the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 13 Sep 2022
Characterization of atmospheric methane release in the outer Mackenzie River Delta from biogenic and thermogenic sources
Abstract. The Mackenzie River Delta is the second largest Arctic river delta in the world. Thin and destabilizing permafrost coupled with vast natural gas reserves at depth, high organic content soils, and a high proportion of wetlands create a unique ecosystem conducive to high rates of methane production from biogenic and thermogenic sources. Hotspots are known to have a significant contribution to summertime CH4 emissions in the region, but little research has been done to determine how often geologic or biogenic methane contributes to CH4 hotspots in the Mackenzie River Delta. In the present study, stable carbon isotope analysis was used to identify the source of CH4 at several aquatic and terrestrial sites thought to be hotspots of CH4 flux to the atmosphere. Source stable carbon isotope (δ13C-CH4) signatures were derived from keeling plots of point samples and ranged from -42 to -88 ‰ δ13C-CH4, identifying both biogenic and thermogenic and mixed biogenic/thermogenic sources. A CH4 source was determined for eight hotspots, two were thermogenic in origin, four were biogenic in origin, and two were from mixed biogenic/thermogenic sources, as evidenced by δ13C-CH4 signatures. This indicates that the largest hotspots of CH4 production in the Mackenzie River Delta are caused by a variety of sources. In addition to biogenic production at the surface we have identified CH4 migration to the surface from the Taglu gas field over an area of approximately 20 km from north to south and two different sites of mixed biogenic/thermogenic CH4 that were approximately 30 km apart.
-
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
(1027 KB)
-
Supplement
(159 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1027 KB) - Metadata XML
-
Supplement
(159 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
CC1: 'Comment on egusphere-2022-549', Laura Lapham, 06 Dec 2022
Summary: The goal of the paper was to take a first look at the source of the hot spots previously found by aerial surveys using stable carbon isotopes within the Mackenzie River Delta (MRD). The authors state that the MRD has thermogenic gas seepage and biogenic methane atmospheric sources. Given there are so many lakes, this is not surprising, but there should be references for that statement. The authors hypothesize (line 84) that the hot spots come from “thermogenic, biogenic, and mixed sources….”. This hypothesis could be strengthened by picking one source based on previous literature. For example, the Kohnert paper clearly suggests that the hot spots are thermogenic because they are so high, but didn’t have any detailed source information to support this. This paper does. Or maybe the fact that there are so many lakes emitting biogenic methane means the atmospheric methane would make one hypothesize these hot spots would be biogenic? The strength of this paper is in using stable carbon isotope ratios of methane in air from ground surveys. The paper could benefit from some more details on the sampling sites, the sampling protocols, the assumptions going into the keeling plots, and overall conclusions drawn from the study. Overall, I think the data is interesting, and definitely novel and worthy of publishing. Hopefully the comments below could help streamline the message:
- The introduction could be streamlined and strengthened. Currently, the text suggests that the MRD has biogenic methane sources, but I couldn’t see any references to support that. Maybe work from Lance Lesack’s group would be helpful here for in and around Inuvik. On line 60, there is also mention of production of methane in the organic rich active layer but has no reference. There was a recent paper to conduct permafrost incubation studies (Lapham et al., 2021) in Tuk, but that didn’t take place in the active layer. In terms of streamlining, the sentence from line 67-68 could be cut; while it is important to measure emissions of methane, since the current study didn’t do this, it’s not necessary.
- The study location and methods sections could be reorganized to be sure the proper information is conveyed in each section. For example, the setting was described in section2, and then there was a “study location” in section 3. The methods section should only give the methods used, and results should be reported in the results section.
- Overall figures are sufficient with some revision. For example, figure 1 could be more informative if the walking transects were shown on the pictures, and the direction of wind and location of seeps were also shown. As it is now, without any labels, it’s difficult to see where the seeps are (unless you know), and at Lake 1, unless you know the picture shows the lake is ice covered, it’s hard to know what is happening. The caption says “prominent ebullition” but unless you know what to look for, it’s not clear with the captions. Furthermore, supplemental figure S2 is not mentioned in the text, yet addresses some concerns about wind direction and the location of the transects. If the quality could be improved, or information combined between figure 1 and figure S2, this could help.
- The chamber flux data seems to be an add-on. Is it necessary for the message of the paper? If so, more details will be needed (like how the samples were analyzed) and some context of what the fluxes mean in terms of other environments (for the discussion). If not needed, please consider taking out of the paper.
Detailed comments:
Line 34: add in “oxidation” after production and before transport. I think it’s important with some of the conclusions drawn to get the idea of oxidation into the text earlier.
