the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Feb 2024
Technical note: A fast and objective autosampler for direct vapor equilibration isotope measurements
Abstract. To investigate water movement in environmental systems, stable water isotopes (2H and 18O) are commonly used tracers. Analyzing the isotopic composition of water bonded to substances like soil or plant tissue necessitates its extraction prior to analysis. One such method, direct vapor equilibration, is popular due to its cost-effectiveness and straightforward sample processing. However, sample analysis requires a significant manual labor, thereby limiting the number of samples that can be analyzed. This limitation is compounded by the fact that stored samples undergo isotopic composition changes over time, and in addition manual measurements may lack objectivity. To address these challenges, we have developed a device that automates the analysis process. Our vapor autosampler for vapour samples, named VapAuSa, features a modular design with up to 350 ports for direct vapor equilibration samples. These ports sequentially connect the prepared samples to the analyzer, enabling continuous automatic measurements. Within the accompanying software, measurement criteria can be specified, facilitating objective analysis. VapAuSa measurements have a similar uncertainty compared to manual analysis of identical samples (VapAuSa δ2H = +/− 4.5 ‰ δ18O = +/− 0.58 ‰ vs. manual δ2H = +/− 5.7 ‰ δ18O = +/− 0.37 ‰). However, the increased sample throughput minimizes storage-induced isotopic changes. Moreover, VapAuSa triples sample throughput per week while reducing direct labor time to just 10 % of manual processing.
- Preprint
(2107 KB) - Metadata XML
-
Supplement
(21 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2024-528', Anonymous Referee #1, 25 Mar 2024
Section 1: General Comments:
The developed automated sampling apparatus is a valuable improvement to the DVE methodology as it increases sample throughput in turn potentially limiting storage impacts on isotope ratio measurements. The possibility of decreased in-lab labour will also be of benefit to labs with large sample throughput needs. This apparatus is a valuable addition to the field and it’s open source design principles using (mostly) market available components is to be commended. The inclusion of the script GUI_Picarro.py (in the supplemental repository "Software/Scripts/") for users of Piccaro CRDS which, is also of great use to users of that analysis approach given the increase in objectivity of measurement it can provide as well as providing the impetus to openly report on measurement parameters, even without the VapAuSa hardware.
Overall, I suggest mostly minor revisions regarding the use of appropriate terminology, the addition of citations for some strong statements contained in the manuscript which currently lack needed support, and some improvements to clarity. There are a few more pressing questions regarding testing for memory effects which need to be addressed. This concise manuscript provides a valuable new technology for users of DVE-CRDS and is appropriate for publication in HESS given the improvements I suggest in the specific comments below.
My review of the supplemental material containing the list of materials, technical drawings, circuit diagram, code, and manuals, necessary for building the VapAuSa” was cursory given my familiarity with some of the file types but the repository appeared to contain all of the required materials and information needed to construct the apparatus. I must admit the construction of such an apparatus falls outside of my expertise so I cannot be fully certain of any missing components, but the repository is well organized and in concert with the figures in the paper looks to provide an excellent starting point for one wishing to replicate this technical note’s apparatus. With the exception of my minor specific comments below the design section of the manuscript clearly details the apparatus, its construction, and function of its components.
Section 2: Specific Comments:
1. Throughout the manuscript the authors use the term ‘stable water isotopes’ while this is a common shorthand, it is incorrect. There are no isotopes of water (molecules do not have isotopes), rather there are isotopes of atoms, or elements (hydrogen, oxygen, nitrogen and so on). The phrase ‘stable isotopes of hydrogen and oxygen’, is more precise (and correct). To quote Zachary Sharp’s 2017 text Principles of Stable Isotope Geochemistry:
“ Writing about 'water isotopes' may sound short and concise, but it is wrong. Just as petrologists don’t talk about ‘rock isotopes’, so hydrologists should avoid talking about ‘water isotopes’.”
