the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 15 Aug 2024
OF-CEAS laser spectroscopy to measure water isotopes in dry environments: example of application in Antarctica
Abstract. Water vapour isotopes are important tools to better understand processes governing the atmospheric hydrological cycle. Their measurement in polar regions is crucial to improve the interpretation of water isotopic records in ice cores. In situ water vapour isotopic monitoring is however an important challenge, especially in dry places of the East Antarctic plateau where water mixing ratio can be as low as 10 ppmv. We present in this article new commercial laser spectrometers based on the optical feedback – cavity enhanced absorption spectroscopy (OF-CEAS) technique, adapted for water vapour isotopic measurement in dry regions. We characterize a first instrument adapted for Antarctic coastal monitoring with an optical cavity finesse of 64,000 (ringdown time of 54 µs), installed at Dumont d’Urville station during the summer campaign 2022–2023, and a second instrument with a high finesse of 116,000 (98 µs ringdown), to be deployed inland East Antarctica. The high finesse instrument demonstrates a stability up to two days of acquisition, with a limit of detection down to 10 ppmv humidity for 𝛿D and 100 ppmv for 𝛿18O.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(1643 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1643 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-2149', Harro A.J. Meijer, 28 Aug 2024
This is a well-written, interesting paper about the 'next level' OF-CEAS instrumentation. I liked reading it, and is carries important new information. I recommend publication, but/and have a set of comments and questions that I invite the authors to respond to, preferably by adapting the paper where appropriate, or else in writing to the editor (and me).
First a general question: how come it is so difficult to perform isotope measurements on water vapour at 10 ppm or even higher, while for example precise isotope measurements on methane (2 ppm) are routine nowadays? Is it because of the humidity range that has to be accomodated?
Then in more detail, going through the paper:
Figure 1 The methane interference can potentially be severe for low humidity (1/100 to 1/1000 of the humidity in this plot). Apart from this plot, this interference has not been discussed or even mentioned anywhere else. Is the methane spectrum always included in, and corrected for in the fits?
Figure 2 It is not clear to me how the residuals in cm-1 relate to the arbitrary normalized intensity (in arbitrary units). Even the x-axis is in free spectra range steps, not in cm-1. Please clarify.
Compared to figure 1, there is a flat baseline here in figure 2. How can that be, as all lines in figure 1 will be proportionally lower?
Line 169 and elsewhere: suddenly decimal commas instead of points. Happens more often, with the declaration of the values for reference waters.
Figure 5 the number of points for the Picarro instrument are much higher than for the AP2E one. Why? DIfferent strategy?
Table 1 should include the number of measurements (so much more for the Picarro I presume)
Lines 190 further. This humidity / mixing ratio dependence is widely observed, not only for water vapour measurements, but also for isotopes in atmospheric CO2. In that field, two ways of dealing with it are in use: the 'ratio method', so building ratios first, and then correct for mixing ratio dependence (this is what you do here), the other method is the isotopologue method, which would first analyze the different isotopologues as different species, calibrating first them, and only then build ratios. See for instance papers:
Flores, et al. Calibration Strategies for FT-IR and Other Isotope Ratio Infrared Spectrometer Instruments for Accurate delta C-13 and delta O-18 Measurements of CO2 in Air. Analytical Chemistry {89}, {3648-3655} (2017).
Steur et al. Simultaneous measurement of δ 13C, δ 18O and δ 17O of atmospheric CO2 – performance assessment of a dual-laser absorption spectrometer. Atmos Meas Tech 14, 4279–4304 (2021).
Both have their pros and cons.
Would that be something to try out here? Or at least to mention and discuss. Or is it not applicable at all in this context?FIgure 8 The Allan deviation is only part of the final uncertainty: the humidty correction and the calibration error also contribute. While that is less important for tracking the diurnal cycle, the calibration error will influence the accuracy of the seasonal cycle.
Line 300 see above: the humidity dependence of dD is relatively speaking larger than that of d18O, and also it calibration uncertainty (fig 7) is, relatively, larger. Would that influence your conclusion?
Line 342 "applying a drying on the humidity generator" ?? unclear to me what you mean.