Line 51: As numbers, delta values can be high or low, positive or negative, but not heavy or light. Please use “high” or “low” d13C values. Be sure to check this throughout and change accordingly.
Line 75: “geologic origin”: maybe define what you mean and that it could be made up of thermogenic methane produced deep and migrates, or biogenic or rather, microbial methane produced in the permafrost? See comment from line 224.
Line 78-80: The sentence starting “interestingly, ….” Is a confusing sentence. How can the sources behave differently than the current understanding? Please reword to make more clear.
Line 117: it would be helpful to have a mark on the map for Tuk.
Lines around 119: describe your sites a little more than just a name. What is “site 9”? There isn’t a description in figure 2, and on the figure 1 map, it looks like it’s upwind of the hot spot from Kohnert. What were the winds like when you sampled it? Also, the sentence here of “Of the five airborne eddy covariance hotspots…..” is a result. It should be moved to that section. In thinking more about the study design, if the idea was to ground truth the Kohnert hot spots, there is a missed opportunity to bring out some novelty of this study. For example, if we look at the “pingo 1” site, the fact that there is a pingo there is important, right? Does that already add information not gained from the Kohnert study? That you observed a pingo there? And what would that mean, what would the observation of a pingo mean for methane emissions? Aren’t they by definition conduits of some sort? Or what about pingos being surrounded by wetlands? Additionally, at wetland 3, the fact that there are wetlands characterizing that hot spot is interesting information. I’d almost envision numbering the hotspots sequentially and then giving them their ground feature names as done on figure 1 (this is just a suggestion and maybe not helpful, it just seems interesting what ground features underlay the aerial hot spots). Yet, such an approach really send home the message that is directly inline with your goal, to groundtruth the aerial survey hot spots.
Table S1: The raw data should be available somewhere for review. Will it be available in a database somewhere, or as a supplemental table here? Also, in this table, for the “source d13C values, R2, and max CH4” in the fourth column, those are the same values as in figure 1. I’m not sure what this table is adding except to give exact locations. Can you replace this with the raw data, from which you derive the y-intercept from keeling plots? Also, Table S1 gives a “site type” as Polar V. How is the Polar V a site type? Also, you give into in this table for the low R2 values, what made you pick the R2 value cut offs you did? For example, you kept an R2 of 0.48 but didn’t talk about the site with R2 of 0.434. This sort of thing should be mentioned in the methods.
Line 123: how did you determine where discrete point samples were collected? What was the strategy? Upwind, downwind, etc? This is a study location section, is this the best place for that information? I think you should be explicit that the strategy was to target the hot spots, and if you adopt a sample numbering scheme like numbering the hot spots, this strategy will be clear.
Line 125: Is “Lake 1” known as another lake by the community? Is that “shot hole” lake or Swiss cheese? And Channel Seep, is that channel seep 1? There could be reports that have some isotope data reported to help aid you in interpretations.
Line 126: the “observations of ebullitions seen in open water in summer”, were those your observations made in this sampling trip or previous knowledge? Cite previous knowledge.
Line 126: how close is “as close as possible” to observed ebullitions? Were you on land, upwind, downwind? How long were walking transects, 10m? This discussion in these 3 lines around 130 are more about sample collection. As such, they should be moved to that section of the paper.
Lines 131-142: This paragraph seems more high level than where it is situated in the paper. I suggest reorganizing this study location section.
Line 149: Are you missing the word “under”? Permafrost “under” Tuk…..
Line 154: what is precision of handheld GPS and what model, make? What is the precision of the licor and the picarro CRDS? It’s important to mention the cavity ring down spectrometer (CRDS) part of the instrument since that is how the measurement is made. And what standards were used to calibrate the concentration and isotope measurements? The placement of the analysis instruments seems out of place here since this is the sample collection section.
Line 156: Please cite the airborne work paper, since it was not done in this study.
Line 158: “photographs of each site….” I think this should be moved to study location, and not sample collection. You are using the photographs to describe the sites, correct? If so, they really set the stage for the setting, which belongs elsewhere.
Line 159: “walking transections…” what was pumping rate? What are dimensions of tubing? Is 6mm OD or ID?
Line 163: “Mixing between sample collection and analysis is limited due to small diameter of tubing”. Have you proved that? Or is there a paper you can cite for this? The reason I ask is that the pumping rate is pretty fast, so I would imagine your sample will smear alone the edges of the tubing and mix along the way it’s filling. Please give more details as to the accuracy of this approach.
Line 166: Why did you pick 1 meter above ground level? Did you ever try to go down to ground level? Did you see any change in the concentration? Or is 1m desirable because things are more mixed and you are trying not to see a ground signal? What is the thought behind this?
