the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 01 Feb 2024
Quantifying H2S with a Picarro CRDS G2201-i and the effect of H2S on carbon isotopes
Abstract. Cavity Ring-Down Spectroscopy (CRDS) is a popular analytical method with important applications in earth sciences including volcanology. A main disadvantage of using CRDS in volcanology is that the presence of H2S distorts some spectral lines causing errors in the measurements. In this study, we investigated the effects of H2S on measurements using a Picarro G2201-i instrument. We defined the interferences caused by H2S on CO2, CH4, and their carbon isotopic compositions. We found that 30 ppb H2S in 1000 ppm CO2 causes a difference of ~1.0 ± 0.2 ‰ on the δ13C-CO2 measurement, while 1 ppm H2S in 1 ppm CH4 per causes a difference of < 0.2 ‰ on the δ13C-CH4 measurement; this agrees with the results from previous studies using other models of Picarro instruments. Characterizing how H2S produces these interferences as a function of concentration, we further developed a series of equations to quantify H2S in gas mixtures in a concentration range of 1 to 270 ppm. We validated our method by analyzing a natural dry gas sample and comparing our results with those of two other independent analytical techniques, namely a CH4-MultiGAS and a “Giggenbach bottle”. When comparing the results between the CH4-MultiGAS and the Picarro G2201-i, we measured differences of ~ 4 %, while when comparing the results between the Giggenbach bottle and the Picarro G2201-i, we measured differences of ~ 9 %. The results of these three techniques show excellent agreement within error of each other. Our study demonstrates that the Picarro G2201-i instrument can accurately and precisely measure CO2, CH4, and H2S concentrations in the gas phase.
- Preprint
(997 KB) - Metadata XML
-
Supplement
(456 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on amt-2023-265', Anonymous Referee #1, 13 Feb 2024
Salas-Navarro and co-workers present an analysis of the potential interference of H2S on CH4 and CO2 carbon isotope measurements. The authors also develop an approach to quantify H2S mixing ratios. These interferences arise owing to overlap of the absorption of H2S and the 12C/13C isotopologues of CH4 and CO2. While the approach is specific to a particular commercial instrument, a Picarro G2201-I, the work will be valuable in vulcanology for quantifying H2S and improving the accuracy of CO2 and CH4 isotope measurements in environments with high H2S mixing ratios.
The authors assess the effect of H2S on CO2 and CH4 measurements across a wide range of mixing ratios of CO2, CH4, and H2S. They demonstrate an appreciable interference of H2S on the quantification of δ13C-CO2; the effect is most marked when CO2 mixing ratios are low (as would be expected). In contrast, H2S has a much smaller effect on measurement of δ 13C-CH4, and the effect is limited to conditions of low CH4 and high H2S mixing ratios. One weakness of the study, which the authors acknowledge, is that the potential interference of H2O is not considered. This work therefore applies only to dry conditions.
The paper is well written and the rationale for the work is clearly motivated. I consider that this paper is suitable for publication after addressing the following issues.
- L149: “room conditions”? Meaning the sample or physical operating environment. Presumably this means room air?
- L78-80: It is unclear how large the Picarro spectral range is and how the spectral analysis is carried out. A few details on these questions would be helpful, especially in reference to interpreting Figure 6.
- L83: Clarify whether the results of the study apply to the other instrument operating modes (CO2 only and CH4 only).
- L96: Was zero air obtained from a gas cylinder or through purification and drying external air? It seems more natural to me to provide the composition and moisture content of zero air here, and not later in L113.
- 3: The uncertainties of the dilution process cover an order of magnitude (2-20%). It is unclear how these uncertainty values are arrived at, what the uncertainty is in the aliquot and dilution volumes, and how such large uncertainties occur in the measured volumes of the components of the gas mixture. Some of the underlying details should be supplied.
- P5 l.140: Determination of H2S is central to the work reported in this paper, so a few sentences would be appropriate here to explain the approach used to correct and calibrate H2S, instead of just referencing another paper.
- L143: How was the spring gas sample dried?
- L260-271, Table 1: The presentation of “errors” and uncertainties is confusing here. The “errors” are not uncertainties as such, but the fractional difference between two measurements from different techniques. Moreover, while the Picarro measurements happen to agree closely with the Multigas measurements, the intrinsic measurement uncertainties of the Picarro ratio measurements are much larger than the difference from the Multigas measurement.
E.g., H2S/CH4: 87±33 (Picarro) vs 91±9 (Multigas). The difference in measurements (87 vs 91) is 4.4% but I understand the uncertainty in the Picarro ratio measurement to be 38% (87±33). The first figure is fortuitous agreement, while the second is a better statement of the overall uncertainty of the Picarro ratio measurement. These issues should be clarified in the text.
Minor corrections:
- L78: “spectral” spelling.
- Terminology: The term “mixing ratio” should be used in the text and figures for measurements in units of ppb and ppm, not “concentration”.
Citation: https://doi.org/10.5194/amt-2023-265-RC1 -
AC1: 'Reply on RC1', Jessica Salas-Navarro, 04 Apr 2024
Response to Anonymous Referee #1
We would like to thank “Anonymous Referee #1” for their valuable and constructive comments that allowed us to significantly improve our manuscript. Below we address all the points raised by the reviewer and we attempt to provide adequate solutions and responses to their concerns in the manuscript.
- L149: “room conditions”? Meaning the sample or physical operating environment. Presumably this means room air?
“Room conditions” does mean room air. We recognize that it might not be clear, therefore we have changed the text from “room conditions” to “room air” as suggested. We made this change in lines 120 and 134.
- L78-80: It is unclear how large the Picarro spectral range is and how the spectral analysis is carried out. A few details on these questions would be helpful, especially in reference to interpreting Figure 6.
We think it is important to highlight that our contribution comes from a user point of view, and it is addressed to users of the Picarro instrument G2201-i. For more specific details of the CRDS technique, the readers can consult our references section. Details about the software calculations for the spectral analysis are not available, at least to the best of our knowledge. Our manuscript includes all the details and information that are available in the literature about this specific instrument. Nevertheless, we did address this issue by modifying section “2.1. Laboratory conditions”. Specifically, we added more details about the instrument and the technique from the user’s guide of this instrument in lines 78 to 85. This section was renamed to “2.1. Laboratory conditions and instrument details”.
- L83: Clarify whether the results of the study apply to the other instrument operating modes (CO2 only and CH4 only).
As mentioned in line 87, all our analyses were conducted using the CO2 and CH4 simultaneous mode. We did not explore the other operating modes. Therefore, we cannot be certain of the effects of H2S on the other operating modes. However, based on the literature and our basic understanding of the operation of these instruments, we feel confident that the H2S effect is the same in the other operating modes.
We did not explore the other operating modes because as mentioned in our motivation, measuring CO2 and CH4 simultaneously is one of the most appealing and useful characteristics of the Picarro G2201-i instrument in the field of volcanology.
- L96: Was zero air obtained from a gas cylinder or through purification and drying external air? It seems more natural to me to provide the composition and moisture content of zero air here, and not later in L113.
Zero air was obtained from a gas cylinder. We provided the zero air tank composition in line 101 and not later in line 117, as you suggested.