Table 2.1 in that text is an excellent resource for correcting often misused terminology. I suggest that the authors review that table and make corrections throughout their manuscript regarding the use of a number of terminological inconsistencies. Some of those issues are noted under this comment:
Introduction first sentence: Same stable water isotopes language issue referenced in comment 1, followed by the symbols for isotope ratios. It would be good to standardize the language throughout the manuscript in regard to isotopes vs isotope vs isotopic vs isotope ratios, and so on. From Zachary Sharps 2017 book:
“2.2.1 ‘Isotope’ vs. ‘Isotopic’ These two words appear to be used randomly and interchangeably because the proper use of one or the other is not immediately clear. My mentor Jim O’Neil, was confronted with this dilemma as a U.S.G.S. employee. He consulted the Technical Reports Unit at the USGS for guidance. After some research, it was decided that ‘isotope’ is used when modified and ‘isotopic’ is used as a stand-alone adjective. One therefore should write “The oxygen isotope composition of . . .” and “The isotopic composition of .. .”. In the first case, ‘oxygen’ modifies ‘isotope’ and in the second ‘isotopic’ stands alone"
If using the symbols for isotope ratios δ18O and δ2H, I would modify the first sentence to say: “Stable isotope ratios (δ18O and δ2H) have found…”. Alternatively, one could say “Stable isotopes of hydrogen and oxygen (2H,1H, 18O, 16O) have found…”, though I would lean towards the first.
Line 130-134: The authors use commonly applied shorthand’s that are carry overs from oral communication: ‘Heavy, and light water’ in this case. I again refer the authours to Table 2.1 in Zachary Sharp’s 2017 text in regard to more precise terminology:
“ As numbers, δ -values can be high or low, positive or negative, but not heavy or light”(Sharp 2017 pg 2-3),
However, that said, I know what you mean when you say ‘heavy or light isotope values’, as do most other folks in our field, so I would leave the decision regarding this use of this terminology in the hands of the Editor, but I think we should strive to be as precise as possible with our language. I suggest the authors check for and address this issue throughout.
Line 165:Terminology issue: depleted signature and enriched signature, use precise terminology per above.
2. Introduction paragraph 2 line ~30 “While most isotope analyzers…” I am confused by the meaning of this sentence. You seem to be saying two things here. First that isotope analyzers are not made for vapour analysis (?), and second that extraction of water from plants or soils is not possible (?)…but you then go on to describe existing extraction methods. Your first statement is not wholly correct as many isotopic analysis systems are built to analyse vapour or liquid samples. Some IRMS methods use vapour equilibration (CO2 and H2 equilibration) and both OA-ICOS and CRDS systems have vapour analysis modes (though not all of them do, especially older models). Are you trying to say that there is not an in-line isotope analysis systems which combines extraction and analysis? Because if that is the case, there is the Picaro induction module which is a peripheral system affixed to a CRDS analyzer that heats samples and inject the water vapour directly into the analyzer. I suggest this sentence be revised to improve clarity of meaning and to represent existing methodologies more accurately.
3. Line 30 &31: Do you have a citation (or other data) which concretely indicates the popularity of CVD and DVE? And the rise in popularity of DVE?
4. Line 35-47: This paragraph is leaning heavily on Grahler et al 2021. I suggest attributing the operating methodology described here to some of the original developers (Wassenaar, Hendry, Barbour, Pratt, etc). The use of silicone blot is an excellent idea though this is the first I’ve encountered it which is to say again a citation showing the introduction of this component would be useful (unless of course you are the progenitors of this technique). I have also seen a loop system used during DVE analysis where after inserting the needle and tubing to the sample bag for the inlet side connection to an OA-ICOS, a second connection is made with a needle and tubing connected to the analyzers outflow port (See Gaj et al 2019 DOI: 2136/vzj2018.04.0083). The latter technique allows for a continuous looping of the vapour sample through the analysis chamber and more stable readings.
5. Line 49-50: I don’t find this statement to be entirely accurate. The Garvelmann study used 8 cm long soil cores (relatively large depending on water content) and the Wassenaar study used water volumes in both soil samples and liquid water samples around 20 ml. There are two main types of CVD. A larger glass and manifold style system like West et al 2006 (DOI: 10.1002/rcm.2456) or Orlowski et al 2013 (DOI:10.5194/jsss-2-179-2013), and a small capillary and vial systems described in Koeniger et al 2011(DOI: 10.1002/rcm.5198) and Millar et al 2018 and 2019 (DOI: 1002/rcm.8530). The Koeniger-style CVD system uses samples volume much smaller than would be required for DVE and in some cases depending on water content the sample volumes used in the larger West/Orlowski-style CVD system may be smaller than required for DVE. Sample size is contingent upon water content. Wassenaar 2008 discuss the required water content needed to limit any atmospheric effects. Hendry et al 2015 (Doi: DOI: 10.5194/hess-19-4427-2015) make mention of issues related to low water content. The Hendry study noting >5% GWC is needed for accurate measurement. In my experience with the DVE system, sample volumes need to be larger than I have used for both types of CVD method (again water content dependant). My main concern with the author’s statement is that by saying “…small sample volume” they don’t account for water volume requirements. I would change the language to exclude the use of ‘small’ and instead indicate that DVE can use sample volumes similar to those used in CVD depending on pre-extraction testing of water contents. This is important to not obfuscate that fact that in some cases relatively large sample volumes of soil or plant material could be required. The second part of this statement is accurate though: one of the great parts of DVE is that it can allow for high spatio-temporal resolution.
