The authors have conducted long-term, high-altitude field observations across the Third Pole under extremely challenging conditions. The resulting comprehensive mercury dataset-covering air, aerosols, precipitation, glaciers, soils, waters, ice cores, and sediment cores-represents a remarkable logistical and scientific achievement. Given the outstanding observational efforts and the overall good quality of the data presented, the manuscript is promising. The clarity of the writing and the value of the dataset are commendable. I recommend acceptance after a moderate revision, as this dataset will serve as a valuable resource for the cryospheric and environmental geochemistry community.

Major concerns

As a data paper, the manuscript lacks critical QA/QC metadata, including detection limits, analytical precision, blank corrections, and reference material results for each measurement matrix (air, aerosols, precipitation, snow, soils, waters, ice cores, sediment cores, and Hg isotopes). A comprehensive QA/QC table should be provided to ensure data transparency and reusability.

The dataset spans a vast and heterogeneous region yet relies on only two air monitoring stations, one ice core, and limited spatial coverage for several media. The authors should add a concise statement in the data description acknowledging these spatial limitations, discuss how the existing data still support the main conclusions, and clarify which findings are regionally robust versus site‑specific. This is essential for proper use of the dataset by the community.

Specific concerns

1. [Lines 87-88 and 644-645] Here, the full name of APCC is written as "Cryospheric Changes," whereas later table and figure captions use "Cryospheric Change." I recommend making this consistent throughout the manuscript.
2. [Lines 99-104, 536-564, and 577] The section arrangement here does not match the actual structure that appears later in the manuscript. The later text first has "5. Dataset limitations and applications," then "6. Data availability," followed by another "6. Conclusions." I recommend revising the section numbering consistently throughout the paper.
3. [Line 115] There is a minor typo here: a space is missing after the colon. I suggest revising it to “including: 2 stations ...”.
4. [Lines 115 and 140] There appears to be a consistency issue: Section 2.3 states 10 sites, whereas the abstract and earlier overview describe aerosol sampling at 9 stations. I recommend checking the station count for the aerosol dataset throughout the manuscript and unifying it across the abstract, methods, and results.
5. [Lines 115-116, 153-156, and 160-163] After reading this carefully, I think the statistical counting here should be rechecked: this sentence states 12 stations, but later Nepal is listed separately as having 5 stations, while the abstract states that precipitation was sampled at 16 stations. Additionally, the numbers of monsoon and westerlies stations given in this sentence do not fully correspond to "6 + 6." I recommend carrying out a dedicated cross-check using a complete summary table.
6. [Lines 162-163 and 613-628] The abbreviation for Gosainkunda is written as GSK here, but as PKR in Table 1. I recommend unifying the abbreviation; otherwise, readers may become confused when trying to identify the station later.
7. [Lines 150-151] The reference information here is clearly incomplete: "Huang et al., 4" appears to be a leftover year or numbering artifact. Please complete it.
8. [Lines 167-168 and 417-418] The overview states 12 glaciers, but the beginning of Section 4.4 later says, “Systematic snowpit surveys conducted at 9 glaciers.” If the latter refers only to a subset of glaciers where snowpits were collected, I recommend explicitly stating in the results section that this is a subset, to avoid giving readers the impression that the two statements are contradictory.
9. [Lines 195-197, 205-208, and 489-490] Here, the surface water section states 9 glacier-fed streams, but Section 4.6 later refers to 7 glacier runoffs. Please verify whether the correct number is 7 or 9, or clarify that the 7 represent a subset for which results are presented.
10. [Line 225] British spelling ("analyser") is used here, but American spelling ("analyzer") appears later on the same page. I recommend using one spelling system consistently throughout the manuscript.
11. [Lines 301-302] I suggest revising “certificated” to the more common and standard expression “certified reference materials.”
12. [Line 307] I have a minor question here: "FLPE bottles" is not a very common material description. Please check whether this should instead be one of the more common terms, such as HDPE, LDPE, or FEP. If FLPE is indeed correct, I suggest explaining it the first time it appears.
13. [Lines 121 and 359] typo, dataset I-13 should be I-1, please verify and correct it.
14. [Lines 115-116, 153-156, and 406-407] Again, the results section states 9 stations, whereas the earlier methods section states 12 stations and the abstract gives 16 stations. I recommend conducting another full consistency check of the precipitation station count.
15. [Lines 167-168 and 417-419] If this refers to a subset of the snowpit survey, I suggest stating it explicitly—for example, "9 of the 12 glaciers"—otherwise, it directly conflicts with the "12 glaciers" given in the methods section.
16. [Lines 168 and 419] I suggest using a consistent format here. Earlier in the manuscript, the form "APCC dataset I-4" is used more frequently. Although "APCC I-4" is understandable in this instance, it would be better to keep the format consistent throughout.
17. [Lines 565-566] From a grammatical perspective, I suggest revising this to either "All Hg datasets presented in this study ..." or "The complete Hg dataset ..." The current wording feels somewhat awkward.
18. [Line 603] There is also a very minor typo here: "by" and "the" are run together. I suggest revising it to "supported by the ...".
19. [Lines 613-628] When reading the table, I suspect the latitude and longitude may have been reversed. In the Kashgar row, “Latitude = 75.45, Longitude = 39.29” is obviously inconsistent with the known GPS positions, and the same issue seems to appear in the MF/TZ/RQ/TX rows below. I recommend checking all latitude and longitude entries systematically and also ensuring they are reported with three decimal places in this table.
20. [Lines 162-163 and 613-628] Here, the abbreviation for Gosainkunda is written as PKR, which is inconsistent with the GSK used in Section 2.4 of the main text. I suggest unifying the two.
21. [Lines 629-635] Here, "zhadang glacier glacier" contains a repeated word. I suggest deleting one of them to correct this small textual error.
22. [Lines 629-635] “Tanggulha” looks like a spelling typo. Please check whether it should be “Tanggula Mts.”.
23. [Lines 636-643 and 823-824] The correct spelling should be "Lingge Co". Please revise accordingly.
24. [Lines 87-88 and 644-645] In the caption of Figure 1, the full name of APCC is still given in the singular form "Change," which is inconsistent with the wording in Section 1 of the main text. Please unify the two.
25. [Line 646] I suggest revising "India monsoon" to the more common expression "Indian monsoon."
26. [Lines 636-643 and 823-824] The lake name here is still inconsistent with Table 3. I recommend carrying out a one-time, unified proofreading of all proper names in the figure captions, main text, and tables.
27. [Lines 1001-1002] This reference entry appears incomplete. Scientific Reports entries should normally include the full volume and article number, and the DOI on this page is also truncated. Please verify the original reference and correct it consistently.