- 3: The uncertainties of the dilution process cover an order of magnitude (2-20%). It is unclear how these uncertainty values are arrived at, what the uncertainty is in the aliquot and dilution volumes, and how such large uncertainties occur in the measured volumes of the components of the gas mixture. Some of the underlying details should be supplied.
The uncertainty at the gas mixture preparation was estimated by considering the error introduced by the different syringes used to make the mixtures, and the error from the gas standard, multiplied by 2. As shown in the following equation:
The error of each syringe was calculated with the following equation:
The order of magnitude difference in these uncertainties is attributed to the very small aliquots of our 100% gas standards (CO2 and CH4) required to create low concentration mixtures, as well as large aliquots of zero air.
- P5 l.140: Determination of H2S is central to the work reported in this paper, so a few sentences would be appropriate here to explain the approach used to correct and calibrate H2S, instead of just referencing another paper.
We modified the section “2.6 Quantifying H2S concentrations” to clarify that we were inspired by the work done by Assan et al. (2017) and Defratyka et al. (2020). They created a method to quantify C2H6 with a Picarro G2201-i. We took a similar approach to attempt for the first time to quantify H2S with a Picarro G2201-i. We added two sentences in lines 142-145 to briefly describe the method of correction and calibration. To avoid repetition, we added a sentence in lines 155-146 stating that all the details about this method can be found in sections 3.2 and 4.3.
- L143: How was the spring gas sample dried?
As mentioned in line 152, this spring gas is mostly CO2 (~80%), and it is a cold spring (~22 °C), therefore in this particular case this gas can be considered dry gas. No water traps were used at the time of sampling or at the time of analysis. It is important to emphasize that, as mentioned in line 153, an aliquot of this gas sample was mixed with zero-air. This will also produce a dry gas mixture since our zero air does not include water. The latter is confirmed by the “H2O” column in the post-data processing file generated by the Picarro instrument, where the lowest H2O values were registered when the Tedlar gas bags with our gas sample mixture were connected to the inlet.
- L260-271, Table 1: The presentation of “errors” and uncertainties is confusing here. The “errors” are not uncertainties as such, but the fractional difference between two measurements from different techniques. Moreover, while the Picarro measurements happen to agree closely with the Multigas measurements, the intrinsic measurement uncertainties of the Picarro ratio measurements are much larger than the difference from the Multigas measurement.
E.g., H2S/CH4: 87±33 (Picarro) vs 91±9 (Multigas). The difference in measurements (87 vs 91) is 4.4% but I understand the uncertainty in the Picarro ratio measurement to be 38% (87±33). The first figure is fortuitous agreement, while the second is a better statement of the overall uncertainty of the Picarro ratio measurement. These issues should be clarified in the text.
This is a very important point, and we are grateful that you brought it up. We attempted to improve our delivery of this crucial information in the text with the following changes:
We modified the results section in line 275 to amend our use of the words “error” and “uncertainty”. In lines 277- 278, we added that the difference between the uncertainties of the techniques is significant. Additionally, we added a sentence that summarizes why the Picarro uncertainties are higher than those from CH4-MultiGAS, in lines 277- 279. Finally, we added a definition of “±’ in the caption of Table 1, to improve the meaning of the uncertainty of our proposed method.
In the discussion section, in lines 367, and 369-370, we explained why the uncertainties among techniques are so different. In lines 402-403 we were more specific in our description of how to reduce the uncertainty from our proposed Picarro method.
In the following lines, we will describe our logic for using the best-fit linear regression slope for our gas ratios, despite showing a relatively higher uncertainty compared to those from the CH4-MultiGAS.
We presented the ratios as the best fit of linear regression in this study to have statistical means to define the accuracy of our proposed method. However, one main drawback was the fact that the Picarro linear regression was calculated with only four points, as shown in Figure S3. By contrast, the linear regression from the CH4-MultiGAS technique includes a minimum of 120 points. To have more points in the linear regression, a larger sample size would be recommended so that more dilution series can be performed. This difference by itself could explain the difference in the uncertainties of these methods. However, we think that the low CH4 concentrations that we were working with and the noisy “PPF_H2S” raw signal also contributed to the higher uncertainty.
As mentioned in our manuscript, the gas composition of our sampling location was challenging because we had large amounts of CO2 and low concentrations of CH4. As shown in Figure S3, we were working with very low CH4 concentrations. As a result, the standard error of our slopes was higher than those calculated from the CH4-MultiGAS technique. Here it is important to point out that the CH4-MultiGAS technique can measure much higher CO2 concentrations than can the Picarro, and therefore the methane concentrations were also higher, producing a lower uncertainty. As mentioned in line 412, higher CH4 concentrations could significantly reduce the uncertainty of the gas ratios presented here.
Additionally, as mentioned in line 397, applying moving averages to the “PPF_H2S” could also decrease this uncertainty.
Another option to calculate our gas ratios (CO2/CH4, CO2/H2S, and H2S/CH4) would be to divide the measured concentrations of each species. For example, in the following table, we present the concentrations of our highest-concentration gas mixture, as presented in Figure S2:
CH4 (ppm)
CO2 (ppm)
H2S (ppm)
3.08
26663.28
271.44
If we simply divide the above concentrations, we get the following ratios:
CO2/H2S
CO2/CH4
H2S/CH4
98.23
8655.88
88.12
These ratios show agreement with the calculated slope and with the other techniques (Table 1). This allows us to eliminate the possibility that the agreement is fortuitous. However, this method of dividing the concentrations does not allow us to define a useful uncertainty for our proposed method.
We think that the uncertainty of the Picarro ratio measurement presented in Table 1, even though it is larger, is a useful reference for our proposed method. Since the conditions of this sample are challenging, there are sparse data points in the regression line and the CH4 concentrations are very low. The uncertainties presented in Table 1 could be considered maximum values for this proposed method due to the above-mentioned difficulties.
Minor corrections:
- L78: “spectral” spelling.: The spelling of this word was corrected.
- Terminology: The term “mixing ratio” should be used in the text and figures for measurements in units of ppb and ppm, not “concentration”.
We understand that this might be a terminology issue, but we purposely decided to use the term “concentration” for the measurement obtained directed from the Picarro instrument analysis in units of “ppm” or the expected concentration of a prepared gas mixture. We think that in this way we can prevent confusion when we introduce the term "gas ratios” as CO2/CH4, CO2/H2S, and H2S/CH4.
Citation: https://doi.org/10.5194/amt-2023-265-AC1
-
RC2: 'Comment on amt-2023-265', Anonymous Referee #2, 23 Feb 2024
Review of Quantifying H2S with a Picarro CRDS G2201-i and the effect of H2S on carbon isotopes by Salas-Navarro et al.
Salas-Navarro et al. present a method for measurement of large quantities of H2S gas in volcanic plumes and gases using a commercial Picarro cavity ring down spectrometer designed to measure 12CO2, 13CO2, 12CH4, 14CH4, and δ13C-CO2 and δ13C-CH4. The technique is potentially useful for simplifying instrumental setups in harsh environments, but the work has several major issues that need to be addressed or considered in terms of publication in AMT.