Lines 355-365 (and figure 6) I agree with your conclusion that while the gradual, linear dependence is indeed caused by a spectral 'misfit' (probably indeed the interference of the very strong lines further in the spectrum, or methane?), the low humidity part strongly indicates sample-to-sample memory effects. This is further supported by the fact that the depleted ref goes up, and the 'enriched' one goes down.
This effect will not fully vanish in the station, because (1) the residual water vapour is probably quite fractionated, and will thus still be different from outside humidity, and (2) you must calibrate your instrument regularly using two waters with very different isotope values.
I would like to see more discussion of these low-humidity part effects (after all, that is the truly innovative part of the instrument and your paper): for example, are the values (in fig 6) influenced by the 'sample history' before one of the reference waters and how would that be different in the station? Would you expect that sample water vapour with isotope values intermediate between the 'high' and 'low' ref waters scale linearly between them (as you suggest in lines 250-253)? What is the added uncertainty in this region?Line 364-365: 'We insist thus on the importance of calibrating the instrument in the field to correct for those artefacts (Casado et al., 2016).' Indeed! May be change the word 'insist' into 'emphasize' ?
line 371-372 "The internal architecture of these analysers therefore reduces the risk of breakdowns during the deployment, but requires an expertise to finely tune them. "
To me this feels like a blessing in disguise, or may be just the exact opposite. Deploying these instruments thus always requires expertise (and time and some equipment) in the field. Any more comments to that? For example negative field experience with fixed mounted equipment such as the Picarro's ?
"Competing interests. The authors declare that they have no conflict of interest."Two of the co-authors work with the (commercial) producer of the instrument. How do they avoid a conflict of interest?
Citation: https://doi.org/10.5194/egusphere-2024-2149-RC1 -
AC1: 'Reply on RC1', Thomas Lauwers, 18 Oct 2024
We thank Harro Meijer for his very pertinent questions concerning the general form of the manuscript as well as the scientific content, which allow us to be exhaustive on the instrumental limits presented in the paper, and give us interesting perspectives for future studies. The answers are in the file attached: they are inserted in the text in red and the citations from the manuscript are indicated in blue.
-
AC1: 'Reply on RC1', Thomas Lauwers, 18 Oct 2024
-
RC2: 'Comment on egusphere-2024-2149', Anonymous Referee #3, 10 Sep 2024
The manuscript titled "OF-CEAS laser spectroscopy to measure water isotopes in dry
environments: example of application in Antarctic" by Lauwers et al. is a much needed study on a promising technique to measure water vapor isotope composition at very low humidity. The manuscript includes all of the information one would hope for to employ their method and requires only minor revisions before publication. Indeed, the authors would go a long way towards publication readiness by clarifying the text and refining the grammar.General comments for clarity:
- Make the figures and tables, including their captions, more independent of the main text. Telling a reader the same information twice (e.g., in the figure and in the main text) helps with clarity.
Instrumentation description:
- I expect this study and instrumentation to be popular among water isotope laboratories and researchers. As such, please describe the specific model of AP2E used. Or, if all three are custom, please indicate so. Consider including a brief description of how these particular OF-CEAS instruments differ from those in Casado et al., 2016. Are they the same?
- Use of the description "V-shape" does not help the reader understand the instrument. Perhaps it would be more informative to say something about the lack of fiber optic cabling in favor of highly reflective mirrors?
2.1.2 Long-term stability at Dumont d’Urville station:
- Do the authors expect the variable water vapor concentration requiring "An additional filtering has been applied to remove the points with a non- stable humidity, i.e. with a humidity value standing outside the 2-𝜎 interval." is due to the LHLG unit or to the OF-CEAS instrument itself?Figure 5:
- De-emphasize the green data by making it a lighter shade and placing them behind the OF-CEAS data.
- I suggest expressing the y-axes as residuals in the same way the authors have already done with Figs 3 and 6. The wide range that must necessarily be used when expressing the values in their absolute terms may hide drift or biases.
- The caption sentence "Each point represents the average of the final minimum 5 minutes of 1000 ppm humidity plateaus and the error bars correspond to the associated standard deviation" is unclear. Does "minimum" imply that sometimes it is longer? Be specific.
- Why would "plateaus" exist with this particular dataset? The caption currently reads as if all of these data were collected at 1000 ppm. Please clarify.
