent, could the higher vMAR inferred from 3HeET indicate enhanced winnowing during the late Miocene and early Pliocene when surface productivity was also higher? While the authors reference prior studies, a brief synthesis of the proposed mechanisms for this divergence would enhance clarity.
In Section 6, the authors discuss several caveats related to the use of 3HeET as a CFP. However, the role of post-deposional dissolution of biogenic material may warrant further attention. The dissolution of biogenic calcareous and siliceous remains occurs widely across the seafloor and varies spatially with depth, ocean chemistry, and circulation, as well as temporally with climate and oceanographic changes. Such processes likely alter the relative proportions of sediment components and thus affect the average concentration of the proxy [CFPavg]. How might this influence the estimation of vMAR?
Minor Comments
Line 93: Typo: “Ee aim to achieve”
Line 211: Consider defining [3HeET] and its unit
Line 214: The unit “pcc” should be defined at first use
Line 250: I do not see “circles with dashed borders” in the figure.
Line 288: “Early work suggested that certain sedimentation patterns…” the term “certain patterns” is vague. Please specify which patterns are being referenced.Citation: https://doi.org/10.5194/egusphere-2026-1583-RC2 -
RC3: 'Comment on egusphere-2026-1583', Pierre-Henri Blard, 12 Jul 2026
This manuscript by Abell et al. is somewhat of a hybrid between a review article and a perspective (viewpoint) paper. Because it is intended for a broad readership, including researchers who are not specialists in extraterrestrial 3He and 230Th, I believe it is a valuable contribution. It provides a clear and pedagogical overview of both the potential and the limitations of these methods. I recommend publication after minor to moderate revisions, taking into account the following comments.
Main comments
- Recommendation to use the Quaternary F_3HeETF as a default flux (Section 6)
I agree that there are clear spatial and temporal variations in extraterrestrial 3He flux F_3HeETF throughout the Cenozoic. However, the range of fluxes reported for the Cenozoic remains substantially lower than the Quaternary value. Therefore, I do not think that recommending the Quaternary flux as the default choice is appropriate.
In this context, I would like to draw the authors' attention to our recent work on Mesozoic marl-limestone alternations (Blard et al., EPSL, 2023). In that study, the total sedimentation rates (carbonate + silicate) reconstructed from extraterrestrial 3He are in excellent agreement with independent biostratigraphic sedimentation rates when using the Cretaceous extraterrestrial flux independently determined by Farley et al. 2012 (~0.1 pcc cm⁻² kyr⁻¹). If the Quaternary flux, which is approximately an order of magnitude higher, had been used instead, the reconstructed sedimentation rates would also have been overestimated by roughly one order of magnitude.
Given that the average F_3HeETF is approximately 0.5 pcc cm⁻² kyr⁻¹ throughout the Cenozoic, compared with about 1 pcc cm⁻² kyr⁻¹ during the Quaternary, available data suggest that the extraterrestrial 3He flux preserved in marine sediments decreases with geological age. Whether this apparent trend reflects preservation biases or genuine temporal variations in the incoming extraterrestrial flux remains an open question.
I therefore suggest revising this section to better acknowledge these long-term variations and the implications for selecting an appropriate reference flux.
- Discussion of the two Cenozoic case studies
The two case studies highlighted in the manuscript are convincing examples of the potential of the method. However, I believe that some of the underlying assumptions and limitations deserve a more thorough discussion.
First, the manuscript suggests that the magnetostratigraphic/biostratigraphic age models may be affected by hiatuses. It would be useful to distinguish between:
- erosional hiatuses, for which extraterrestrial 3He cannot recover the missing sediment, and
- periods of non-deposition, which may indeed be identified using extraterrestrial 3He.
This distinction is important because the implications are fundamentally different.
Second, these records span several million years. Over such timescales, variations in extraterrestrial 3He flux are not uncommon in the geological record. How can the authors justify assuming a constant extraterrestrial flux throughout these intervals? I think this issue deserves further discussion.
- Interpretation of Figure 2 (compilation of 230Th focusing factors)
Figure 2 shows that more sites have focusing factors greater than one than less than one. In other words, the mean focusing factor of the compiled dataset appears to be significantly larger than unity.
This is somewhat surprising. At the scale of an ocean basin, sediment focusing and winnowing should compensate each other, since sediment removed from one location is redistributed elsewhere. Therefore, one would expect the basin-wide average focusing factor to be close to one, assuming conservation of sediment mass.
Could the authors comment on this apparent imbalance? Does it reveal a methodological bias, for example a systematic bias in the assumed 230Th production rate?
One interesting possibility would be to use this global compilation to constrain the production rate itself, by imposing that the mean focusing factor equals one and treating PCFP as an unknown parameter to be calibrated. If feasible, this would considerably strengthen the impact of the study.
Minor comments
Line 5: Fossils and biominerals may partly overlap. Please check whether these categories should be distinguished more clearly.
Line 82: Additional limitations include (i) environments characterized by highly variable sedimentation rates (e.g., turbidite-hemipelagite alternations), and (ii) sedimentary archives that cannot easily be dated because suitable chronostratigraphic markers (e.g., carbonate-rich material) are absent.