Major Issues:
Novelty- The authors state that a previous Picarro instrument has been previously investigated for use in measuring H2S. They even go so far to report that the various models of the Picarro instrument utilize the same absorption lines of the gases to perform the measurements. This means that while the authors have taken a slightly different approach to producing the standards for calibration, this is essentially a repeat of the same work by previous authors just with a different variant of a commercial instrument that is intended to work the same. This does not then appear to meet the level of novelty and originality necessary for publication in AMT.
Approach to instrument outputs and results:
The authors treat the outputs of the instrument close to the outputs of a black-box instrument. What would have been more illustrative and perhaps more novel is a further discussion of the fact that a cavity ring down device, and especially one with a step scan laser source, can be used to generate spectral fits of many species that fit on top of each other. Therefore, if the instrument is outputting an H2S signal of some sort, then it is fitting that signal to some sort of absorption cross-section to get results. The authors show these cross sections in Figure 6 – and give the amount of H2S and the line shapes and overlaps, some cross-talk is likely expected. I would assume that further collaboration with the instrument company would be warranted to actually help the instrument do a better job of fitting, rather than trying to make the instrument do something that it isn’t intended to do. This would be similar to using an SO2 flash fluorescence instrument to purposely try to measure NO, since that is a possible interferant – but at least with H2S you have a chance of getting selective concentration information since there are specific absorption bands that can be fitted.
Minor comments:
In general, the paper is well written and organized.
Line 33-35: For all of these examples the authors summarize whole fields of application with a single reference for fields like atmospheric chemistry where there may be upwards of 50 or more papers using CRDS.
Line 305: The authors describe the effect on the absorption spectrum as a distortion. This is less a distortion (unless some how the presence of H2S is actually distorting the absorption band physically) and more just the additional absorption of H2S under the CH4 spectrum. This should be clarified. (more like the description on line 322).
Figure 6: The log scales make for confusing graphs here. Also, I wonder how real some of the dips and sharp valleys are in some of the absorption spectra depending on the resolution of the lines (FWHM).
Citation: https://doi.org/10.5194/amt-2023-265-RC2 -
AC2: 'Reply on RC2', Jessica Salas-Navarro, 04 Apr 2024
Response to Anonymous Referee #2
We are grateful to “Anonymous Referee #2” for the constructive comments and for pointing out key issues to be improved in the paper. We have attempted to address the issues that were raised to improve the paper as suggested and we include a detailed response to the reviewers’ questions below.
Major Issues:
- Novelty- The authors state that a previous Picarro instrument has been previously investigated for use in measuring H2S. They even go so far to report that the various models of the Picarro instrument utilize the same absorption lines of the gases to perform the measurements. This means that while the authors have taken a slightly different approach to producing the standards for calibration, this is essentially a repeat of the same work by previous authors just with a different variant of a commercial instrument that is intended to work the same. This does not then appear to meet the level of novelty and originality necessary for publication in AMT.
Thank you for raising this point. We agree it is crucial to clarify the novelty component of our contribution throughout the manuscript.
We explicitly included in the text, in lines 41-49, 138-140, and 323-325, details about other studies that were used as references for this paper. These modifications attempt to clarify that our contribution is not a repetition of the studies cited in our manuscript. Our manuscript is the first contribution that investigates in depth the interference from H2S on CO2 and CH4 and their isotopic composition using a Picarro instrument G2201-i. Additionally, it is the very first attempt to provide a method to quantify H2S with this instrument.
We modified our introduction in lines 41-49 to clarify that the work by Malowany et al., (2015) studied the effects of H2S on CO2 in volcanic environments using an older version of a Picarro instrument model G1101-i. This instrument was able to measure 12CO2, 13CO2, CH4, and H2O concentrations and isotopic compositions of δ13C-CO2. The G1101-i does not measure the carbon isotopes of methane. Malowany et al. (2015) is the only contribution in the literature that addressed H2S interferences in volcanic environments. Their contribution is now almost a decade old. Since then, very little has been published about the details of newer Picarro instruments.
In our contribution, instead of removing the H2S from the gas stream as did Malowany et al. (2015), we decided to take advantage of this linear inference to quantify H2S in the gas phase to improve laboratory routines. In lines 138-140, we stated that we were inspired by the work of Assan et al. (2017) and Defratyka et al. (2020). They created a method to quantify C2H6 with a Picarro G2201-I because C2H6 is an interference in δ13C-CH4. We took a similar approach to attempt, for the first time, to quantify H2S with a Picarro G2201-i.
Our manuscript presents the first detailed experiments to evaluate the effect of H2S on CH4 concentrations and δ13C-CH4 in volcanic environments in CRDS instruments. In lines 323-325, we described the study by Rella et al. (2015) that evaluates the effects of C2H6 on δ13C-CH4. In this contribution, the authors presented a table with an estimated effect on δ13C-CH4 caused by different gases including H2S. In Table 1 from Rella et al. (2015), the estimated effect on δ13C-CH4 caused by H2S was defined as < 0.2 ‰ ppm CH4 (ppm H2S)−1. That row in Table 1 (Rella et al., 2015) is the only information available in the literature regarding the relationship between H2S and CH4 in the CRDS technique. Rella et al. (2015) do not explain how they found this interference or in which range of H2S concentrations their experiments were run. In this case, this lack of detail was understandable because H2S was not their main interest. Due to the lack of details about the possible cross-interference between H2S and CH4, we decided to develop our study to fill that gap. Filling this gap is crucial because the CDRS technique and more specifically the G2201-i instrument have become very popular in the volcanological community. However, there also has been increasing uncertainty among researchers about the accuracy of the methane measurements and its isotopic composition.
The fact that Picarro has been utilizing the same absorption lines in the different instrument models does not mean that the interferences are the same throughout all the models. The different versions of Picarro instruments are based on the same optical lines. However, continuous operational changes have been applied to the mathematical models and software which are now embedded in the new versions of the instrument. We added lines 78-85 to give more details about the instrument and to point out that converting the absorption intensity to concentration is performed by the instrument’s software. The instrument is programmed to model the absorption peak from other surrounding molecules and subtract contributions to the absorption from these surrounding molecules.
Water vapor interference is a fitting example. Despite maintaining the same spectral lines, interferences can be mathematically corrected. More specifically, Chen et al. (2010) detected water vapor interference, and they derived a water correction function that could be applied universally to a given model of Picarro CRDS (Nara et al., 2012). According to Pang et al. (2016) in 2014 the company launched an update of the software to correct the water interference. In their contribution, Pang et al. (2016) were able to identify the correction embedded in the instrument software while the spectral lines were kept the same. Despite the constant improvement of the water vapor correction, a more recent contribution from Reum et al. (2019) has reminded us that more and novel contributions, like the one we are proposing, are needed to improve the cross-interferences in the different versions of the instrument, depending on the specific application.