6. Line 55-56: The end of this sentence is not accurate. The research by Nehemy et al 2019 does not involve correction of isotopic data generated by DVE-LS as impacted by co-extracted VOCs. That research (and their follow up work in Millar et al 2021 (1002/rcm.9118)) were focused on developing a method for detecting spectral interference during DVE-LS analysis since a tool for detection of spectral contamination during vapour analyses for OA-ICOS did not yet exist. They do not suggest a correction approach using their detection method. Revisions here are needed to correct for the inaccuracy of attributing the Nehemy et al reference with creating isotopic data correction techniques for spectrally contaminated samples.
7. Line 59: Yes, I agree! The lack of standardization is a major issue across many aspects of ecohydrological study. The need for improvements in standardization of methods has been previously discussed in Ceperly et al 2024 (DOI: 10.1002/wat2.1727) and in Millar et al 2022 (DOI: 10.1002/hyp.14698).
8. Line 70: Regarding the use of ambient air for flushing and memory effect. I see below that you used a drying canister during air flushing testing but found no improvements to your data. Our laboratory uses drierite canisters in between analysis of each sample to lower the analysis chamber water contents and to limit any issues related to memory effects. I am curious if during your testing you discovered any issues related to memory effects or if you did any testing to check for memory effects, given the use of ambient air for flushing? Did you run each sample more than once, or stager your samples so that waters with different isotope values were ran in an intermixed fashion? I would be concerned about any impacts from water in the ambient atmospheric air used during flushing given humidity levels can change drastically throughout a day, which may impact your measurements or contribute to some memory effects. I would also like to know if you tested for memory effects. The IAEA WICO survey from Wassesnaar et al 2021 (DOI: 10.1002/rcm.9193) indicated that some of the best performing laboratories were those that carried out multiple injections per sample keeping only the latter readings as a means of dealing with memory effect. The WICO tests were for liquid analyses, but the issue of memory effect is still relevant here, especially with such a high sample throughput!
9. Line 84: 391 samples! Wow! An incredible improvement on daily sample throughput. This is great.
10. Line 106: Beginning “Within an active sequence…”. I am slightly confused about the flushing sequence. If I am understanding this sentence and the preceding set up explanation, the CRDS is drawing air into its analysis chamber from the open ambient air flushing valve though the valve block to flush out any potential water, with the H2O value being measured in the CRDS analysis chamber yes? Or is there another senor in the valve block which measures the customizable H2O value? This needs clarification and is perhaps contradicting your discussion section 4.5 “Don’t buy Cheap” which indicates that ambient air can cause measurement errors. Please clarify this. Second, when you say customizable H2O value, as user of such systems I am aware of this water values function (indicates the water content in the analysis chamber), but it may be worth adding a short sentence explaining what the H2O value is for non-expert users.
11. Line 127-128: How was airtight vapour flow ensured? If a specialized check was used here (beyond the use of the silicone blots), it bears mentioning so others may attempt to replicate this.
12. Line 170: As a counterpoint to the statement about accuracy vs measurement repeatability. Its bears noting that one can only present accuracy data if they have a know truth value to compare the subsequent measurements against. In cases where water to water extractions are carried out, truth values can be collected and accuracy assessed. However, if a study is using plant material, or soils, (and not carrying out a spiking experiment with all its well discussed issues Gaj et al 2017, Thielman 2019, Oerter et al 2014 etc. all reviewed in Millar et al 2022), a truth value may not exist and so repeatability is the only option is to present uncertainty and error. Wassenaar et al 2021 (Doi: 1002/rcm.9193) discuss the standards of error propagation calculations that we should all aspire to.