This manuscript presents a comprehensive mercury dataset comprising air, aerosols, precipitation, glaciers, soils, surface waters, glacier ice cores, and lake sediment cores collected across the Third Pole. The authors have made a commendable effort to establish the integrated monitoring network through the Atmospheric Pollution and Cryospheric Change (APCC) program, thereby making it accessible to collect samples from harsh environments across the Tibetan Plateau and its surrounding regions. The data are publicly available. I believe this dataset will be of high interest to the Earth system science community, particularly for improving predictions of mercury cycling in cryosphere environments and evaluating its impact under rapid climate warming and cryosphere ablation on the Third Pole. My comments are as follows:

1.    Section 2. Observation site descriptions and section 3. Field sampling and measurement methods share a similar structure for introducing various environmental matrices. Is it possible to combine these two sections?

2.    Observation site descriptions: 1) L152: format error for Huang et al., 4. 

3.    Field sampling and measurement methods: 1) Vegetation is a useful environmental compartment to explore mercury cycling between the atmosphere and terrestrial ecosystem. Why didn’t the authors collect vegetation samples? 2) L237: HgP is not defined. 3) L305: If the same analytical procedures were applied as for aerosol and cryoconite samples, these analytical details should be presented above when you describe the aerosol and cryoconite samples. 4) L317: What is MOS-grade? 5) L336-338: The authors did not specify which samples were processed for isotope analysis. For aerosol samples, aqua regia digestion cannot eliminate the influence of interfering elements remaining in samples, which may cause the lower recovery of samples and have an impact on the accuracy of mercury isotope analysis. The authors need to point out this limitation. 