Nowadays, almost a decade after Malowany et al. (2015) reported the H2S interference using a G1101-i for the first time, we, the users of the G2201-i, are not aware of the most recent updates of the instrument’s software can reduce or eliminate this linear effect on the newer commercial versions. Even if the absorption lines are the same, it became crucial to identify and quantify the H2S interference in the G2201-i for CO2 and CH4 and their carbon isotopes to ensure that our results (as well as those from other users) are accurate when using this instrument in volcanic environments. Additionally, due to the lack of recent literature on this topic, we aim to contribute an updated manuscript to better inform other users.
Approach to instrument outputs and results:
- The authors treat the outputs of the instrument close to the outputs of a black-box instrument. What would have been more illustrative and perhaps more novel is a further discussion of the fact that a cavity ring down device, and especially one with a step scan laser source, can be used to generate spectral fits of many species that fit on top of each other. Therefore, if the instrument is outputting an H2S signal of some sort, then it is fitting that signal to some sort of absorption cross-section to get results. The authors show these cross sections in Figure 6 – and give the amount of H2S and the line shapes and overlaps, some cross-talk is likely expected. I would assume that further collaboration with the instrument company would be warranted to actually help the instrument do a better job of fitting, rather than trying to make the instrument do something that it isn’t intended to do. This would be similar to using an SO2 flash fluorescence instrument to purposely try to measure NO, since that is a possible interferant – but at least with H2S you have a chance of getting selective concentration information since there are specific absorption bands that can be fitted.
We think it is important to highlight that our contribution comes from a user point of view, and it is addressed to users of the Picarro instrument G2201-i. For more specific details of the CRDS technique, the readers can consult our references section. Details about the software calculations for the spectral analysis are not available, at least to the best of our knowledge. Our manuscript includes all the details and information that are available in the literature about this specific instrument. We modified section “2.1. Laboratory conditions”. We added more specific details about the instrument and the CDRS technique in lines 78 to 85 from the user’s guide of this instrument. This section was consequently renamed to “2.1. Laboratory conditions and instrument details”.
As mentioned in our previous answer, we added lines 78-85 to provide more details about the instrument to address this comment. This text shows that the company has developed models and mathematical corrections for poorly resolved peaks. The instrument is programmed to model the absorption peak from other surrounding molecules, and it subtracts contributions to the absorption from surrounding molecules, such as water vapor. We can assume that this is how the G2201-i outputs an H2S signal. In this contribution, we offer a new method to make the most out of this signal. We can also assume that exploring the potential of this signal is not a priority for Picarro Inc., since H2S is not present in atmospheric concentrations. Therefore, we attempted to find the meaning of this signal and we propose a method to make this signal useful.
This study is useful for other scientific teams that do not have an instrument to measure H2S in the gas phase, but do have a Picarro G2201-i. In the case of volcanic gases, the traditional way to measure H2S concentration does not measure this molecule in the gas phase. Therefore, for groups that already have this instrument, the analytical routines can significantly decrease their analysis time, profiting from what is considered the biggest disadvantage of the CRDS technique in volcanic studies.
Our goals were to ensure that the G2201-i measurements were accurate when using this instrument in volcanic settings and to maximize the instrument’s output. As we mentioned before, we were inspired by other studies that have pushed the limits of this same version of the Picarro instrument and used the output signal of other gases to quantify the interference, instead of removing it from the gas stream.
Minor comments:
In general, the paper is well written and organized.
- Line 33-35: For all of these examples the authors summarize whole fields of application with a single reference for fields like atmospheric chemistry where there may be upwards of 50 or more papers using CRDS.
We agree our approach is minimalistic, but our goal was not to summarize a whole field in one example. Our goal was to emphasize the broad fields of application that have been using this technique for years, rather than focusing on the number of contributions per field.
- Line 305: The authors describe the effect on the absorption spectrum as a distortion. This is less a distortion (unless some how the presence of H2S is actually distorting the absorption band physically) and more just the additional absorption of H2S under the CH4 spectrum. This should be clarified. (more like the description on line 322).
We agree distortion is not the correct way to describe the effect on the spectral lines. We have changed the word “distortion” to “overlap” in lines 11, 39, and 414.
- Figure 6: The log scales make for confusing graphs here. Also, I wonder how real some of the dips and sharp valleys are in some of the absorption spectra depending on the resolution of the lines (FWHM).
We used log scales to make this figure comparable to the ones available in the literature (Malowany et al., 2015; Rella et al., 2015). Studies about Picarro instruments and the CRDS technique use log scales when they plot HITRAN data, so we think that we should maintain this presentation for simple comparisons among papers. We used the data directly from the HITRAN database, and the data were plotted without any filters. Our Figure 6 agrees with the HITRAN data shown by Malowany et al. (2015) and Rella et al. (2015). These studies performed experiments with different versions of the Picarro instrument, but as discussed above, the spectral lines are the same.
References
Assan, S., Baudic, A., Guemri, A., Ciais, P., Gros, V., and Vogel, F. R.: Characterization of interferences to in situ observations of δ13CH4 and C2H6 when using a cavity ring-down spectrometer at industrial sites, Atmos. Meas. Tech., 10, 2077-2091, https://doi.org/10.5194/amt-10-2077-2017, 2017.
Chen, H., Winderlich, J., Gerbig, C., Hoefer, A., Rella, C. W., Crosson, E. R., Van Pelt, A. D., Steinbach, J., Kolle, O., Beck, V., Daube, B. C., Gottlieb, E. W., Chow, V. Y., Santoni, G. W., and Wofsy, S. C.: High-accuracy continuous airborne measurements of greenhouse gases (CO2 and CH4) using the cavity ring-down spectroscopy (CRDS) technique, Atmos. Meas. Tech., 3, 375-386, https://10.5194/amt-3-375-2010, 2010.
Defratyka, S. M., Paris, J. D., Yver-Kwok, C., Loeb, D., France, J., Helmore, J., Yarrow, N., Gros, V., and Bousquet, P.: Ethane measurement by Picarro CRDS G2201-i in laboratory and field conditions: potential and limitations, Atmos. Meas. Tech. Discuss., 2020, 1-24, https://doi.org/10.5194/amt-2020-410, 2020.
Malowany, K., Stix, J., Van Pelt, A., and Lucic, G.: H2S interference on CO2 isotopic measurements using a Picarro G1101-i cavity ring-down spectrometer, Atmos. Meas. Tech., 8, 4075-4082, https://doi-org/10.5194/amt-8-4075-2015, 2015.
Nara, H., Tanimoto, H., Tohjima, Y., Mukai, H., Nojiri, Y., Katsumata, K., and Rella, C. W.: Effect of air composition (N2, O2, Ar, and H2O on CO2 and CH4 measurement by wavelength-scanned cavity ring-down spectroscopy: calibration and measurement strategy, Atmos. Meas. Tech., 5, 2689-2701, https://doi.org/10.5194/amt-5-2689-2012, 2012.
Pang, J., Wen, X., Sun, X., and Huang, K.: Intercomparison of two cavity ring-down spectroscopy analyzers for atmospheric 13CO2/ 12CO2 measurement, Atmos. Meas. Tech., 9, 3879-3891, https://doi.org/10.5194/amt-9-3879-2016, 2016.