13. Line 176: on your new systems accuracy: Your accuracies fit within the ranges described for other DVE publications in Table 2 in Millar et al 2022 which looks to have added new publications for vegetation extraction and new methods from what was first shown in Springer et al 2015. Allen and Kirchner 2022 (DOI: 10.1002/hyp.14483) discuss the issue of accuracy/error and range of end members, while lower accuracy is not always desirable, high throughput has a benefit and, in some cases, depending on end member range it may not be a huge issue if the error is 7‰ for δ2H for example.
14. Line 180: The Orlowski study used zip-top plastic bags as opposed to the new standards set by Grahler: (aluminum laminated with plastic coffee-style bags) and discussed through a lens of diffusion rates in Millar et al 2022). I expect the cost of these aluminum bags ,while higher than standard market available resealable zip-top bags, are still lower than the cost of the consumable glassware, and liquid nitrogen utilized in the CVD approach. Are there any newer citations which also discuss cost and time of set up given Grahler’s findings on appropriate bags? If no citation exists can the authors discuss briefly a cost comparison from their own use of such systems?
15. Line 183: The automation of this process will be a boon to users of the DVE-CRDS system! If only it existed for OA-ICOS as well.
16. Comments for Section 4.3 on storage time:
Line 186: “Uncertainty might cancel out due to lower sample storage time”. This kind of speculation is not useful without data to back up it up. It is hopeful to say that may be the case but as scientists we do not rely on might or could be. I would revise this sentence or provide data/ citations which support it.
Line 187: “Over time these changes appear with every…” Do you have a citation for this? Such a statement requires one. I know that some users heat seal their Al-Plastic bags as opposed to using the zip-top. Is there any data on leaks from heat sealed bags?
Issue continued in line 190: Again scientist should not assume, they should provide data or citations which confirm their reasoning. Research has been carried out assessing temperature effects on isotopic drift. Millar et al 2022 thoroughly review storage material effects and temperature effect of the isotope values of stored samples. I suggest the authors review the references detailed in that work instead of making assumptions regarding freezer storage as some research had already been done regarding this issue.
17. Section 4.4 On objectivity. This is very good. The field need more transparent reporting of all parameters applied during extractions. Extraction parameters should be reported in all publications using them.
18. Line 200: The statement regarding operator body heat seems speculative, but I could be wrong. Do the authors have data or citations to support this?
19. Section 4.5 Don’t buy cheap: Mentions that even a small amount of ambient air leaking into the system can cause errors dung vapour analysis. But this contradicts the use of ambient air to flush their system. Given my comments above regarding clarification of the ambient air flushing system, if ambient air is not in fact pulled into the measurement chamber during flushing, then this is not a contradiction. Clarification is needed above so the readers are fully certain that ambient air is not being pulled into the measurement cavity during flushing of the block. However, if ambient air is pulled in the measurement chamber during flushing, then this section is contradictory to the above design. Please clarify.
20. Line 217: The authors should not suggest use of cheap plastic bags in their conclusion. Grahler's research on appropriate materials for sample storage and analysis during DVE-LS seems conclusive as does the discussions in Millar et al 2022 indicating that zip-top plastic bags are not appropriate for use with DVE-LS. The addition of this statement in the conclusion does not take into account the fact that some samples collected in zip-top bags may undergo periods of storage or require transport (time) which increase evaporative fractionation risks. Introducing this idea in the conclusion is not appropriate but should they wish to make this point it should be expanded upon in the discussion and the leak risk compared against other already published studies which indicate these types of bags are not appropriate.
Section 3: Technical Corrections:
1. Language correction: In abstract ~ line 5: “However, sample analysis requires a significant manual labor, thereby limiting the number of samples that can be analyzed.”. Remove the ‘a’ between ‘requires’ and ‘significant’.
2. Language correction: Introduction paragraph 2 line ~30 “While most isotope analyzers are build…” should be ‘built’ not ‘build’.
3. Introduction paragraph 2. I think the structure of this paragraph could be reworked to improve readability. You begin with a confusing statement which is not accurate depending on your meaning (discussed in my comments above), briefly mention DVE, then immediately change topic to discussing CVD and how it works. The paragraph ends and the next takes up the discussion of DVE. I understand that you may want to mention the popularity of DVE and CVD together, but I find this disjointed.
4. Line 49: “Additionally on a small sample…” space required between volume and similar
5. Line 30-31: DVE-LS is only partially defined. I know that LS = laser spectroscopy but other readers may not, so the acronym needs to be fully defined.
