Dear Sebastiaan van de Velde, dear authors,

As requested, I read and reviewed the manuscript titled “Coral Skeletal Proxy Records Database for the Great Barrier Reef, Australia” submitted by Ariella Arzey and colleagues for publication in Earth System Science Data. In their manuscript, the authors present and discuss the new online Great Barrier Reef Coral Skeletal Records Database (GBRCD) which compiles a large quantity of data gathered in studies into skeletal coral records in the Great Barrier Reef in Australia. The main aim of this study, and the database, is to compile this proxy data from various sources to make it available and machine readable for future research projects. Given the wide distribution of this type of data in different formats and in various repositories and supplements, this database presents an important effort to making the hard work of coral researchers more easily available for meta-analyses and comparison studies.

The authors provide a clear introduction explaining the need for a comprehensive database of coral records in the Great Barrier Reef and a thoughtful background section on the reef setting and the types of proxies contained in the dataset. A small point of feedback would be that it is not immediately clear that this is an open dataset and that more data can be added. However, this becomes clear at the end of the manuscript. The two formats of the database (LiPd + CSV) ensure ease of access of all the data both through manual downloading and data processing and by machine reading. Figures 3-5 give a nice sneak peek into the spread and location of geochemical records in GBR.

In conclusion, I believe this manuscript and the database it presents represent a valuable contribution to the field of (coral) sclerochronology and paleoclimatology and I would be happy to support its publication more or less as is. Below, I detail a few minor textual comments I had while reading through the manuscript, but beside these I think the manuscript is in pretty good shape.

Minor comments

Line 152: “The GBRCD is along the lines of” should probably be rephrased to something like “The aim of the GBRCD is similar to”

Line 157: The past tense in this criterion and others suggests that the database is done and not a living product. Perhaps this can be rephrased.

Line 208: It is unclear to me how a Mg/Ca ratio can have a value <0, since it is a ratio of two concentrations. Perhaps the authors can use another example of “abnormal data”.

Best wishes,

Niels de Winter

General comments

This manuscript provides details on a new coral geochemistry database for the Great Barrier Reef, the first for a particular location and the most studied reef in the field of coral-based paleoclimatology. The paper is well-written and comprehensive and the database is easy to use and well-documented. This database will be useful to coral and climate researchers, especially those working with climate models for many decades. The authors intend to update the database annually and provide a contact email on the GitHub page but not a person to contact. It would be useful to have this information in the manuscript and NOAA Paleo website. Additionally, please add details for continued maintenance at the institution (AIMS)? or PI level so data can be submitted 10 to 20 years from now. There should also be a mechanism for corrections to be database to be submitted to those maintaining the database.

Specific comments

Why just coral records to the Holocene? What about the IODP coral records and other GBR coral records that go that further in time? (e.g., Felis 2022 doi:10.1029/2021GL096495, Brenner 2020 doi:10.1029/2020PA003962, Brenner 2017, doi:10.1002/2016PA002973, Felis 2014 doi:10.1038/ncomms5102). perhaps these could be added in the next update to the database.

The authors also included growth data (extension per year, density, calcification) in the database, please mention this in the introduction. These parameters help understand the coral geochemical data and should be archived together. A repository of X-radiographs from the corals and cores would also be useful, but this reviewer understands hosting the large file sizes is an issue but should be included as part of a permanent archive or a next update project.

The authors note that we are losing valuable coral data as corals become increasingly more threatened, which is true. However, we are also losing coral data as researchers retire or computer systems age, and thus data and paper files are lost, unreadable on newer computers, or destroyed. Rescuing these data sets is also important and may require transcribing paper files to digital formats. I would encourage the authors to note this as well and may be used to persuade institutions to save these important files as well as funding agencies to support these efforts. 

I would also encourage the authors to make physical coral cores and their location a part of the database or a separate database, with International Generic Sample Numbers (IGSN) (Dassie et al., 2017). The IGSN numbers should be included in the metadata as well.

Regarding diagenesis in the discussion section. There are best practices for detecting diagenesis that are well rooted in geology including thin sections, SEM, and XRD as well as X-radiographs, and these techniques are taught at the undergraduate and graduate level in geology. Many studies have urged researchers to assess their corals for diagenesis, see the study of Quinn et al., 2006 suggested that “...practice of publishing coral geochemical records, especially those that depict abrupt and/or large changes in tropical SST, without additional documentation of the pristine nature of corals used in their study should be avoided.” See the recent paper of Weerabaddana et al., 2024 https://doi.org/10.1029/2023PA004730. Yes, this should be standard practice and researchers should be and are trained to detect diagenesis in their corals and reconstructions, and in this reviewer’s experience, it has been done in the last 20 years. It may not be mentioned in the main text but is given in supplemental files. Some short-form journals (Nature, Science, GRL, Geology, etc.) do not give authors the space to add these details to the paper so they get left out. This does not mean the authors did not screen for diagenesis.

This reviewer likes the comment in the discussion about the minimum metadata that should be included in a publication. I hope journals start to insist on these metadata for coral paleo papers. Additionally, this reviewer appreciates the additional information included in the Appendices.

The conclusions section is redundant and not needed, please remove it.

Technical corrections, typing errors

Figure 1 Are the selected rivers those studied by research in the database or just author preference? If they are studied rivers, perhaps rephrase “select” to “studied”.

Line 118 Revise “…as they incorporate trace elements, varying isotopes, and organic materials in proportion to environmental variations…”  Each element has several isotopes and it is the ratio (e.g., δ¹⁸O and δ¹³C) or absolute amount (e.g., ¹⁴C, ^239+240Pu) that are used for climate and environmental proxies whereas other isotopes are used for dating (²³⁰Th).

Line 128 Revise “…and were at the forefront of the development of new technology and methods to measure these coral properties.”

The data files for one Tree Island and Dip and Stanley Reefs on the NOAA Paleoclimatology website are tab-delimited, not CSV files as stated in the manuscript lines 254, 198, 191, 73, 34, etc.). CSV files use commas to separate values, not tabs. “# Data line format - tab-delimited text, variable short name as header”  

Line 257 and elsewhere (not every instance is noted, use find and replace). Revise “between” to “among”. Between is used for two items, among for more than two items.

Line 293 Spell out maximum and minimum.

Table 3 and text description of “Age” lines 266-270. Are the monthly data given as the middle of the month or the beginning of the month? After looking at several of the data files, the age or time step varies by record. Some records are resampled to even monthly time steps and others are not. This reviewer assumes the authors are using the original age given by the original authors. Please clarify this in the text and note that some users may need to resample or linearly interpolate to get data on the same time scale.

Table 7 geo_longitude Description. “All values are positive due to the GBR’s location east of the Prime Meridian.” This could be confusing, in general, longitude east is positive and east is negative, it has nothing to do with GBR’s location. Please revise. Same for geo_latitude.

Table 11 Is there a field for records that have been resampled to even time steps? This reviewer did not see this mentioned in the text either.

Tables 9 and 12 The authors use 1 and 2 sigma for analytical precision, most likely given by the original authors. Why not make it easier and just make all precisions 1 sigma as in previous databases (Iso2k and Coralhydro2k)? This will help users who may not understand the differences.

Table 13 calib_SSTdata and text. It would be useful to also include version numbers especially for ERSST since each version can be quite different. Additionally, OISST has different resolutions that would be useful to know.

Figure 3 This is a nice figure but panels A and C are hard to read, perhaps switch to a log scale so the fossil corals can be seen for their group and resolution.