Line 171: Where are the flux chamber measurement data? Also, this section is confusing as written. It looks like 2 chambers were used, the automated one from licor, but I can’t tell what was used for Lake 1, using manual extraction of samples. And the dimensions are for a flux chamber, but there is only one, yet it seems two chambers were used? Also, can you discuss how only allowing 1 hour between collars installed and measurement might change your results?
Line 178: “Keeling plot analysis”. I am not sure this is the right terminology. You analyzed keeling plots to determine the stable carbon isotope signature of the methane source. That phrase, keeling plot analysis, is used several times, so maybe it is the right term but it could be better to say “We constructed keeling plots with the discrete transect data to determine ….” And also cite the “common approach” of using keeling plots. I agree that it is common, but I am not sure it is common to take discrete measurements in the horizontal direction versus the vertical direction. Meaning, I thought keeling plots were always done in the vertical collection of air in a forest canopy, for example. If that is not true, it’s probably still making a note of this since it does seem a bit novel to use this technique for the walking transects. Or maybe there are papers to show this approach used in this way.
Line 193: Do you have the bubble isotope values from historical data by the GSC to compare to your point measurements?
Line 197: Why is walking transect data shown as average data? The maximum value of 12ppm is very interesting and seems important to know where you were in comparison to the “wetland”. Is it possible to do this from the GPS location data you obtained?
Line 203: The sentence of “estimates of source….” Seems a bit premature. Since this is the methods section, could you first say that the keeling plots are shown in figure 3, and show intercepts of X, Y, Z, which indicate the source of the methane? And also give the R2 values? As it is now, you don’t mention figure 3, so it’s unclear where these numbers come from. The figure 3 caption mentions a “grey region” but there are no grey regions on the figures. And finally, there is a formatting issue with the equation written on the figures. It is also important to mention the cut off you used for the R2 values.
Line 208: “Keeling plot values” is not quite accurate. Maybe say “keeling plot y-intercepts”. It’s interesting the seasonal component of these values. Is seems reasonable to think that during the winter, there isn’t as much methane oxidation, which leads to the lower values. I think you mention that in discussion.
Line 211: “Flux rates” isn’t accurate. A flux is a concentration per area per time, a rate is distance per time. I think you just mean fluxes here. It also seems like the chamber fluxes are an afterthought since the fluxes aren’t given here. And they are put in a supplemental table. Are they needed?
Lines 215-218: change “values” to “concentrations”.
Line 217: take out “were”. What is an “observation” in this context? There wasn’t mention of observations before now. Where does the 2013 number come from? Do you mean the discrete measurements? Same question for the 1850 observations at pingo 1. This section is a bit confusing since it’s also written like results, but yet in the discussion section.
Line 224: For the general reader, it might help to define geologic source in the introduction to put this conclusion in more context.
Line 240: Can you give values for the pingo 1 and 2 site in the text? It would help the reader not have to flip back to the figures.
Line 241: Is this new data? I didn’t see the reporting of the methane concentrations over the pingo features themselves. And this sentence is also a bit confusing as to what you mean. Did you do 2 transects for Pingo 1? Seems like supplemental figure S2 (top right) could be helpful to show what you mean here. You could refer to that figure here, but that figure quality needs to be improved.
Line 243: Only wind speed is reported in table 1. Wind direction is also key that would be important to show in that table. Or did you always collect samples downwind? Do you see a correlation with wind speed? I would think that the higher the wind speed, the further away the source of that gas could be.
Paragraph starting at line 254: This is a great reason you chose to sample pingos. Could you move it to the introduction to help set up the “why” for your study? After reading the paper again, it seems you didn’t set out to study pingos, per say, but you found those pingos at the ground features under the aerial hot spots. If you present these ground features as part of the results, I don’t think you need to describe why you chose pingos to study, but instead, it will be clear why you sampled them.
Line 262: “these sites had no obvious geologic….” What sites are being referred to here: wetland 1, 2 and 3? If so, is there a hot spot at wetland 1? It’s not obvious on the map. And wetland 3 is very close to swiss cheese seeps, correct? Seems like a potential geologic source.
Line 268: Can you add in a sentence after “….Wetland 3.”? Please consider adding in: “Our data is consistent with knowledge that wetlands produce significant methane from microbial degradation. This carbon is also probably recent in age versus geologic methane.” And you can give some citations for that.
Line 269: For the lack of signal at site 9, were you downwind of the wetland?
Line 273: Can you give reference for this?
Line 288: What is the evidence for thermogenic methane at this site? Is there any reason to think that the seep could be thermogenic? -53 from the keeling intercept still seems quite low for thermogenic. It just seems that this is most likely an oxidation signal. But as you say, it is still possible there could be thermogenic.