6. Line 74: Language issue for sentence: “For 23 sample bags are connected with cannulas to the valves through 1/8 inch PTFE tubing”. The use of ‘For’ at the start of sentence is incorrect or something is missing later in the sentence. Depending on desired meaning ‘For’ could be removed, simply saying “23 sample bags can be connected via cannulae to…”
Citation: https://doi.org/10.5194/egusphere-2024-528-RC1 -
RC2: 'Comment on egusphere-2024-528', Anonymous Referee #2, 29 Mar 2024
Thank you for letting me review this manuscript. The authors present an autosampler for measuring the water isotope composition of water (and potentially, soil and plant) samples using the direct vapor equilibration method. I know and appreciate the work of the group and I am user, hence, very familiar with the DVE-LS method as well. Providing an open-source autosampler for increasing sample throughput and minimizing labor and cost is simply amazing and I applaud the others for developing and providing all codes publicly. However - and I am sorry to say this - the presented study has severe inconsistencies and shortcomings, which simply need to be addressed before publishing. In brief, I could summarize these as follows: i.) for a convincing proof-of-concept, the results for measuring 21 samples is simply insufficient; ii.), the ms. refers to soil samples, CVD and artifacts throughout, but does not actually report any results for using the autosampler on soil sample equilibrations, which would be much more interesting and relevant for the actual application of the method; and iii.) potentially relevant factors were not tested for nor commented on (e.g., the effect of leaving the samples to be measured pinched by the canules). Additionally, the structure of the ms is currently confusing and is missing a red line (mainly due to ii.) I believe). I hope my comments and suggestions help in improving the ms. Having that said, I totally support publication of the autosampler, after the data foundation for a true objective evaluation is improved.
General/Main comments:
- Were only 21 samples measured in total? This is very low for assessing accuracy/uncertainty and makes me seriously wonder – the authors have the autosampler, why would they rely on such a small number of samples for proofing its accuracy and uncertainties?
- The manuscript mentions soil samples, CVD, and artifacts introduced by extraction multiple times and quite prominent; however, no soil data is presented in the manuscript at all. This is confusing me extremely. Firstly, CVD is not of relevance for the autosampler at all for the purpose of proof-of-concept. Secondly, I would’ve loved to see comparisons between manual injections and autosampler results for soil samples! I don’t understand why this is not part of the study.
- 155-165: This chapter is completely unrelated to the actual study, which was done using liquid water, not soil samples. Also, the ‘bashing’ of CVD and explanation of the differences or not of any importance for the presented study. I strongly suggest to focus on comparing and testing the accuracy and uncertainty of the data obtained with the autosampler vs. manual injection and liquid water measurements for comparable matrices, i.e., liquid, non-extracted water.
- Same as above is true for chapter 4.2: The study did not compare data obtain from CVD with data obtained by DVE-LS – why is this discussed in the manuscript? I suggest again to focus on the actual purpose of the study: Provide an autosampler for water vapor isotope measurements and proof its accuracy, benefits and caveats. The methodological artifacts introduced by such a study must be eliminated, i.e., by only using water samples.
- Avoid colloquial language and vague statements (“some”, “lots of”, e.g., l.195)
Minor comments:
- Abstract:
- Vapor or vapour? à use consistent BE or AE
- 12: the reported uncertainty of both manual and automated measurement for d2H seems pretty high, compared to liquid injections. It would be great to see some elaborations on the reasons and/or recommendations to improve this somewhere in the ms.
- Please report here on the methods also (how was the reported accuracy etc. elaborated – how many samples of which matrix were tested)
- Why is the later mentioned accuracy not reported in the abstract? (and also, repeatability)
- Text:
- 20: obtain estimate
- 24: distribution composition
- 26: partitioning effects à unclear what is meant with this, please rephrase or explain. Suggestion: water partitioning
- 29/30: this sentence is unclear, it is very much possible to extract water from soils and plants; isotope analyzers are also not built to measure specifically liquid water; in fact, the lasers all measure water vapor. The autosamplers though are build for liquid water, I guess this is what is meant here. Please rephrase.