4.    Data descriptions. 1) There are no mercury isotope data from other matrices, why? 2) L352: Do you mean yearly TGM rather than monthly TGM? 3) L410: There is no MeHg in Fig. 3b. 3) L440: here and elsewhere, is the difference significant? 4) L457-463: The expression of “n” is not consistent. 5) L531-536: There is no description of lake sediment core δ202Hg values. Based on Δ199Hg, I think it cannot illustrate the rising anthropogenic Hg emissions originating from South Asia, especially for the Tanglha Lake sediment, mercury photochemical reactions can also lead to the positive Δ199Hg value. Anthropogenic Hg emission signal may be modified by the atmospheric redox before deposition into the lake sediment. The last sentence, “The dataset offers empirical constraints on Hg distribution, with an emphasis on observational evidence rather than source attribution”, looks strange in the current context. 6) There is no further discussion about the mercury isotope data of aerosols and soils. (7) I think many of the Hg isotope data from the third pole, including air and vegetation, are not included. 

5.    Dataset limitations and applications: The discussion about potential applications is very broad and vague. Please try to be more specific. 

    

Figures: Fig. 4 is not referred to in the main text. Fig. 7 only shows the sampling location. Is it possible to add THg concentrations like other figures? Fig. 9 only shows the lakes from the northern slopes. Why not show the data from the southern slopes? 

Tables: The information in tables doesn’t match well with the main text. The authors need to carefully check all the details. 

General comment: This is a valuable and difficult-to-reproduce mercury dataset across multiple environmental compartments of the Third Pole. The breadth of the dataset is impressive and potentially well suited to ESSD. However, the manuscript still reads partly as a synthesis of previous Hg studies rather than as a data-description paper. I suggest that the authors more clearly refocus the manuscript on dataset structure, reuse, file-level metadata, and practical guidance for users.

Lines 96–104 and Section 4: The stated aim is to provide an overview of previous studies and present an organized Hg dataset. This is appropriate, but Section 4 often emphasizes interpretation of previous findings rather than documenting the dataset itself. To improve ESSD suitability, each APCC subset could follow a more consistent data-description format: repository file, matrix, variables, sample number, temporal coverage, spatial coverage, method summary, descriptive statistics, limitations, and recommended use.

Lines 40–44 and 565–577 / Data availability: The manuscript states that the data are archived in standardized Excel format and provides a DOI, but the relationship between the manuscript, repository files, and data products remains unclear. Please add a concise repository overview table listing each APCC subset, file name, environmental matrix, number of records/samples, temporal coverage, spatial coverage, main variables, and associated publication. This would complement the more detailed file-format and accessibility issues noted by other reviewers.

Lines 261–267 and 319–322 / Derived variables: HgP is calculated as HgT–HgD. Please clearly mark derived variables in the repository files and document the calculation method. Similar documentation is needed for deposition fluxes, accumulation rates, chronology-derived variables, spatial gradients, interpolated fields, etc.

Lines 390–416 / Precipitation Hg: Please state explicitly whether the repository contains event-level precipitation data, seasonal averages, annual fluxes, or only summary statistics. If fluxes are included, please document the equations and precipitation inputs used to calculate them.

Lines 500–518 / Ice core and lake sediment cores: Please provide more information on chronology and uncertainty. For sediment- and ice-core data, the repository should include or link to age-depth models, dating methods, dating uncertainties, accumulation-rate calculations, and whether Hg fluxes are measured or derived.

Lines 520–536 / Stable Hg isotopes: Other reviewers have raised important scientific questions about isotope interpretation and missing isotope matrices. In addition, from a data-reuse perspective, the manuscript should explicitly document what is included in the isotope files: δ²⁰²Hg, Δ¹⁹⁹Hg, Δ²⁰⁰Hg, Δ²⁰¹Hg, analytical uncertainty, replicate measurements, sample matrix, sampling date, site, associated Hg concentration, notation, and reference standard.

Lines 537–564 / Dataset limitations and applications: Building on comments from other reviewers, this section would be more useful if it provided practical guidance for users. Please clarify which subsets are suitable for regional spatial comparisons, seasonal analysis, model evaluation, deposition-flux estimation, source attribution, or climate-change impact studies. Also identify uses that are not recommended because of sparse coverage, non-synchronous sampling, method differences, or limited temporal resolution.

Lines 565–577 / Data availability: Please add repository title, dataset version, release date or access date, license, whether future updates are expected, and whether all data needed to reproduce the manuscript figures are included.