Line 294: I believe the location of Lake 1 is the same as “swiss cheese” that has been visited by the GSC before. Are there reports that report the bubble signature isotope value? It might be informative in your discussion of your values.
References:
Lapham, L.L., Dallimore, S., Magen, C., Henderson, L.C., Powers, L., Gonsior, M., Clark, B., Cote, M., Fraser, P., and Orcutt, B.N. (2021). Microbial greenhouse gas dynamics associated with warming coastal permafrost, western Canadian Arctic. Frontiers in Earth Sciences https://doi.org/10.3389/feart.2020.582103.
Citation: https://doi.org/10.5194/egusphere-2022-549-CC1 - AC1: 'Reply on CC1', Daniel Wesley, 11 Apr 2023
-
RC1: 'Comment on egusphere-2022-549', Anonymous Referee #1, 13 Dec 2022
This manuscript aims to determine the contribution of biogenic and thermogenic methane (CH4) to CH4 fluxes from the Mackenzie River Delta (MRD) into the atmosphere. Therefore, the authors collected surface air samples from several sampling sites in the MRD and analysed their CH4 concentration and the δ13C value of CH4. To differentiate between thermogenic and biogenic CH4 they used two thresholds, assuming that CH4 with a carbon stable isotope value of > -50‰ is of thermogenic origin and CH4 with a value < -70‰ of biogenic origin. Values between -50‰ and -70‰ would indicate a mixture of both sources. The main conclusions of the manuscript are based on this assumption, which is, however, a substantial oversimplification. There are numerous studies demonstrating carbon stable isotope signatures of biogenic CH4 of > -70‰, in particular CH4 from acetoclastic methanogenesis (see e.g. Bréas et al. (2001), Chanton & Smith (1993), Conrad (2005)), including from permafrost affected wetlands (Nakagawa et al., 2002). Furthermore, CH4 emitted from highly heterogeneous wetlands as the one studied here are affected by microbial CH4 oxidation, which causes the carbon stable isotope signature of released CH4 to increase, also to values above -50‰. There are many studies about the impact of CH4 oxidation in northern wetlands on CH4 fluxes and the carbon stable isotope signatures of released CH4 (e.g. Happell et al. (1994), Vaughn et al. (2016)), but the effect of CH4 oxidation on carbon stable isotope values of CH4 is mentioned only very briefly. Since the carbon stable isotope values of released methane may vary strongly, e.g. due to different CH4 production pathways, CH4 transport and CH4 oxidation, carbon stable isotope values between -42‰ and -88‰, as presented in this manuscript, may be explained by biogenic sources alone and are also reported for northern wetlands not affected by fluxes of thermogenic methane. Hence, I do not see that carbon stable isotope values of released methane alone provide robust information to answering the central research question of this manuscript, the contribution of biogenic and thermogenic methane to methane release in the MRD. To give substantial information on this question, further data are needed, e.g. the δD signatures of CH4, its 14C age, or the concentration of further hydrocarbons.
Furthermore, methods of gas sampling and analysis and calculation of the source δ13C value should be described in more detail. What was the gas flow while flushing the Synflex tube, how often was it flushed with the air sample to ensure that no contaminations remained? How was gas collected with the LI-7810, how often and at which positions? Why were gas samples collected in the Synflex tube, if they were not analysed for d13C of CH4? How far is ‘as close as possible’? Please clearly describe which samples were collected for which analysis. Particularly for the Keeling-plots and the calculation of the δ13C source values it should be clearly explained from which collected sample the CH4 concentrations and δ13C values were analysed.
Finally I suggest restructuring the Results and Discussion section. In the current version of the manuscript a substantial part of the results are presented (or repeated) in the Discussion.
Specific comments:
L11: To my understanding, CH4 is released but not produced from thermogenic sources. Please clarify
L30: What means ‘conductive for biogenic CH4 production’? Please clarify
L56f: This assumption is an oversimplification (see above)
L85: Do mixed sources contain other CH4 than biogenic and thermogenic? Please clarify.
L159f: The sampling of surface gas with the aluminium tubing is unclear to me. How was the tube filled and how it was possible to analyse discrete samples from this tube? Please explain in more detail.
L 177: value not ratio
L 275f: This might just indicate a higher contribution of CH4 oxidation in summer than in winter, when the surface soil is frozen.
L282f: Methane oxidation in permafrost-affected wetlands is most important in the ice-free summer. High CH4 oxidation might even cause the lack of ebullition and explain the high δ13C value of CH4.
L306: It is unclear, which data indicate the multiple sources of CH4. Please clarify.
L311 f: The second part of this sentence is unclear.