Rella, C. W., Hoffnagle, J., He, Y., and Tajima, S.: Local- and regional-scale measurements of CH4, δ13CH4, and C2H6 in the Uintah Basin using a mobile stable isotope analyzer, Atmos. Meas. Tech., 8, 4539-4559, https://doi-org/10.5194/amt-8-4539-2015, 2015.
Reum, F., Gerbig, C., Lavric, J. V., Rella, C. W., and Göckede, M.: Correcting atmospheric CO2 and CH4 mole fractions obtained with Picarro analyzers for sensitivity of cavity pressure to water vapor, Atmos. Meas. Tech., 12, 1013-1027, https://doi-org/10.5194/amt-12-1013-2019, 2019.
Citation: https://doi.org/10.5194/amt-2023-265-AC2
-
AC2: 'Reply on RC2', Jessica Salas-Navarro, 04 Apr 2024
Status: closed
-
RC1: 'Comment on amt-2023-265', Anonymous Referee #1, 13 Feb 2024
Salas-Navarro and co-workers present an analysis of the potential interference of H2S on CH4 and CO2 carbon isotope measurements. The authors also develop an approach to quantify H2S mixing ratios. These interferences arise owing to overlap of the absorption of H2S and the 12C/13C isotopologues of CH4 and CO2. While the approach is specific to a particular commercial instrument, a Picarro G2201-I, the work will be valuable in vulcanology for quantifying H2S and improving the accuracy of CO2 and CH4 isotope measurements in environments with high H2S mixing ratios.
The authors assess the effect of H2S on CO2 and CH4 measurements across a wide range of mixing ratios of CO2, CH4, and H2S. They demonstrate an appreciable interference of H2S on the quantification of δ13C-CO2; the effect is most marked when CO2 mixing ratios are low (as would be expected). In contrast, H2S has a much smaller effect on measurement of δ 13C-CH4, and the effect is limited to conditions of low CH4 and high H2S mixing ratios. One weakness of the study, which the authors acknowledge, is that the potential interference of H2O is not considered. This work therefore applies only to dry conditions.
The paper is well written and the rationale for the work is clearly motivated. I consider that this paper is suitable for publication after addressing the following issues.
- L149: “room conditions”? Meaning the sample or physical operating environment. Presumably this means room air?
- L78-80: It is unclear how large the Picarro spectral range is and how the spectral analysis is carried out. A few details on these questions would be helpful, especially in reference to interpreting Figure 6.
- L83: Clarify whether the results of the study apply to the other instrument operating modes (CO2 only and CH4 only).
- L96: Was zero air obtained from a gas cylinder or through purification and drying external air? It seems more natural to me to provide the composition and moisture content of zero air here, and not later in L113.
- 3: The uncertainties of the dilution process cover an order of magnitude (2-20%). It is unclear how these uncertainty values are arrived at, what the uncertainty is in the aliquot and dilution volumes, and how such large uncertainties occur in the measured volumes of the components of the gas mixture. Some of the underlying details should be supplied.
- P5 l.140: Determination of H2S is central to the work reported in this paper, so a few sentences would be appropriate here to explain the approach used to correct and calibrate H2S, instead of just referencing another paper.
- L143: How was the spring gas sample dried?
- L260-271, Table 1: The presentation of “errors” and uncertainties is confusing here. The “errors” are not uncertainties as such, but the fractional difference between two measurements from different techniques. Moreover, while the Picarro measurements happen to agree closely with the Multigas measurements, the intrinsic measurement uncertainties of the Picarro ratio measurements are much larger than the difference from the Multigas measurement.
E.g., H2S/CH4: 87±33 (Picarro) vs 91±9 (Multigas). The difference in measurements (87 vs 91) is 4.4% but I understand the uncertainty in the Picarro ratio measurement to be 38% (87±33). The first figure is fortuitous agreement, while the second is a better statement of the overall uncertainty of the Picarro ratio measurement. These issues should be clarified in the text.
Minor corrections:
- L78: “spectral” spelling.
- Terminology: The term “mixing ratio” should be used in the text and figures for measurements in units of ppb and ppm, not “concentration”.
Citation: https://doi.org/10.5194/amt-2023-265-RC1 -
AC1: 'Reply on RC1', Jessica Salas-Navarro, 04 Apr 2024
Response to Anonymous Referee #1
We would like to thank “Anonymous Referee #1” for their valuable and constructive comments that allowed us to significantly improve our manuscript. Below we address all the points raised by the reviewer and we attempt to provide adequate solutions and responses to their concerns in the manuscript.
- L149: “room conditions”? Meaning the sample or physical operating environment. Presumably this means room air?
“Room conditions” does mean room air. We recognize that it might not be clear, therefore we have changed the text from “room conditions” to “room air” as suggested. We made this change in lines 120 and 134.
- L78-80: It is unclear how large the Picarro spectral range is and how the spectral analysis is carried out. A few details on these questions would be helpful, especially in reference to interpreting Figure 6.
We think it is important to highlight that our contribution comes from a user point of view, and it is addressed to users of the Picarro instrument G2201-i. For more specific details of the CRDS technique, the readers can consult our references section. Details about the software calculations for the spectral analysis are not available, at least to the best of our knowledge. Our manuscript includes all the details and information that are available in the literature about this specific instrument. Nevertheless, we did address this issue by modifying section “2.1. Laboratory conditions”. Specifically, we added more details about the instrument and the technique from the user’s guide of this instrument in lines 78 to 85. This section was renamed to “2.1. Laboratory conditions and instrument details”.
- L83: Clarify whether the results of the study apply to the other instrument operating modes (CO2 only and CH4 only).
As mentioned in line 87, all our analyses were conducted using the CO2 and CH4 simultaneous mode. We did not explore the other operating modes. Therefore, we cannot be certain of the effects of H2S on the other operating modes. However, based on the literature and our basic understanding of the operation of these instruments, we feel confident that the H2S effect is the same in the other operating modes.
We did not explore the other operating modes because as mentioned in our motivation, measuring CO2 and CH4 simultaneously is one of the most appealing and useful characteristics of the Picarro G2201-i instrument in the field of volcanology.
- L96: Was zero air obtained from a gas cylinder or through purification and drying external air? It seems more natural to me to provide the composition and moisture content of zero air here, and not later in L113.
Zero air was obtained from a gas cylinder. We provided the zero air tank composition in line 101 and not later in line 117, as you suggested.
- 3: The uncertainties of the dilution process cover an order of magnitude (2-20%). It is unclear how these uncertainty values are arrived at, what the uncertainty is in the aliquot and dilution volumes, and how such large uncertainties occur in the measured volumes of the components of the gas mixture. Some of the underlying details should be supplied.
The uncertainty at the gas mixture preparation was estimated by considering the error introduced by the different syringes used to make the mixtures, and the error from the gas standard, multiplied by 2. As shown in the following equation:
The error of each syringe was calculated with the following equation:
The order of magnitude difference in these uncertainties is attributed to the very small aliquots of our 100% gas standards (CO2 and CH4) required to create low concentration mixtures, as well as large aliquots of zero air.
- P5 l.140: Determination of H2S is central to the work reported in this paper, so a few sentences would be appropriate here to explain the approach used to correct and calibrate H2S, instead of just referencing another paper.