Table 1:
- If these variance estimates are from the data presented in Fig 5, I suggest deleting this table and stating these values in the Fig 5 caption text. If the authors choose to keep the table, state in the table caption that the variance estimates are from the data presented in Fig 5. If they are not from the data in Fig 5, please describe their origin.Figure 6:
- Please describe in the caption what is meant by "ref" in the y-axis titles.
Table 2:
- Does "lightly depleted" and "highly depleted" refer to standards shown in Fig 6? Please describe in the table caption what is meant, or better yet, remove the ambiguous terms in favor of the actual reference water names.
Figure 8:
- The caption is confusing. The authors are not plotting Allan deviation after 24 hours, they are plotting the predicted standard deviation after 24 hours of integration as predicted from an Allan variance analysis.
- The wording describing the dotted lines is confusing. It currently sounds as if they are the predicted range one would expect during the course of a 24 hour period. The main text, however, suggests these dotted lines are the authors desired "noise threshold".
- Is this threshold one standard deviation or perhaps a 95 % confidence interval? Please reword the caption.
- Make the right side axis title and tick labels red, in the same way the authors made the left side blue.
- Lastly, the font size of the stations on the map need to be much larger.
Section 3.1
- This section would be better positioned after the discussion since it is a proposed application of what the authors intend to do after having completed the current study.Figure 9:
- Similar to my comment for Fig 8, reword "The noise is obtained from short-term Allan deviations at 𝝉 = 2 min..." to something like "Standard deviation is predicted from an allan variance analysis for the 2 minute integration time...".
Discussion Line 325:
- The sentence "It shows an optimum stability range of ~ 15 min, followed by a drift period in the hour range and finally a stabilisation of the signal in the day range" is confusing. Are the authors referring to the #1169 instrument at 500 ppm H2O and d18O? The blue series is the only series that does seem to be achieving stability in the day range. All others continue to drift or have insufficient data to make any conclusions about their trajectory (e.g., dD #1169).
Interesting features for field operation:
- This is an important section and seems to compare OF-CEAS to Picarro's CRDS. The glue, the inability to clean, lack of needed software. I suggest the authors state plainly that AP2E's design and software are more conducive to remote field deployment than Picarro's CRDS.
Citation: https://doi.org/10.5194/egusphere-2024-2149-RC2 - AC2: 'Reply on RC2', Thomas Lauwers, 18 Oct 2024
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-2149', Harro A.J. Meijer, 28 Aug 2024
This is a well-written, interesting paper about the 'next level' OF-CEAS instrumentation. I liked reading it, and is carries important new information. I recommend publication, but/and have a set of comments and questions that I invite the authors to respond to, preferably by adapting the paper where appropriate, or else in writing to the editor (and me).
First a general question: how come it is so difficult to perform isotope measurements on water vapour at 10 ppm or even higher, while for example precise isotope measurements on methane (2 ppm) are routine nowadays? Is it because of the humidity range that has to be accomodated?
Then in more detail, going through the paper:
Figure 1 The methane interference can potentially be severe for low humidity (1/100 to 1/1000 of the humidity in this plot). Apart from this plot, this interference has not been discussed or even mentioned anywhere else. Is the methane spectrum always included in, and corrected for in the fits?
Figure 2 It is not clear to me how the residuals in cm-1 relate to the arbitrary normalized intensity (in arbitrary units). Even the x-axis is in free spectra range steps, not in cm-1. Please clarify.
Compared to figure 1, there is a flat baseline here in figure 2. How can that be, as all lines in figure 1 will be proportionally lower?
Line 169 and elsewhere: suddenly decimal commas instead of points. Happens more often, with the declaration of the values for reference waters.
Figure 5 the number of points for the Picarro instrument are much higher than for the AP2E one. Why? DIfferent strategy?
Table 1 should include the number of measurements (so much more for the Picarro I presume)
Lines 190 further. This humidity / mixing ratio dependence is widely observed, not only for water vapour measurements, but also for isotopes in atmospheric CO2. In that field, two ways of dealing with it are in use: the 'ratio method', so building ratios first, and then correct for mixing ratio dependence (this is what you do here), the other method is the isotopologue method, which would first analyze the different isotopologues as different species, calibrating first them, and only then build ratios. See for instance papers:
Flores, et al. Calibration Strategies for FT-IR and Other Isotope Ratio Infrared Spectrometer Instruments for Accurate delta C-13 and delta O-18 Measurements of CO2 in Air. Analytical Chemistry {89}, {3648-3655} (2017).