L318f: What are ‘eddy covariance hotspot locations’ and which data of this study verify these?
Cited literature:
Bréas O, Guillou C, Reniero F, Wada E (2001) The Global Methane Cycle: Isotopes and Mixing Ratios, Sources and Sinks. Isotopes in Environmental and Health Studies, 37, 257-379.
Chanton JP, Smith LK (1993) Seasonal variations in the isotopic composition of methane associated with aquatic macrophytes. In: Biogeochemistry of Global Change, Radiatively Active Trace Gases. (ed Oremland RS) pp 618-632. London, Chapman & Hall Inc.
Conrad R (2005) Quantification of methanogenic pathways using stable carbon isotopic signatures: a review and a proposal. Organic Geochemistry, 36, 739–752.
Happell JD, Chanton JP, Showers WS (1994) The influence of methane oxidation on the stable isotopic composition of methane emitted from Florida swamp forests. Geochimica et Cosmochimica Acta, 58, 4377-4388.
Nakagawa F, Yoshida N, Nojiri Y, Makarov VN (2002) Production of methane from alasses in eastern Siberia: Implications from its C-14 and stable isotopic compositions. Global Biogeochemical Cycles, 16.
Vaughn LJS, Conrad ME, Bill M, Torn MS (2016) Isotopic insights into methane production, oxidation, and emissions in Arctic polygon tundra. Global Change Biology, 22, 3487-3502.
Citation: https://doi.org/10.5194/egusphere-2022-549-RC1 - AC2: 'Reply on RC1', Daniel Wesley, 11 Apr 2023
-
RC2: 'Comment on egusphere-2022-549', Anonymous Referee #2, 21 Feb 2023
Review of Characterization of atmospheric methane release in the outer Mackenzie River Delta from biogenic and thermogenic sources
Synopsis:
The authors present results of ground-based measurements of atmospheric methane concentrations and 13C isotope signatures made in the Mackenzie River Delta, NWT, CN in 2019 and 2021 near locations which had anomalously high fluxes during airborne eddy covariance surveys in 2012 and 2013. They interpret their findings to mean that these previously determined “hotspot” areas were generated either by thermogenic CH4 sources, biogenic CH4 sources, or mixtures of both sources. They conclude that their study only provided a snapshot and that more robust methods are needed to more confidently confirm the origins of the CH4 hotspots identified by the airborne surveys.
The authors’ dataset is indeed rare, is collected in a climate-sensitive environment, and could be valuable to the scientific community. However, their incomplete methods and lack of description regarding some of the assumptions they make about their sampling approach greatly limit the interpretation of their results. I believe this study will make an important contribution, but not before the authors fully articulate some critical shortcomings about their sampling and whether or not it can actually be used to interpret previously-discovered methane emission hotspots in the Mackenzie River Delta.
General Comments:
This study highlights the challenges of relating remotely sensed patterns to those which can be observed from the ground. A proper ground-truthing of remotely sensed signals requires careful consideration of the limitations of both approaches. Aside from a comment that it is impractical to sample the airborne-derived hotspots from the ground, there was little discussion regarding how representative their ground sampling was and whether it could be used to interpret the airborne-derived hotspots at all. The significant spatial disparity between the previous airborne survey and the present ground survey, the minimum spatial resolution at which the airborne survey can be confidently interpreted, and the limited description/presentation of the ground-based survey’s methods all raise considerable uncertainty in the study’s findings. At a minimum, this study should more explicitly (and quantitatively) describe their findings in space relative to the remotely-observed hotspots. For example, can the walking transect data be represented in figure 1? Were samples collected up-wind or down-wind of the hotspots? ...etc.
How representative are their discrete samples of the airborne-observed hotspots to which they attempt to attribute their findings? From Figure 1, one can see that the airborne hotspots are several kilometers across. Are the authors attempting to use singular samples from discrete locations to describe CH4 sources from kilometers-wide “hotspots?” The ground-based sampling methods should further discuss the assumptions, strategy, and potential biases involved in their plan. For example, was sampling meant to describe individual features (e.g. pingos) on the ground or the large-area hotspots more generally? How do the authors reconcile these scale differences? If they intended to apply isotopic signatures and flux rates to individual features on the ground, why not use chambers or other methods to explicitly isolate the air from those sources? If they intended on sampling the air to infer something about the surrounding landscape, why not conduct a more formal wind analysis or “footprint” determination and then describe the features within the footprint? It seems that the approaches fall somewhere in between these two objectives, but as a result, fail to robustly describe the environment in which they were made.