We modified the section “2.6 Quantifying H2S concentrations” to clarify that we were inspired by the work done by Assan et al. (2017) and Defratyka et al. (2020). They created a method to quantify C2H6 with a Picarro G2201-i. We took a similar approach to attempt for the first time to quantify H2S with a Picarro G2201-i. We added two sentences in lines 142-145 to briefly describe the method of correction and calibration. To avoid repetition, we added a sentence in lines 155-146 stating that all the details about this method can be found in sections 3.2 and 4.3.
- L143: How was the spring gas sample dried?
As mentioned in line 152, this spring gas is mostly CO2 (~80%), and it is a cold spring (~22 °C), therefore in this particular case this gas can be considered dry gas. No water traps were used at the time of sampling or at the time of analysis. It is important to emphasize that, as mentioned in line 153, an aliquot of this gas sample was mixed with zero-air. This will also produce a dry gas mixture since our zero air does not include water. The latter is confirmed by the “H2O” column in the post-data processing file generated by the Picarro instrument, where the lowest H2O values were registered when the Tedlar gas bags with our gas sample mixture were connected to the inlet.
- L260-271, Table 1: The presentation of “errors” and uncertainties is confusing here. The “errors” are not uncertainties as such, but the fractional difference between two measurements from different techniques. Moreover, while the Picarro measurements happen to agree closely with the Multigas measurements, the intrinsic measurement uncertainties of the Picarro ratio measurements are much larger than the difference from the Multigas measurement.
E.g., H2S/CH4: 87±33 (Picarro) vs 91±9 (Multigas). The difference in measurements (87 vs 91) is 4.4% but I understand the uncertainty in the Picarro ratio measurement to be 38% (87±33). The first figure is fortuitous agreement, while the second is a better statement of the overall uncertainty of the Picarro ratio measurement. These issues should be clarified in the text.
This is a very important point, and we are grateful that you brought it up. We attempted to improve our delivery of this crucial information in the text with the following changes:
We modified the results section in line 275 to amend our use of the words “error” and “uncertainty”. In lines 277- 278, we added that the difference between the uncertainties of the techniques is significant. Additionally, we added a sentence that summarizes why the Picarro uncertainties are higher than those from CH4-MultiGAS, in lines 277- 279. Finally, we added a definition of “±’ in the caption of Table 1, to improve the meaning of the uncertainty of our proposed method.
In the discussion section, in lines 367, and 369-370, we explained why the uncertainties among techniques are so different. In lines 402-403 we were more specific in our description of how to reduce the uncertainty from our proposed Picarro method.
In the following lines, we will describe our logic for using the best-fit linear regression slope for our gas ratios, despite showing a relatively higher uncertainty compared to those from the CH4-MultiGAS.
We presented the ratios as the best fit of linear regression in this study to have statistical means to define the accuracy of our proposed method. However, one main drawback was the fact that the Picarro linear regression was calculated with only four points, as shown in Figure S3. By contrast, the linear regression from the CH4-MultiGAS technique includes a minimum of 120 points. To have more points in the linear regression, a larger sample size would be recommended so that more dilution series can be performed. This difference by itself could explain the difference in the uncertainties of these methods. However, we think that the low CH4 concentrations that we were working with and the noisy “PPF_H2S” raw signal also contributed to the higher uncertainty.
As mentioned in our manuscript, the gas composition of our sampling location was challenging because we had large amounts of CO2 and low concentrations of CH4. As shown in Figure S3, we were working with very low CH4 concentrations. As a result, the standard error of our slopes was higher than those calculated from the CH4-MultiGAS technique. Here it is important to point out that the CH4-MultiGAS technique can measure much higher CO2 concentrations than can the Picarro, and therefore the methane concentrations were also higher, producing a lower uncertainty. As mentioned in line 412, higher CH4 concentrations could significantly reduce the uncertainty of the gas ratios presented here.
Additionally, as mentioned in line 397, applying moving averages to the “PPF_H2S” could also decrease this uncertainty.
Another option to calculate our gas ratios (CO2/CH4, CO2/H2S, and H2S/CH4) would be to divide the measured concentrations of each species. For example, in the following table, we present the concentrations of our highest-concentration gas mixture, as presented in Figure S2:
CH4 (ppm)
CO2 (ppm)
H2S (ppm)
3.08
26663.28
271.44
If we simply divide the above concentrations, we get the following ratios:
CO2/H2S
CO2/CH4
H2S/CH4
98.23
8655.88
88.12
These ratios show agreement with the calculated slope and with the other techniques (Table 1). This allows us to eliminate the possibility that the agreement is fortuitous. However, this method of dividing the concentrations does not allow us to define a useful uncertainty for our proposed method.
We think that the uncertainty of the Picarro ratio measurement presented in Table 1, even though it is larger, is a useful reference for our proposed method. Since the conditions of this sample are challenging, there are sparse data points in the regression line and the CH4 concentrations are very low. The uncertainties presented in Table 1 could be considered maximum values for this proposed method due to the above-mentioned difficulties.
Minor corrections:
- L78: “spectral” spelling.: The spelling of this word was corrected.
- Terminology: The term “mixing ratio” should be used in the text and figures for measurements in units of ppb and ppm, not “concentration”.
We understand that this might be a terminology issue, but we purposely decided to use the term “concentration” for the measurement obtained directed from the Picarro instrument analysis in units of “ppm” or the expected concentration of a prepared gas mixture. We think that in this way we can prevent confusion when we introduce the term "gas ratios” as CO2/CH4, CO2/H2S, and H2S/CH4.
Citation: https://doi.org/10.5194/amt-2023-265-AC1
-
RC2: 'Comment on amt-2023-265', Anonymous Referee #2, 23 Feb 2024
Review of Quantifying H2S with a Picarro CRDS G2201-i and the effect of H2S on carbon isotopes by Salas-Navarro et al.
Salas-Navarro et al. present a method for measurement of large quantities of H2S gas in volcanic plumes and gases using a commercial Picarro cavity ring down spectrometer designed to measure 12CO2, 13CO2, 12CH4, 14CH4, and δ13C-CO2 and δ13C-CH4. The technique is potentially useful for simplifying instrumental setups in harsh environments, but the work has several major issues that need to be addressed or considered in terms of publication in AMT.
Major Issues:
Novelty- The authors state that a previous Picarro instrument has been previously investigated for use in measuring H2S. They even go so far to report that the various models of the Picarro instrument utilize the same absorption lines of the gases to perform the measurements. This means that while the authors have taken a slightly different approach to producing the standards for calibration, this is essentially a repeat of the same work by previous authors just with a different variant of a commercial instrument that is intended to work the same. This does not then appear to meet the level of novelty and originality necessary for publication in AMT.
Approach to instrument outputs and results:
The authors treat the outputs of the instrument close to the outputs of a black-box instrument. What would have been more illustrative and perhaps more novel is a further discussion of the fact that a cavity ring down device, and especially one with a step scan laser source, can be used to generate spectral fits of many species that fit on top of each other. Therefore, if the instrument is outputting an H2S signal of some sort, then it is fitting that signal to some sort of absorption cross-section to get results. The authors show these cross sections in Figure 6 – and give the amount of H2S and the line shapes and overlaps, some cross-talk is likely expected. I would assume that further collaboration with the instrument company would be warranted to actually help the instrument do a better job of fitting, rather than trying to make the instrument do something that it isn’t intended to do. This would be similar to using an SO2 flash fluorescence instrument to purposely try to measure NO, since that is a possible interferant – but at least with H2S you have a chance of getting selective concentration information since there are specific absorption bands that can be fitted.