Lines 89–91: You could also mention that cobalt accumulation has been proposed as a long-term MAR proxy, based on the assumption of a nearly constant hydrothermal cobalt input (Kyte et al., 1993; Dunlea et al., 2015). This approach is mainly applicable in distal pelagic settings where terrigenous cobalt inputs remain negligible.
Line 93: "We" instead of "Ee"?
Line 156: Please provide the units of the production rate. Which parameters are required for its calculation? Since 230Th scavenging depends on particle surface area, does sediment grain size influence the estimate?
Line 168: Are you sure this statement is correct? Since scavenging is surface-area dependent, I imagine that the granulometry may affect the restuts. Can 230Th be measured reliably in sandy fractions as well as in hemipelagic muds?
Lines 183 - 184: Please define the units of CFP and PCFP. I assume these are atoms g⁻¹ and atoms cm⁻² yr⁻¹, respectively.
Figure 2: Would it be worthwhile performing a similar compilation using published extraterrestrial 3He datasets?
Line 213: Please provide the units of both 3HeET and F3He.
Lines 220 - 221: You may wish to cite our study on Jurassic-Cretaceous marl-limestone alternations (Blard et al., 2023), as well as Lucas' PhD thesis (Oxford, 2021) on the Cenomanian Turonian transition. Alternatively, if the list is intended only as illustrative, consider introducing it with "e.g.".
Line 228: You may also wish to cite Tomchovska et al., GCA, 2025, who demonstrated contrasting behaviours of 3He and 230Th with respect to sediment focusing and scavenging.
Line 237: I suggest rephrasing this sentence. It could be interpreted as implying that duration studies do not require assumptions about extraterrestrial 3He flux. In fact, studies reconstructing event durations also assume a value for F3He. Converting duration into MAR is then straightforward, requiring only sediment thickness and dry bulk density.
Figure 4: Why not extend this inventory to pre-Cenozoic sites?
Line 306: "Export" is rather vague. Please specify both the source reservoir and the receiving reservoir.
Line 339: It could also be mentioned that both the duration and the mechanisms responsible for the initial phase of the PETM remain debated.
Line 370: Please define what is meant by a "typical" sample size.
Lines 370–375: Another possible strategy would be to adapt the sample size according to the expected extraterrestrial 3He concentration, thereby ensuring that enough IDPs are analysed to achieve the desired counting statistics (Poisson uncertainty; Fenisse et al., 2025).
Line 380: When the measured bulk 3He/4He ratio approaches the terrestrial end-member, the calculated extraterrestrial 3He concentration becomes not only potentially biased but also much less precise. The propagated uncertainty derived from analytical uncertainties used in equation 10 may reach 100%, implying that 3HeET may be below the detection limit in some cases.
Line 388: You may wish to cite Fenisse et al. (2025), who showed from granulometric analyses of Concordia firn that the vast majority of extraterrestrial 3He is carried by IDPs smaller than 25 µm.
Line 435: I do not think that Figure 5 convincingly demonstrates that extraterrestrial 3He is the only robust MAR proxy beyond the Quaternary. As discussed above, this conclusion may be affected by uncertainties associated with hiatuses and long-term variations in extraterrestrial flux.
Lines 446 - 450: This paragraph might fit better in the Introduction, as it primarily provides background and motivation for the review rather than discussing its conclusions.
Citation: https://doi.org/10.5194/egusphere-2026-1583-RC3
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 361 | 133 | 31 | 525 | 30 | 21 |
- HTML: 361
- PDF: 133
- XML: 31
- Total: 525
- BibTeX: 30
- EndNote: 21
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General comments
Good research on paleoclimate. Please, follow my specific comments to improve your manuscript.
Specific comments
Lines 34-36. “High resolution archives that are also temporally expansive are integral for understanding the modern Earth system”. Insert further and recent literature on Cenozoic climate changes registered in the sedimentary record:
- Medici, G., Marianelli, D., Gori, F., Cornacchia, I., Brandano, M. 2026. Multi-disciplinary approach to paleokarst occurrence in the Eocene–Oligocene succession of the Apulia Carbonate Platform (Salento, Italy). Facies, doi: 10.1007/s10347-026-00729-5
- Jaramillo-Vogel D, Strasser A, Frijia G, Spezzaferri S (2013) Neritic isotope and sedimentary records of the Eocene-Oligocene greenhouse–icehouse transition: the Calcare di Nago Formation (northern Italy) in a global context. Palaeogeogr Palaeoclimatol Palaeoecol 369:361–376
Line 94. “Goals”. I would stick with “objectives” as you have done a few lines below.
Line 134. You have many equations and this one is obvious. Can you avoid at least Equation 3?
Line 212. Eleven equations are too many for me in a non-engineering manuscript. Some equations are very basic in this manuscript.
Line 464. A clear conclusion that re-calls your good four objectives would be useful.
Figures and tables
Figure 2. The figure can be divided in two separate figures, one with the maps and one with the graphs.
Figure 3. This figure works also in smaller size.
Figure 3. The caption is too long. Some detail can be moved to the main body of the manuscript.
Figure 4. Increase the graphic resolution in terms of dpi.
Figure 6. The caption is too long also for this figure. Some detail can be moved to the main body of the manuscript.
Figure 7. The timescale is missing and unclear.