The authors have not clearly articulated their assumptions regarding separating background air from the air that contains mixtures of CH4 originating from their hotspots. Can their methods sufficiently distinguish between the two? How do they know that the background air wasn’t already a mixture of other local sources? How do variable wind conditions (velocity and direction) affect their sampling approach? There is little description regarding the direction of the wind and how this could affect their goal to interpret the hotspots.
No reference is made to Figure S1 or Figure S2 in the manuscript text. This was surprising, especially since Figure S2 contains the walking transect data.
Specific Comments:
Line 39: A more recent/updated reference is available for the global CH4 budget. Saunois, M., Stavert, A.R., Poulter, B., Bousquet, P., Canadell, J.G., Jackson, R.B., Raymond, P.A., Dlugokencky, E.J., Houweling, S., Patra, P.K. and Ciais, P., 2020. The global methane budget 2000–2017. Earth system science data, 12(3), pp.1561-1623.
Line 76: Without context, a “hotspot” can have an ambiguous definition. Since 5 mg CH4 m-2 d-1is not a particularly high flux in the context of typical wetland/lake emissions, I think that the authors need to emphasize on what spatial scale this would be considered a high flux or a hotspot. This would also help set the stage for placing the new observations in the proper spatial context.
Lines 81 - 86: Check verb tenses in this paragraph. It currently reads like this work has not happened yet- as one would write in a proposal.
Line 102: Bowen et al. (2008) is plural, so probably change “has” to “have”
Lines 119 - 121: Results should be moved to the results section.
Line 125: What does “focused” mean here? It’s also misspelled. This sentence could use a rewrite to improve clarity.
Line 155: Could sampling only in wetlands bias the observations towards one emission pathway (or source) over another? Do thermogenic emissions only occur in wetlands and/or lakes? Or can they occur in dry areas as well? Some discussion and communication of assumptions around this sampling strategy are warranted (see general comment).
Lines 159 – 164: Is this an “air core?” or something else? This description is somewhat vague. It raises several questions. What is the objective of the “walking transect?” How was the tubing filled? What are “analytical determinations?” Mixing of what?
Line 168: The aerial EC hotspot sites? Or on walking transects, or both? Be more specific about the spatial representativeness of the observations.
Line 175: How were fluxes derived from chamber data? How many vials were analyzed per flux observation? For how long? Linear extrapolation? Was non-linearity observed? If so, how was this handled in data QA/QC? These parameters are typically reported alongside chamber flux data.
Line 221: What is “reasonable confidence?” Can you be more quantitative? Correlation statistic? Otherwise, this is not very informative.
Line 222: Phrases like “substantially enriched” and “relatively close” are not very informative and weaken the impact of the claims being made. Please quantify where possible.
Line 266: How was “background” determined?
Citation: https://doi.org/10.5194/egusphere-2022-549-RC2 - AC3: 'Reply on RC2', Daniel Wesley, 11 Apr 2023
Interactive discussion
Status: closed
-
CC1: 'Comment on egusphere-2022-549', Laura Lapham, 06 Dec 2022
Summary: The goal of the paper was to take a first look at the source of the hot spots previously found by aerial surveys using stable carbon isotopes within the Mackenzie River Delta (MRD). The authors state that the MRD has thermogenic gas seepage and biogenic methane atmospheric sources. Given there are so many lakes, this is not surprising, but there should be references for that statement. The authors hypothesize (line 84) that the hot spots come from “thermogenic, biogenic, and mixed sources….”. This hypothesis could be strengthened by picking one source based on previous literature. For example, the Kohnert paper clearly suggests that the hot spots are thermogenic because they are so high, but didn’t have any detailed source information to support this. This paper does. Or maybe the fact that there are so many lakes emitting biogenic methane means the atmospheric methane would make one hypothesize these hot spots would be biogenic? The strength of this paper is in using stable carbon isotope ratios of methane in air from ground surveys. The paper could benefit from some more details on the sampling sites, the sampling protocols, the assumptions going into the keeling plots, and overall conclusions drawn from the study. Overall, I think the data is interesting, and definitely novel and worthy of publishing. Hopefully the comments below could help streamline the message:
- The introduction could be streamlined and strengthened. Currently, the text suggests that the MRD has biogenic methane sources, but I couldn’t see any references to support that. Maybe work from Lance Lesack’s group would be helpful here for in and around Inuvik. On line 60, there is also mention of production of methane in the organic rich active layer but has no reference. There was a recent paper to conduct permafrost incubation studies (Lapham et al., 2021) in Tuk, but that didn’t take place in the active layer. In terms of streamlining, the sentence from line 67-68 could be cut; while it is important to measure emissions of methane, since the current study didn’t do this, it’s not necessary.
- The study location and methods sections could be reorganized to be sure the proper information is conveyed in each section. For example, the setting was described in section2, and then there was a “study location” in section 3. The methods section should only give the methods used, and results should be reported in the results section.
