Review of “No evidence for tephra in Greenland from the historic eruption of Vesuvius in 79 CE: Implications for geochronology and paleoclimatology” by Plunkett et al.

General comments

This manuscript presents a novel methodological approach for studying past volcanic eruptions that is intersectional and of relevance for many research areas. Combining sulphate and cryptotephra records with modelling of volcanic ash, the authors present a convincing argument that is well-structured and well-supported. Given the existing constraints, e.g. limited or no reference data available for many eruptions or volcanoes, I think this presents a thorough summary that has tested the limits of their hypothesis appropriately and represents an important step forward for work in this area. I also agree strongly with the concluding comment of the abstract here regarding a need for formal acceptance of the revised ice-core chronology.

Specific comments

97: It’d be nice to have a comment here on your choice of VEI 4 as a limit, and whether VEI 3 eruptions could erupt ash into the stratosphere in the right conditions (e.g. with a seasonally low tropopause). I don’t know if modern observations can speak to this?

Figure 4: Is there a reason why mean values for Mt. Spurr are plotted when single-point data are available, for some modern eruptions at least (e.g. Crater Peak 1992)? If the average data are used I think the uncertainties need to be plotted as well.

169: While the point data that you show here do seem broadly offset between OC1-5 and QUB 1832/33, there is still a good degree of overlap with the compositional field that is plotted from published data. If you’re writing this off as a source it would be useful to have a comment explaining why the published data field isn’t seen as reliable here.

193-194: As you have whole rock data for Mount Spurr plotted, I think this point should be clarified here (e.g. specifying glass, or single point data, etc).

I’m not overly familiar with data for this volcano myself, but the AVO website’s geochemistry search (https://avo.alaska.edu/geochem/search) shows whole rock data are available from three additional references by Nye et al. (listed below), which may be comparable to the data from George et al. (2003). Two samples listed as tephra fall pyroclasts from the 1953 eruption (85CNS16 & 17, the latter looks like it was included in George et al.).

Nye, C.J., Beget, J.E., Layer, P.W., Mangan, M.T., McConnell, V.S., McGimsey, R.G., Miller, T.P., Moore, R.B., and Stelling, P.L., 2018, Geochemistry of some quaternary lavas from the Aleutian Arc and Mt. Wrangell: Alaska Division of Geological & Geophysical Surveys Raw Data File 2018-1, 29 p. https://doi.org/10.14509/29843

Nye, C. J., Harbin, M. L., Miller, T. P., Swanson, S. E., and Neal, C. A., 1995, Whole-rock major- and trace-element chemistry of 1992 ejecta from Crater Peak, Mount Spurr volcano, Alaska: in Keith, T. E. C., (ed.), The 1992 eruptions of Crater Peak Vent, Mount Spurr volcano, Alaska, U.S. Geological Survey Bulletin 2139, p. 119-128.

Nye, C. J., and Turner, D. L., 1990, Petrology, geochemistry, and age of the Spurr volcanic complex, eastern Aleutian arc: Bulletin of Volcanology, v. 52, n. 3, p. 205-226.

209: I would appreciate some discussion here of the parameters used for modelling shard size and shape. Were the grain sizes reported here for the cryptotephra converted into the diameter of an equivalent sphere? What does the fine tephra measurement of ~30 μm relate to here? Cryptotephra data have been used to show that the transport distances of modelled particles are affected by sphericity (e.g. Saxby et al., 2020) so a sentence or two that comments on these details as they relate to your work would be of value here.

208-209: Where do these eruption parameters come from? It’s clarified for some of the following examples, but not here.

228-230: Related to the previous point, why is the grain size distribution detailed here for this source but not for the other two?

SI suggestion: related to my two previous comments on eruption source parameters, it might be useful to add a summary table that lists these (with references) for all three sources to the SI.

233: I think that your approach here, and what you describe at the start of this paragraph, is really important. Given that you’re investigating an eruption that we don’t currently have records of the assumption that it’ll be like other eruptions that we do know about may not be valid. The individual initial runs are therefore likely too narrow in range and the testing with 1000 random events gives useful probabilistic bracketing data. I would suggest that this understanding could be emphasised earlier – that you’re trying a best approximation from known data but it may fall short and it’s more useful to test a range of values – because I think it’s the only way we can really usefully study past eruptions. A sentence along these lines could be added to the methods, e.g. around lines 118-121.

Technical corrections

56: The use of ‘rather’ here seems like an odd choice of a qualifying word. From what you go on to say, I think this should be stronger.

74: What exactly is meant by an ‘unambiguous tephra’? Does the ambiguity (or lack of) relate to whether the grains are indeed volcanic glass, the number and size of shards, their geochemistry, or whether they generally constitute a useful or reliable marker ‘horizon’?

99: I think this should be ‘…none of which are historically dated…’ instead of ‘none of which is…’.

244: It might be worth stating that this reference is a data archive/record, as this wasn’t apparent to me at first glance and it seemed odd that the data weren’t included in this publication. I can now see why, having accessed it!

Figure 4: It’s a little confusing that the Aniakchak points and the Spurr range are in similar orange colours - my first impression was that they were the compositional range and point data for the same source (as is the case for the other volcanoes). I assume you’ve chosen similar colours to show they’re both from Alaska/Aleutian Arc, but there might be a clearer alternative.