Minor comments:
In general, the paper is well written and organized.
Line 33-35: For all of these examples the authors summarize whole fields of application with a single reference for fields like atmospheric chemistry where there may be upwards of 50 or more papers using CRDS.
Line 305: The authors describe the effect on the absorption spectrum as a distortion. This is less a distortion (unless some how the presence of H2S is actually distorting the absorption band physically) and more just the additional absorption of H2S under the CH4 spectrum. This should be clarified. (more like the description on line 322).
Figure 6: The log scales make for confusing graphs here. Also, I wonder how real some of the dips and sharp valleys are in some of the absorption spectra depending on the resolution of the lines (FWHM).
Citation: https://doi.org/10.5194/amt-2023-265-RC2 -
AC2: 'Reply on RC2', Jessica Salas-Navarro, 04 Apr 2024
Response to Anonymous Referee #2
We are grateful to “Anonymous Referee #2” for the constructive comments and for pointing out key issues to be improved in the paper. We have attempted to address the issues that were raised to improve the paper as suggested and we include a detailed response to the reviewers’ questions below.
Major Issues:
- Novelty- The authors state that a previous Picarro instrument has been previously investigated for use in measuring H2S. They even go so far to report that the various models of the Picarro instrument utilize the same absorption lines of the gases to perform the measurements. This means that while the authors have taken a slightly different approach to producing the standards for calibration, this is essentially a repeat of the same work by previous authors just with a different variant of a commercial instrument that is intended to work the same. This does not then appear to meet the level of novelty and originality necessary for publication in AMT.
Thank you for raising this point. We agree it is crucial to clarify the novelty component of our contribution throughout the manuscript.
We explicitly included in the text, in lines 41-49, 138-140, and 323-325, details about other studies that were used as references for this paper. These modifications attempt to clarify that our contribution is not a repetition of the studies cited in our manuscript. Our manuscript is the first contribution that investigates in depth the interference from H2S on CO2 and CH4 and their isotopic composition using a Picarro instrument G2201-i. Additionally, it is the very first attempt to provide a method to quantify H2S with this instrument.
We modified our introduction in lines 41-49 to clarify that the work by Malowany et al., (2015) studied the effects of H2S on CO2 in volcanic environments using an older version of a Picarro instrument model G1101-i. This instrument was able to measure 12CO2, 13CO2, CH4, and H2O concentrations and isotopic compositions of δ13C-CO2. The G1101-i does not measure the carbon isotopes of methane. Malowany et al. (2015) is the only contribution in the literature that addressed H2S interferences in volcanic environments. Their contribution is now almost a decade old. Since then, very little has been published about the details of newer Picarro instruments.
In our contribution, instead of removing the H2S from the gas stream as did Malowany et al. (2015), we decided to take advantage of this linear inference to quantify H2S in the gas phase to improve laboratory routines. In lines 138-140, we stated that we were inspired by the work of Assan et al. (2017) and Defratyka et al. (2020). They created a method to quantify C2H6 with a Picarro G2201-I because C2H6 is an interference in δ13C-CH4. We took a similar approach to attempt, for the first time, to quantify H2S with a Picarro G2201-i.
Our manuscript presents the first detailed experiments to evaluate the effect of H2S on CH4 concentrations and δ13C-CH4 in volcanic environments in CRDS instruments. In lines 323-325, we described the study by Rella et al. (2015) that evaluates the effects of C2H6 on δ13C-CH4. In this contribution, the authors presented a table with an estimated effect on δ13C-CH4 caused by different gases including H2S. In Table 1 from Rella et al. (2015), the estimated effect on δ13C-CH4 caused by H2S was defined as < 0.2 ‰ ppm CH4 (ppm H2S)−1. That row in Table 1 (Rella et al., 2015) is the only information available in the literature regarding the relationship between H2S and CH4 in the CRDS technique. Rella et al. (2015) do not explain how they found this interference or in which range of H2S concentrations their experiments were run. In this case, this lack of detail was understandable because H2S was not their main interest. Due to the lack of details about the possible cross-interference between H2S and CH4, we decided to develop our study to fill that gap. Filling this gap is crucial because the CDRS technique and more specifically the G2201-i instrument have become very popular in the volcanological community. However, there also has been increasing uncertainty among researchers about the accuracy of the methane measurements and its isotopic composition.
The fact that Picarro has been utilizing the same absorption lines in the different instrument models does not mean that the interferences are the same throughout all the models. The different versions of Picarro instruments are based on the same optical lines. However, continuous operational changes have been applied to the mathematical models and software which are now embedded in the new versions of the instrument. We added lines 78-85 to give more details about the instrument and to point out that converting the absorption intensity to concentration is performed by the instrument’s software. The instrument is programmed to model the absorption peak from other surrounding molecules and subtract contributions to the absorption from these surrounding molecules.
Water vapor interference is a fitting example. Despite maintaining the same spectral lines, interferences can be mathematically corrected. More specifically, Chen et al. (2010) detected water vapor interference, and they derived a water correction function that could be applied universally to a given model of Picarro CRDS (Nara et al., 2012). According to Pang et al. (2016) in 2014 the company launched an update of the software to correct the water interference. In their contribution, Pang et al. (2016) were able to identify the correction embedded in the instrument software while the spectral lines were kept the same. Despite the constant improvement of the water vapor correction, a more recent contribution from Reum et al. (2019) has reminded us that more and novel contributions, like the one we are proposing, are needed to improve the cross-interferences in the different versions of the instrument, depending on the specific application.
Nowadays, almost a decade after Malowany et al. (2015) reported the H2S interference using a G1101-i for the first time, we, the users of the G2201-i, are not aware of the most recent updates of the instrument’s software can reduce or eliminate this linear effect on the newer commercial versions. Even if the absorption lines are the same, it became crucial to identify and quantify the H2S interference in the G2201-i for CO2 and CH4 and their carbon isotopes to ensure that our results (as well as those from other users) are accurate when using this instrument in volcanic environments. Additionally, due to the lack of recent literature on this topic, we aim to contribute an updated manuscript to better inform other users.
Approach to instrument outputs and results:
- The authors treat the outputs of the instrument close to the outputs of a black-box instrument. What would have been more illustrative and perhaps more novel is a further discussion of the fact that a cavity ring down device, and especially one with a step scan laser source, can be used to generate spectral fits of many species that fit on top of each other. Therefore, if the instrument is outputting an H2S signal of some sort, then it is fitting that signal to some sort of absorption cross-section to get results. The authors show these cross sections in Figure 6 – and give the amount of H2S and the line shapes and overlaps, some cross-talk is likely expected. I would assume that further collaboration with the instrument company would be warranted to actually help the instrument do a better job of fitting, rather than trying to make the instrument do something that it isn’t intended to do. This would be similar to using an SO2 flash fluorescence instrument to purposely try to measure NO, since that is a possible interferant – but at least with H2S you have a chance of getting selective concentration information since there are specific absorption bands that can be fitted.