- Overall figures are sufficient with some revision. For example, figure 1 could be more informative if the walking transects were shown on the pictures, and the direction of wind and location of seeps were also shown. As it is now, without any labels, it’s difficult to see where the seeps are (unless you know), and at Lake 1, unless you know the picture shows the lake is ice covered, it’s hard to know what is happening. The caption says “prominent ebullition” but unless you know what to look for, it’s not clear with the captions. Furthermore, supplemental figure S2 is not mentioned in the text, yet addresses some concerns about wind direction and the location of the transects. If the quality could be improved, or information combined between figure 1 and figure S2, this could help.
- The chamber flux data seems to be an add-on. Is it necessary for the message of the paper? If so, more details will be needed (like how the samples were analyzed) and some context of what the fluxes mean in terms of other environments (for the discussion). If not needed, please consider taking out of the paper.
Detailed comments:
Line 34: add in “oxidation” after production and before transport. I think it’s important with some of the conclusions drawn to get the idea of oxidation into the text earlier.
Line 51: As numbers, delta values can be high or low, positive or negative, but not heavy or light. Please use “high” or “low” d13C values. Be sure to check this throughout and change accordingly.
Line 75: “geologic origin”: maybe define what you mean and that it could be made up of thermogenic methane produced deep and migrates, or biogenic or rather, microbial methane produced in the permafrost? See comment from line 224.
Line 78-80: The sentence starting “interestingly, ….” Is a confusing sentence. How can the sources behave differently than the current understanding? Please reword to make more clear.
Line 117: it would be helpful to have a mark on the map for Tuk.
Lines around 119: describe your sites a little more than just a name. What is “site 9”? There isn’t a description in figure 2, and on the figure 1 map, it looks like it’s upwind of the hot spot from Kohnert. What were the winds like when you sampled it? Also, the sentence here of “Of the five airborne eddy covariance hotspots…..” is a result. It should be moved to that section. In thinking more about the study design, if the idea was to ground truth the Kohnert hot spots, there is a missed opportunity to bring out some novelty of this study. For example, if we look at the “pingo 1” site, the fact that there is a pingo there is important, right? Does that already add information not gained from the Kohnert study? That you observed a pingo there? And what would that mean, what would the observation of a pingo mean for methane emissions? Aren’t they by definition conduits of some sort? Or what about pingos being surrounded by wetlands? Additionally, at wetland 3, the fact that there are wetlands characterizing that hot spot is interesting information. I’d almost envision numbering the hotspots sequentially and then giving them their ground feature names as done on figure 1 (this is just a suggestion and maybe not helpful, it just seems interesting what ground features underlay the aerial hot spots). Yet, such an approach really send home the message that is directly inline with your goal, to groundtruth the aerial survey hot spots.
Table S1: The raw data should be available somewhere for review. Will it be available in a database somewhere, or as a supplemental table here? Also, in this table, for the “source d13C values, R2, and max CH4” in the fourth column, those are the same values as in figure 1. I’m not sure what this table is adding except to give exact locations. Can you replace this with the raw data, from which you derive the y-intercept from keeling plots? Also, Table S1 gives a “site type” as Polar V. How is the Polar V a site type? Also, you give into in this table for the low R2 values, what made you pick the R2 value cut offs you did? For example, you kept an R2 of 0.48 but didn’t talk about the site with R2 of 0.434. This sort of thing should be mentioned in the methods.
Line 123: how did you determine where discrete point samples were collected? What was the strategy? Upwind, downwind, etc? This is a study location section, is this the best place for that information? I think you should be explicit that the strategy was to target the hot spots, and if you adopt a sample numbering scheme like numbering the hot spots, this strategy will be clear.
Line 125: Is “Lake 1” known as another lake by the community? Is that “shot hole” lake or Swiss cheese? And Channel Seep, is that channel seep 1? There could be reports that have some isotope data reported to help aid you in interpretations.
Line 126: the “observations of ebullitions seen in open water in summer”, were those your observations made in this sampling trip or previous knowledge? Cite previous knowledge.
Line 126: how close is “as close as possible” to observed ebullitions? Were you on land, upwind, downwind? How long were walking transects, 10m? This discussion in these 3 lines around 130 are more about sample collection. As such, they should be moved to that section of the paper.
Lines 131-142: This paragraph seems more high level than where it is situated in the paper. I suggest reorganizing this study location section.
Line 149: Are you missing the word “under”? Permafrost “under” Tuk…..
