General comments:

The manuscript “Tephra data from…” by Beckett et al., 2023 aims at assembling a dataset of tephra layers found in (varved) European lake records during the time interval 25 ka BP to 8 ka BP. This is a valuable approach, as tephra layers can be excellent isochronous marker horizons within sedimentary records, which can be utilized to improve chronologies and to enable inter-site paleoenvironmental comparisons. Since record specific tephra datasets are most commonly published in individual publications, combining published results in a database can foster data accessibility and reusability.

In the present form, it is not clear if this contribution reports on a novel modification/extension of the existing database VARDA or if the presented results are a summary of a VARDA database query. Comparing the structure of VARDA presented in Ramisch et al., 2020 with the actual database website and with the information given about data collection within the manuscripts, some fields (geochemistry) may have been added. However in the manuscripts it is also stated, that this is beyond the scope of this paper. Therefore, I strongly suggest to stress clearly, what is a new or has been modified and what is just a summary of a database query.

It is stated that the project focuses on varved records, but for the integrity of a robust tephra-dataset it is extremely important to consider also data of non-varved records. Important information about the age and glass geochemical composition, which are needed for correct correlations and precise chronologies, may come from such non-varved records and thus would be not considered otherwise. The authors already have partially identified this issue, including data from Lake Ohrid/Prespa, although their non-varved character is not consequently reported (see also community comment, as this seems to be the case for other records). However, the approach stills appears incomplete. As the manuscript is reported to be the initial phase, setting the basis for an overarching project, this needs to be considered in depth for the overall long-term project aim to construct global frameworks.

With regard to the compiled dataset of tephra layers, there is no discussion about the quality of the data collected (quality of geochemical analyses, ages). Even if the authors state that the best age of a tephra/eruption may be a subjective feature, the general quality of available ages could be addressed, as in the end for the application of tephrochronology a single age is needed to unify and align chronologies and their records (and it was done in Figure 4).

Specific comments:

Title: Consider rephrasing as with the focus set to European volcanism and varved lakes, the dataset collection rather reports on a regional than a global tephra framework. Also, there is no discussion about existing ages, so that chronologies were not improved yet.

Abstract:

Consider rephrasing with regard to point out if this contribution represents modification or a query of the VARDA database.

Please check the given numbers about records and tephra layers. Not all 19 records represent varved records, further Figure 3 shows more than 49 tephra layers…

Introduction:

Figure 1 is not crucially needed with respect to the dataset compiled.

Methods:

It appears that only records registered within VARDA were considered, which may be insufficient to present a full list of (varved) records. Using only data given in VARDA strongly depends on data quality, maintenance and update of this database. This is not discussed within the manuscript. For example, for the Lake Ohrid tephra data the latest results 2019-2023 are not included in VARDA (and were potentially also missed by the google scholar search). Therefore, to provide a reasonable and critical review of existing data to compile a dataset about tephra layers of the LGIT, there should be not only one database considered, but also additional non-database listed references included to ensure completeness and quality of the presented data.

I would suggest rephrasing of the data collection paragraph in order to point out, how VARDA was modified (see general comment).

For the new data fields, there are some fields listed in the supplementary data (such as data_availability, datset, lake, geochemistry_availability), which are not given in the tables of the manuscript. There are also only major element data fields (Table 1), whereas also trace element data is given in the attached dataset. For Table 2: What about adding a field for importing uncalibrated radiocarbon ages in order to simplify recalibration of radiocarbon ages using the same IntCal curve. Please consider adding the information based on which calibration curve the age was calibrated.

Further, for the tables presented I would consider avoiding colour coding of mandatory and optional fields in order to make figures accessible to readers with colour-blindness.

Results:

Referring to the community comment, I also welcome the provided original references of the compiled individual datasets. Please check if your list of references (p.6) is complete and includes latest references (and the consequences for Figure 3).

Based on the tephra correlations presented in Figure 3, eight of them are reported within the focussed time-interval. Consider if it is applicable to report on the different geochemical results of these eight (everywhere the same composition, variations?) and about their potential ages, findings may also be picked up in the implications section.

Figure 3: Is it necessary to report tephra layers well beyond this interval? Otherwise, I would suggest highlighting the focussed time-interval.

Implications:

I like the idea of comparing the known distribution of an ash cloud with the location of available records to identify potential new targets for (crypto)-tephra investigations. Please specify how the list of (7) new locations was compiled. Maybe it is worthwhile to consider a function for VARDA to report also non-successful cryptotephra investigations, which did not yield any (crypto-)tephra findings. Also these (negative) findings may help to improve knowledge about ash distribution, but also avoid unnecessary investigations by others.

Beckett et al. report on the addition of information about the occurrence and geochemical compositions of tephra layers within European varve sequences reported in the Varve Database (VARDA). Tephra layers offer the potential to connect varve chronologies at single moments in time with the potential to compare and transfer differential and absolute dating information between sites and increase overall dating precision by replication. The authors state the aim to incorporate information from tephra in varves globally over the coming 5 years, as well as increase the window of time for which data is included. The reasoning for the addition of tephra information is well presented in the manuscript, although as VARDA already includes relevant palaeo proxy datasets for many sites, it is unclear why this particular “proxy” requires a stand-alone publication. The addition of the “event” layer bar in the GICC05 panel of the VARDA home page is a useful tool for quick reference however, and perhaps as the database grows it will be able to provide a useful online reconnaissance tool for project design and field site selection.

Whilst this paper addresses the inclusion of tephra data, I found myself browsing VARDA more generally and found myself confused by the inclusion of many non-varved lakes (I’ll just note some of the ones I am familiar with as I have worked on them: Lake Victoria, Lake Tanganyika, Lake Bled). I felt a clearer description of the VARDA database itself was probably needed to make sense of the datasets entered to date.