We think it is important to highlight that our contribution comes from a user point of view, and it is addressed to users of the Picarro instrument G2201-i. For more specific details of the CRDS technique, the readers can consult our references section. Details about the software calculations for the spectral analysis are not available, at least to the best of our knowledge. Our manuscript includes all the details and information that are available in the literature about this specific instrument. We modified section “2.1. Laboratory conditions”. We added more specific details about the instrument and the CDRS technique in lines 78 to 85 from the user’s guide of this instrument. This section was consequently renamed to “2.1. Laboratory conditions and instrument details”.
As mentioned in our previous answer, we added lines 78-85 to provide more details about the instrument to address this comment. This text shows that the company has developed models and mathematical corrections for poorly resolved peaks. The instrument is programmed to model the absorption peak from other surrounding molecules, and it subtracts contributions to the absorption from surrounding molecules, such as water vapor. We can assume that this is how the G2201-i outputs an H2S signal. In this contribution, we offer a new method to make the most out of this signal. We can also assume that exploring the potential of this signal is not a priority for Picarro Inc., since H2S is not present in atmospheric concentrations. Therefore, we attempted to find the meaning of this signal and we propose a method to make this signal useful.
This study is useful for other scientific teams that do not have an instrument to measure H2S in the gas phase, but do have a Picarro G2201-i. In the case of volcanic gases, the traditional way to measure H2S concentration does not measure this molecule in the gas phase. Therefore, for groups that already have this instrument, the analytical routines can significantly decrease their analysis time, profiting from what is considered the biggest disadvantage of the CRDS technique in volcanic studies.
Our goals were to ensure that the G2201-i measurements were accurate when using this instrument in volcanic settings and to maximize the instrument’s output. As we mentioned before, we were inspired by other studies that have pushed the limits of this same version of the Picarro instrument and used the output signal of other gases to quantify the interference, instead of removing it from the gas stream.
Minor comments:
In general, the paper is well written and organized.
- Line 33-35: For all of these examples the authors summarize whole fields of application with a single reference for fields like atmospheric chemistry where there may be upwards of 50 or more papers using CRDS.
We agree our approach is minimalistic, but our goal was not to summarize a whole field in one example. Our goal was to emphasize the broad fields of application that have been using this technique for years, rather than focusing on the number of contributions per field.
- Line 305: The authors describe the effect on the absorption spectrum as a distortion. This is less a distortion (unless some how the presence of H2S is actually distorting the absorption band physically) and more just the additional absorption of H2S under the CH4 spectrum. This should be clarified. (more like the description on line 322).
We agree distortion is not the correct way to describe the effect on the spectral lines. We have changed the word “distortion” to “overlap” in lines 11, 39, and 414.
- Figure 6: The log scales make for confusing graphs here. Also, I wonder how real some of the dips and sharp valleys are in some of the absorption spectra depending on the resolution of the lines (FWHM).
We used log scales to make this figure comparable to the ones available in the literature (Malowany et al., 2015; Rella et al., 2015). Studies about Picarro instruments and the CRDS technique use log scales when they plot HITRAN data, so we think that we should maintain this presentation for simple comparisons among papers. We used the data directly from the HITRAN database, and the data were plotted without any filters. Our Figure 6 agrees with the HITRAN data shown by Malowany et al. (2015) and Rella et al. (2015). These studies performed experiments with different versions of the Picarro instrument, but as discussed above, the spectral lines are the same.
References
Assan, S., Baudic, A., Guemri, A., Ciais, P., Gros, V., and Vogel, F. R.: Characterization of interferences to in situ observations of δ13CH4 and C2H6 when using a cavity ring-down spectrometer at industrial sites, Atmos. Meas. Tech., 10, 2077-2091, https://doi.org/10.5194/amt-10-2077-2017, 2017.
Chen, H., Winderlich, J., Gerbig, C., Hoefer, A., Rella, C. W., Crosson, E. R., Van Pelt, A. D., Steinbach, J., Kolle, O., Beck, V., Daube, B. C., Gottlieb, E. W., Chow, V. Y., Santoni, G. W., and Wofsy, S. C.: High-accuracy continuous airborne measurements of greenhouse gases (CO2 and CH4) using the cavity ring-down spectroscopy (CRDS) technique, Atmos. Meas. Tech., 3, 375-386, https://10.5194/amt-3-375-2010, 2010.
Defratyka, S. M., Paris, J. D., Yver-Kwok, C., Loeb, D., France, J., Helmore, J., Yarrow, N., Gros, V., and Bousquet, P.: Ethane measurement by Picarro CRDS G2201-i in laboratory and field conditions: potential and limitations, Atmos. Meas. Tech. Discuss., 2020, 1-24, https://doi.org/10.5194/amt-2020-410, 2020.
Malowany, K., Stix, J., Van Pelt, A., and Lucic, G.: H2S interference on CO2 isotopic measurements using a Picarro G1101-i cavity ring-down spectrometer, Atmos. Meas. Tech., 8, 4075-4082, https://doi-org/10.5194/amt-8-4075-2015, 2015.
Nara, H., Tanimoto, H., Tohjima, Y., Mukai, H., Nojiri, Y., Katsumata, K., and Rella, C. W.: Effect of air composition (N2, O2, Ar, and H2O on CO2 and CH4 measurement by wavelength-scanned cavity ring-down spectroscopy: calibration and measurement strategy, Atmos. Meas. Tech., 5, 2689-2701, https://doi.org/10.5194/amt-5-2689-2012, 2012.
Pang, J., Wen, X., Sun, X., and Huang, K.: Intercomparison of two cavity ring-down spectroscopy analyzers for atmospheric 13CO2/ 12CO2 measurement, Atmos. Meas. Tech., 9, 3879-3891, https://doi.org/10.5194/amt-9-3879-2016, 2016.
Rella, C. W., Hoffnagle, J., He, Y., and Tajima, S.: Local- and regional-scale measurements of CH4, δ13CH4, and C2H6 in the Uintah Basin using a mobile stable isotope analyzer, Atmos. Meas. Tech., 8, 4539-4559, https://doi-org/10.5194/amt-8-4539-2015, 2015.
Reum, F., Gerbig, C., Lavric, J. V., Rella, C. W., and Göckede, M.: Correcting atmospheric CO2 and CH4 mole fractions obtained with Picarro analyzers for sensitivity of cavity pressure to water vapor, Atmos. Meas. Tech., 12, 1013-1027, https://doi-org/10.5194/amt-12-1013-2019, 2019.
Citation: https://doi.org/10.5194/amt-2023-265-AC2
-
AC2: 'Reply on RC2', Jessica Salas-Navarro, 04 Apr 2024
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 359 | 78 | 36 | 473 | 52 | 27 | 26 |
- HTML: 359
- PDF: 78
- XML: 36
- Total: 473
- Supplement: 52
- BibTeX: 27
- EndNote: 26
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1