Steur et al. Simultaneous measurement of δ 13C, δ 18O and δ 17O of atmospheric CO2 – performance assessment of a dual-laser absorption spectrometer. Atmos Meas Tech 14, 4279–4304 (2021).
Both have their pros and cons.
Would that be something to try out here? Or at least to mention and discuss. Or is it not applicable at all in this context?FIgure 8 The Allan deviation is only part of the final uncertainty: the humidty correction and the calibration error also contribute. While that is less important for tracking the diurnal cycle, the calibration error will influence the accuracy of the seasonal cycle.
Line 300 see above: the humidity dependence of dD is relatively speaking larger than that of d18O, and also it calibration uncertainty (fig 7) is, relatively, larger. Would that influence your conclusion?
Line 342 "applying a drying on the humidity generator" ?? unclear to me what you mean.
Lines 355-365 (and figure 6) I agree with your conclusion that while the gradual, linear dependence is indeed caused by a spectral 'misfit' (probably indeed the interference of the very strong lines further in the spectrum, or methane?), the low humidity part strongly indicates sample-to-sample memory effects. This is further supported by the fact that the depleted ref goes up, and the 'enriched' one goes down.
This effect will not fully vanish in the station, because (1) the residual water vapour is probably quite fractionated, and will thus still be different from outside humidity, and (2) you must calibrate your instrument regularly using two waters with very different isotope values.
I would like to see more discussion of these low-humidity part effects (after all, that is the truly innovative part of the instrument and your paper): for example, are the values (in fig 6) influenced by the 'sample history' before one of the reference waters and how would that be different in the station? Would you expect that sample water vapour with isotope values intermediate between the 'high' and 'low' ref waters scale linearly between them (as you suggest in lines 250-253)? What is the added uncertainty in this region?Line 364-365: 'We insist thus on the importance of calibrating the instrument in the field to correct for those artefacts (Casado et al., 2016).' Indeed! May be change the word 'insist' into 'emphasize' ?
line 371-372 "The internal architecture of these analysers therefore reduces the risk of breakdowns during the deployment, but requires an expertise to finely tune them. "
To me this feels like a blessing in disguise, or may be just the exact opposite. Deploying these instruments thus always requires expertise (and time and some equipment) in the field. Any more comments to that? For example negative field experience with fixed mounted equipment such as the Picarro's ?
"Competing interests. The authors declare that they have no conflict of interest."Two of the co-authors work with the (commercial) producer of the instrument. How do they avoid a conflict of interest?
Citation: https://doi.org/10.5194/egusphere-2024-2149-RC1 -
AC1: 'Reply on RC1', Thomas Lauwers, 18 Oct 2024
We thank Harro Meijer for his very pertinent questions concerning the general form of the manuscript as well as the scientific content, which allow us to be exhaustive on the instrumental limits presented in the paper, and give us interesting perspectives for future studies. The answers are in the file attached: they are inserted in the text in red and the citations from the manuscript are indicated in blue.
-
AC1: 'Reply on RC1', Thomas Lauwers, 18 Oct 2024
-
RC2: 'Comment on egusphere-2024-2149', Anonymous Referee #3, 10 Sep 2024
The manuscript titled "OF-CEAS laser spectroscopy to measure water isotopes in dry
environments: example of application in Antarctic" by Lauwers et al. is a much needed study on a promising technique to measure water vapor isotope composition at very low humidity. The manuscript includes all of the information one would hope for to employ their method and requires only minor revisions before publication. Indeed, the authors would go a long way towards publication readiness by clarifying the text and refining the grammar.General comments for clarity:
- Make the figures and tables, including their captions, more independent of the main text. Telling a reader the same information twice (e.g., in the figure and in the main text) helps with clarity.
Instrumentation description:
- I expect this study and instrumentation to be popular among water isotope laboratories and researchers. As such, please describe the specific model of AP2E used. Or, if all three are custom, please indicate so. Consider including a brief description of how these particular OF-CEAS instruments differ from those in Casado et al., 2016. Are they the same?