Figure 6: What are the triangle points shown on this plot – Aniakchak? They’re not included in the key.

Figure 8: I’d find some extra labels on the plots useful here (e.g. all months, Nov-Feb). When I first looked at this, I was also a little confused because I assumed the four parts would be read left to right in two rows, not in columns. If you keep this layout, I suggest that you highlight the labels more clearly. Lastly, I don’t think the blue line in the key matches the blue line on the plot.

General statement: The manuscript is well written and well-structured and outlines evidence that will be relevant to a wide community, specifically those reconstructing volcanism over the past 2 millennia and those attempting to synchronise records to Greenland’s GICC05 timescale. Resolving regional climate responses to short-lived events (e.g. volcanic events) between records (such as ice and tree rings) at the annual requires reducing uncertainties of each record. Using sound geochemical evidence, the Vesuvius 79 CE tie-point has been refuted, and its removal will improve the matching of events and future ice-core chronologies. Those using the high-res ice core timescale should be aware of these dating biases when transferring ages. 

Specific comments:

Line 73: I think you should also mention the work of Adolphi and Muscheler (2016, 2018), as they also published age adjustments the GICC05 timescale for the Holocene. I believe their recommended adjustment of GICC05 is by -11 years at 79 CE, which is consistent with the corrected age of 88 CE by Sigl (2015) and McConnell (2018).

Line 116: In a recent paper by Barker et al 2019 who modelled Taupo fallout, they simplified source parameters - i.e. did not: 'include the effects of atmospheric rainfall on ash removal and assumed that 100% of the erupted mass rose buoyantly into the atmosphere.  Were there similar simplifications/assumptions in this model? Can they be mentioned somewhere or referenced?

Line 121: Is there a reference for the observational source data from the 3 different locations, or is it the same as the reference link from line 115?

Line 125: What were the shard concentrations like for these samples? Were they a few hundred or higher.

Line 131: I think it would be useful to have the depths or age ranges here.

Line 154: I found lines 153-160 slightly confusing to read. I haven't plotted the data, but from Fig 4 a-d, it looks like there are perhaps 4 or 5 grains from 1859 that resemble 1832/3 when major elements are compared. You mention there are just 2. I know there are only 2 shards that could be compared on their trace elements.

Line 210: this is an interesting finding. Past tephra searches in Greenland, which concentrated on tracing VEI >5 events probably found less tephras than are potentially there. If continuous sampling is not adopted, then events with an estimated VEI of 4 should be included in the search criteria.

Line 220: The VEI is not known, but tephra from 1 event from Kuriles has been traced in NEEM (Bourne et al., 2016), around 23 kyr BP.

Line 234: References needed here for observed data?? or, are you still referring to data from Global Volcanism Program, 2013 for Chikurachki and Siebe and Machias 2004 for these parameters? Would be good to make this clear.

Line 264:  I would suggest to add that the Barbante et al 2013 analyses were by the SEM-EDS method. Maybe you can discuss why this method is not so robust, compared to WDS, with less analytical prevision, higher detection limits on top of the added challenge of dealing especially when dealing with 6 small grains of about 5 microns. There were secondary standards available, did you also assess the quality of data from these analyses?

Line 267: Do you mean due to measurement error (instrument sensitivity, detection, small glass area for bombardment or assay?) in the analysis of such small grains?

Line 279: This seems to follow the trends of late Holocene cryptotephra deposits found in different North Hemisphere, high latitude records from Greenland, North America, Newfoundland and Svalbard. Are the distal deposits found in these records all > VEI 4? Would be good to mention if so, as it is consistent with the model results.

Line 293 –ref to Adolphi and Muscheler (2016) here?

Line 339: Were you also able to test this model on any recent events from the different locations – with known parameters such as eruption duration, plume height, grain size and deposition pathway? e.g. eyjafjallajokull. If so, did you find consistency between model output and the observations? Perhaps this has been demonstrated in other studies using the same model?

Line 348: can you give details about this? Which records showed the response and is there a duration or temp inference?

Typing/Technical

Line 19 – suggestion: add ‘deposits’ or ‘layers’ here?

Line 28 Suggestion: ‘assigning’? this volcanic event to 79 CE.

Line 34_ suggestion to simplify.

Line 44 – refs needed here

Line 51 – ref needed here

Line 53 – suggestion: tephra ‘deposit’?

Line 62 – suggestion: ‘major element’ geochemistry

Line 112 – suggestion –use of ‘vanishingly small’? is not commonplace outside of native English speakers?

Line 137 -  Masaya, ‘Nicaragua’.

Line 144 - Should the regions of origin accompany the names here? E.g. Kamchatka.

Line 154 – Suggestion: of ‘the previously published’ QUB-1859 from….etc.

Line 160 - Fig 4?

Line 165 – Full stop after Pink Pumice – long paragraph,

Line 179 –Lesser Antilles (Caribbean)

Line 179- The sentence starting with Although is a bit confusing. But it seems both points suggest that there is no correlation with Japanese sources?

Line 207: Suggested insert - at least two other VEI 4 eruptions have occurred ‘at Aniakchak’.

Line 919 – remove the repeat mention of ‘in this paper’ Fig 1a.

Fig 4 – can you add (a) – (d) to the biplots.