My comments here are organised under the headings indicated by the ESSD Review guidelines, followed by some minor editorial notes.

1. Significance

This particular compilation of tephra layers reported in varve sequences in Europe is unique in that draws together commonalities in the records and could be a great time-saver in looking up sites and articles. Highlighting the value of tephra layers to varve researchers is also beneficial.  Additional value could be achieved by including specific and relevant search tools, such as those from the RESET database (Bronk Ramsey et al., 2015) that are used within the article to show the connectivity between records using tephra layers and maps (e.g. KDE) of the sites where tephra layers have been reported. The RESET database is problematic as it is no longer maintained, but in terms of a tephra data repository it is more complete and contains critical data missing from the VARDA database as presented (see comments under Data Quality). The VARDA team might be better to find a means to connect VARDA with that database, rather than starting again to record all of the published tephra layers in Europe within a new repository.

One note that caught my attention was a sentence in the conclusions about the opportunity to explore machine learning approaches to tephra compositional analyses. As there is no mention of this in the body of the paper, it needs further exploration and justification. If there are additional novel tools being created that could really add to the uniqueness and usefulness of the growing compilation.

2. Data Quality

Data in VARDA is easily searched and clearly presented and downloadable. I applaud the inclusion of EPMA analytical conditions but I cannot understand the exclusion of secondary standard data, which is critical to evaluating whether one can compare to another tephra dataset or not. The authors referred more than once in the manuscript (e.g. within Table 2 that sets out mandatory and optional criteria for metadata) to the inclusion of “Standards used for analytical calibration, e.g. Lipari Obsidian”. Two types of standards are used in EPMA work and there seems to be confusion here. Primary standards are usually a suite of minerals or oxides with known elemental compositions, which are used to calibrate the instrument. The publication of primary standard data is not conventional, as what matters is that the data is accurate, not which minerals were used for which elements. Secondary standards are materials of known composition that are analysed before, during and after a run of analyses on an unknown sample, in order to demonstrate the accuracy and precision of the calibration. These are usually matrix-matched to the material being analysed, so the Lipari Obsidian, for example, is a commonly used secondary standard for the analysis of volcanic glass. The tephrochronology community has long called for the inclusion of secondary standard analyses alongside ALL tephra compositional datasets, so that the data may be trusted to make comparisons between tephra datasets generated at different times and on different instruments (e.g. Hunt and Hill., 1996; Kuehn et al., 2011; Wallace et al., 2022). At present, I couldn’t find any secondary standard data within the VARDA database, which means that if I were to use it to trace tephra layers, I would immediately have to open the original article and extract the data from there, rather than from VARDA. Those less conscientious might unintentionally propagate poor data and miscorrelations. I strongly recommend that the database authors amend the database to include either i. a clear statement that *only* data with secondary standards within 2 standard deviations of published assays are included in the database (a lot of work for data stewards), or, ii. secondary standard analyses for all tephra datasets, alongside a link to published assays.

In addition, reviewing the criteria for recording tephra geochemical data, I would recommend that data type (e.g. single grain, whole rock) and material (e.g. glass shards, mineral, pumice) are added as essential criteria. This is also essential metadata for ensuring like is being compared with like and whilst most data will be single-grain glass shard analyses, it should not be a given.

3. Presentation

The article and the database are clearly presented and easily accessible. No issues here at all.

4. Minor editorial notes

Title: I would focus here on European and LGIT tephra data in varve records as the potential and value of a global inventory isn’t obvious from the article and data at this stage.

Abstract: There is inconsistent use of capitalisation and hyphenation in “last Glacial-Interglacial transition” between the paper title and the abstract that needs correcting one way or the other.

Introduction, Line 39: the term “well defined” needs explaining.

Methods, Line 73 and 98/99: References to standards used for calibration, rather than secondary standards, needs correcting.

Figure 2 and 3: Bled, Ohrid and Prespa (at least) are not varved and their inclusion needs an explanation. If non-varved lakes are included, then what do we get from VARDA that is unique? There are asterisks noting that Ohrid and Prespa are non-varved in Fig 3, but not Bled. Other sites I am less familiar with.

Results, lines 126 and 133: The Mediterranean does not describe a volcanic region and it would be better to define to at least Italian and Hellenic Arc, if not specific volcanic fields.

Line 151: I would replace “tephra plume” with “tephra fallout area” to avoid any indication that the sites studied faithfuly capture the plume dispersal of an eruption. This is especially pertinent given that within the screen shot of Askja-S sites, Iceland (therefore the volcano) is not included in the shaded envelope.

References used:

Ramsey, Christopher Bronk, Rupert A. Housley, Christine S. Lane, Victoria C. Smith, and A. Mark Pollard. "The RESET tephra database and associated analytical tools." Quaternary Science Reviews 118 (2015): 33-47.

Hunt, John B., and Peter G. Hill. "An inter-laboratory comparison of the electron probe microanalysis of glass geochemistry." Quaternary International 34 (1996): 229-241.

Kuehn, Stephen C., Duane G. Froese, Phil AR Shane, and INTAV Intercomparison Participants. "The INTAV intercomparison of electron-beam microanalysis of glass by tephrochronology laboratories: results and recommendations." Quaternary International 246, no. 1-2 (2011): 19-47.

Wallace, Kristi L., Marcus I. Bursik, Stephen Kuehn, Andrei V. Kurbatov, Peter Abbott, Costanza Bonadonna, Katharine Cashman et al. "Community established best practice recommendations for tephra studies—from collection through analysis." Scientific data 9, no. 1 (2022): 447.