General comment on figure/data traceability: Please improve traceability between the figures and archived data by stating which repository files and variables were used to generate each figure. If figures use processed or unpublished intermediate data, those data should also be archived or clearly documented.

Comment on APCC dataset I-7 / ice and lake sediment cores: The core dataset should include depth information for both ice-core and lake-sediment-core records. At present, the table includes age-related fields such as “Ice age (year)” and “210Pb-inferred Year,” plus Hg concentration and flux columns, but it does not provide core depth or depth intervals. Please add, at minimum, core depth top, core depth bottom, interval thickness/resolution, and units for each record. For sediment cores, depth intervals are necessary to evaluate the age-depth model, accumulation-rate calculations, and Hg concentration profiles. For the ice core, depth intervals are needed to link Hg concentrations and deposition fluxes to the chronology and sampling resolution.

Comment on APCC dataset I-7 / mixed core structure: The current core table appears to combine ice-core and lake-sediment-core variables in the same rows and columns, including “Ice age (year),” “Hg Concentration (ng L⁻¹),” “210Pb-inferred Year,” “µg yr⁻¹ m⁻²,” and “ng g⁻¹.” This structure is difficult to interpret because ice-core concentrations are reported in ng L⁻¹, whereas sediment concentrations are reported in ng g⁻¹. Please separate ice-core and sediment-core records into distinct tables, or add a clear “archive type” column and ensure that non-applicable fields are blank and well documented.

Comment on APCC dataset I-7 / ambiguous flux columns: The columns “ug/m2.yr” and “ug.yr-1.m-2” appear to represent fluxes, but their meanings are unclear and the two unit formats may refer to the same quantity. Please rename these fields using standard notation, for example “Hg deposition flux (µg m⁻² yr⁻¹),” and specify what calculation method was used to derive these values.

Comment on APCC dataset I-7 / missing site metadata: Many rows in the core table appear to lack region, abbreviation, elevation, or reference information. Please ensure that every record includes core/site name, archive type, latitude, longitude, elevation or lake altitude, reference, and chronology method. Repeating metadata in every row is acceptable and makes the table easier to reuse programmatically.

Comment on APCC dataset I-6 / summary statistics mixed with raw-style fields: The river and lake water table mixes sampling time, concentration ranges, average ± standard deviation values, sample counts, and water-type categories. For reuse, it would be better to split these into separate columns: mean, standard deviation, minimum, maximum, number of samples, sampling start date, sampling end date, and water type. Values such as “5.44+-6.16” and “0.64–32.96” should not be stored as combined text strings.

Comment on APCC dataset I-4 / glacier data units and depths: The glacier table combines snowpits, surface snow, and cryoconite, but the concentration header is “Hg concentrations (ng/L).” This is appropriate for melted snow/ice samples but not for cryoconite, which is typically reported on a mass basis, such as ng g⁻¹. Please add a concentration-unit column or separate tables by sample type. For snowpit samples, the table should also clearly distinguish total pit depth, sample depth interval, and vertical resolution.

Comment on APCC dataset I-3 / zero versus missing values: The precipitation table contains zero values in concentration and flux fields for some months. Please clarify whether zero means no precipitation/no sample/no Hg detected, or whether these should be missing values. This distinction is important for calculating deposition fluxes and seasonal averages.

Comment on all APCC data tables: Please avoid storing numerical values as combined text strings. Ranges, means, standard deviations, sample counts, and units should be separated into machine-readable columns. For example, instead of reporting a value as “5.44 ± 6.16 ng L⁻¹” or “0.64–32.96 ng L⁻¹,” please use separate columns such as THg_mean_units, THg_sd_units, THg_min_units, THg_max_units, and sample_count. Here, sample_count refers to the number of samples or observations used to calculate the summary statistics. This would greatly improve reusability and reduce ambiguity.

Comment on blank cells/missing values: Blank cells should be avoided unless their meaning is clearly documented. Please distinguish among: missing but expected data, not-applicable fields, below-detection-limit values, and not-measured variables. These cases should not all appear as empty cells because they have different implications for reuse. A data dictionary could define standardized entries such as NA = missing value, not_applicable = field does not apply to this sample type, not_measured = variable was not analyzed, and a detection-limit flag such as below_detection_limit = yes/no. This would make the tables much easier to parse and interpret.