- Use of the description "V-shape" does not help the reader understand the instrument. Perhaps it would be more informative to say something about the lack of fiber optic cabling in favor of highly reflective mirrors?
2.1.2 Long-term stability at Dumont d’Urville station:
- Do the authors expect the variable water vapor concentration requiring "An additional filtering has been applied to remove the points with a non- stable humidity, i.e. with a humidity value standing outside the 2-𝜎 interval." is due to the LHLG unit or to the OF-CEAS instrument itself?Figure 5:
- De-emphasize the green data by making it a lighter shade and placing them behind the OF-CEAS data.
- I suggest expressing the y-axes as residuals in the same way the authors have already done with Figs 3 and 6. The wide range that must necessarily be used when expressing the values in their absolute terms may hide drift or biases.
- The caption sentence "Each point represents the average of the final minimum 5 minutes of 1000 ppm humidity plateaus and the error bars correspond to the associated standard deviation" is unclear. Does "minimum" imply that sometimes it is longer? Be specific.
- Why would "plateaus" exist with this particular dataset? The caption currently reads as if all of these data were collected at 1000 ppm. Please clarify.
Table 1:
- If these variance estimates are from the data presented in Fig 5, I suggest deleting this table and stating these values in the Fig 5 caption text. If the authors choose to keep the table, state in the table caption that the variance estimates are from the data presented in Fig 5. If they are not from the data in Fig 5, please describe their origin.Figure 6:
- Please describe in the caption what is meant by "ref" in the y-axis titles.
Table 2:
- Does "lightly depleted" and "highly depleted" refer to standards shown in Fig 6? Please describe in the table caption what is meant, or better yet, remove the ambiguous terms in favor of the actual reference water names.
Figure 8:
- The caption is confusing. The authors are not plotting Allan deviation after 24 hours, they are plotting the predicted standard deviation after 24 hours of integration as predicted from an Allan variance analysis.
- The wording describing the dotted lines is confusing. It currently sounds as if they are the predicted range one would expect during the course of a 24 hour period. The main text, however, suggests these dotted lines are the authors desired "noise threshold".
- Is this threshold one standard deviation or perhaps a 95 % confidence interval? Please reword the caption.
- Make the right side axis title and tick labels red, in the same way the authors made the left side blue.
- Lastly, the font size of the stations on the map need to be much larger.
Section 3.1
- This section would be better positioned after the discussion since it is a proposed application of what the authors intend to do after having completed the current study.Figure 9:
- Similar to my comment for Fig 8, reword "The noise is obtained from short-term Allan deviations at 𝝉 = 2 min..." to something like "Standard deviation is predicted from an allan variance analysis for the 2 minute integration time...".
Discussion Line 325:
- The sentence "It shows an optimum stability range of ~ 15 min, followed by a drift period in the hour range and finally a stabilisation of the signal in the day range" is confusing. Are the authors referring to the #1169 instrument at 500 ppm H2O and d18O? The blue series is the only series that does seem to be achieving stability in the day range. All others continue to drift or have insufficient data to make any conclusions about their trajectory (e.g., dD #1169).
Interesting features for field operation:
- This is an important section and seems to compare OF-CEAS to Picarro's CRDS. The glue, the inability to clean, lack of needed software. I suggest the authors state plainly that AP2E's design and software are more conducive to remote field deployment than Picarro's CRDS.
Citation: https://doi.org/10.5194/egusphere-2024-2149-RC2 - AC2: 'Reply on RC2', Thomas Lauwers, 18 Oct 2024
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 279 | 74 | 62 | 415 | 17 | 20 |
- HTML: 279
- PDF: 74
- XML: 62
- Total: 415
- BibTeX: 17
- EndNote: 20
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 136 | 31 |
| France | 2 | 71 | 16 |
| Germany | 3 | 31 | 7 |
| China | 4 | 29 | 6 |
| Netherlands | 5 | 20 | 4 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 136
Cited
Thomas Lauwers
Elise Fourré
Olivier Jossoud
Daniele Romanini
Frédéric Prié
Giordano Nitti
Mathieu Casado
Kévin Jaulin
Markus Miltner
Morgane Farradèche
Valérie Masson-Delmotte
Amaëlle Landais
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1643 KB) - Metadata XML