- 31: DVE is not really an extraction, it’s an equilibration
- 32-34: the explanation of CVD is explained very sloppy; from the current explanation it is absolutely not clear how CVD works. Also, the citation used is very old – the method has been much improved meanwhile (classic citation Koeniger et al., 2011, but also many others)
- 38: For other users, a reference to the company and size of the laminated bags would be useful here
- 49: volume similar (space missing)
- 49/50: CVD requires much less soil compared to DVE-LS, only 10g are sufficient for one replicate in many instances. Having that said, it should be stated somewhere how much soil is needed. Also, the constraint for high spatial and temporal sampling is often the digging, not the sample storage or amount.
- 74: please provide details of the cannulas (diameter) – it might be a constraint for providing the 35ml/l to the Picarro and induce errors if diameter is too small
- 127: It seems like the cannulas all have to be connected to the bags in the beginning of the measurement cycle. This means that the later they are in the sampling line, the more potential exists for contaminating the sample because the sample was pinched. It would be great to have some comments on this. Also, silicone might affect the absorption spectra of the measurements – could this be explanations for the relatively high uncertainties of the method? It would be nice to see some elaborations on this.
- 129: Just referring to the table is not sufficient, please explain what the stability criteria actually checks.
- 135: next to uncertainty, the accuracy of the method should also be reported in the abstract
- 152/153: Perhaps the authors are right, but this needs to be further discussed. As stated before, there are potential influencing factors which could cause a difference of the manual vs. automated injections; e.g., the time until a sampling is analyzed (intrusion of atmospheric air?) and the potential contamination with silicone (could affect only one isotope and hence, explain different behavior of d2H and d18O).
- 173/174: I agree that ONLY reporting repeatability is masking the actual outcomes, but not checking for repeatability is equally questionable when providing a proof-of-concept. Why not report both?
- General suggestion for the discussion: Try to stay objective in the statements. Reading the discussion, I felt that the authors are trying to convince me how good the proposed method is, rather than objectively reporting and discussing the pros and cons.
- 4.3: The elaborations here are good, but please comment on the potential effect of pinched bags during the measurement cycle
- 216: may be is ;)
- 219/220: “So far we tested the system with liquid and soil samples only” Why are these results not presented here?
Citation: https://doi.org/10.5194/egusphere-2024-528-RC2 -
RC3: 'Comment on egusphere-2024-528', Fabio Marzaioli, 09 Apr 2024
The authors have developed an automated system, called VapAuSa, for water isotope measurements by direct vapour equilibration (DVE-LS). The system is described with a lot of detail, but the data reporting and commenting is still poor in my opinion. I see a strong problem in the definition of accuracy and the results obtained are hardly commentable in this sense. The data are there, but some work should be done to produce scientifically meaningful results.
Detailed comments.
Abstract
Ln 7: What do the authors mean for lack of objectivity?
Ln 12: Can you please specify what is your precision estimator? Are they comparable according to a Fisher test?
Intro
Ln 29: I would suggest not using “isotope analysers”. It goes in the verse of a measurement black box while if one knows principles of measurements for apparatuses can be a great help to produce accurate data.
Ln 29: Please avoid the usage of isotopic composition (refer to isotope ratio od delta).
Ln 32: Please rephrase, how does a vial contain vacuum? The whole process happens under vacuum.
Ln 37: add Gas before tight.
Ln 38: I would not see the 2) as a different point form the 1).
Ln 42: Isotopically Equilibrate: change to “can reach isotope equilibrium”.
Ln 49: volumesimilar: volume similar?
Par 3.1
Ln 123: 3 not three.
Table 1: Tables can not be used without IS units.
Period starting at line 130: There should be uncertainties for the reported water samples. Only primary standards are reported with no uncertainties.
Ln 135 and soon after: Accuracy cannot be a difference between 2 values. Of course, since a and l are measured, they are affected by uncertainties. Therefore, an unbiased way to define accuracy is to define a variable whose numerator is the reported difference (accuracy d18O formula) and whose denominator is the propagated uncertainty of such a difference.
Par 4.1 (Accuracy)
The section is not homogeneous. The introductory part should be included in the intro. Also, if some authors (ln 170) used repeatability as an indicator of accuracy, it should be stated that this is a misinterpretation. Also, it is not a matter of ranking accuracy vs repeatability, they are 2 different concepts.
Citation: https://doi.org/10.5194/egusphere-2024-528-RC3
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 195 | 44 | 21 | 260 | 47 | 10 | 11 |
- HTML: 195
- PDF: 44
- XML: 21
- Total: 260
- Supplement: 47
- BibTeX: 10
- EndNote: 11
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1