Line 154: what is precision of handheld GPS and what model, make? What is the precision of the licor and the picarro CRDS? It’s important to mention the cavity ring down spectrometer (CRDS) part of the instrument since that is how the measurement is made. And what standards were used to calibrate the concentration and isotope measurements? The placement of the analysis instruments seems out of place here since this is the sample collection section.
Line 156: Please cite the airborne work paper, since it was not done in this study.
Line 158: “photographs of each site….” I think this should be moved to study location, and not sample collection. You are using the photographs to describe the sites, correct? If so, they really set the stage for the setting, which belongs elsewhere.
Line 159: “walking transections…” what was pumping rate? What are dimensions of tubing? Is 6mm OD or ID?
Line 163: “Mixing between sample collection and analysis is limited due to small diameter of tubing”. Have you proved that? Or is there a paper you can cite for this? The reason I ask is that the pumping rate is pretty fast, so I would imagine your sample will smear alone the edges of the tubing and mix along the way it’s filling. Please give more details as to the accuracy of this approach.
Line 166: Why did you pick 1 meter above ground level? Did you ever try to go down to ground level? Did you see any change in the concentration? Or is 1m desirable because things are more mixed and you are trying not to see a ground signal? What is the thought behind this?
Line 171: Where are the flux chamber measurement data? Also, this section is confusing as written. It looks like 2 chambers were used, the automated one from licor, but I can’t tell what was used for Lake 1, using manual extraction of samples. And the dimensions are for a flux chamber, but there is only one, yet it seems two chambers were used? Also, can you discuss how only allowing 1 hour between collars installed and measurement might change your results?
Line 178: “Keeling plot analysis”. I am not sure this is the right terminology. You analyzed keeling plots to determine the stable carbon isotope signature of the methane source. That phrase, keeling plot analysis, is used several times, so maybe it is the right term but it could be better to say “We constructed keeling plots with the discrete transect data to determine ….” And also cite the “common approach” of using keeling plots. I agree that it is common, but I am not sure it is common to take discrete measurements in the horizontal direction versus the vertical direction. Meaning, I thought keeling plots were always done in the vertical collection of air in a forest canopy, for example. If that is not true, it’s probably still making a note of this since it does seem a bit novel to use this technique for the walking transects. Or maybe there are papers to show this approach used in this way.
Line 193: Do you have the bubble isotope values from historical data by the GSC to compare to your point measurements?
Line 197: Why is walking transect data shown as average data? The maximum value of 12ppm is very interesting and seems important to know where you were in comparison to the “wetland”. Is it possible to do this from the GPS location data you obtained?
Line 203: The sentence of “estimates of source….” Seems a bit premature. Since this is the methods section, could you first say that the keeling plots are shown in figure 3, and show intercepts of X, Y, Z, which indicate the source of the methane? And also give the R2 values? As it is now, you don’t mention figure 3, so it’s unclear where these numbers come from. The figure 3 caption mentions a “grey region” but there are no grey regions on the figures. And finally, there is a formatting issue with the equation written on the figures. It is also important to mention the cut off you used for the R2 values.
Line 208: “Keeling plot values” is not quite accurate. Maybe say “keeling plot y-intercepts”. It’s interesting the seasonal component of these values. Is seems reasonable to think that during the winter, there isn’t as much methane oxidation, which leads to the lower values. I think you mention that in discussion.
Line 211: “Flux rates” isn’t accurate. A flux is a concentration per area per time, a rate is distance per time. I think you just mean fluxes here. It also seems like the chamber fluxes are an afterthought since the fluxes aren’t given here. And they are put in a supplemental table. Are they needed?
Lines 215-218: change “values” to “concentrations”.
Line 217: take out “were”. What is an “observation” in this context? There wasn’t mention of observations before now. Where does the 2013 number come from? Do you mean the discrete measurements? Same question for the 1850 observations at pingo 1. This section is a bit confusing since it’s also written like results, but yet in the discussion section.
Line 224: For the general reader, it might help to define geologic source in the introduction to put this conclusion in more context.
Line 240: Can you give values for the pingo 1 and 2 site in the text? It would help the reader not have to flip back to the figures.
Line 241: Is this new data? I didn’t see the reporting of the methane concentrations over the pingo features themselves. And this sentence is also a bit confusing as to what you mean. Did you do 2 transects for Pingo 1? Seems like supplemental figure S2 (top right) could be helpful to show what you mean here. You could refer to that figure here, but that figure quality needs to be improved.
Line 243: Only wind speed is reported in table 1. Wind direction is also key that would be important to show in that table. Or did you always collect samples downwind? Do you see a correlation with wind speed? I would think that the higher the wind speed, the further away the source of that gas could be.
Paragraph starting at line 254: This is a great reason you chose to sample pingos. Could you move it to the introduction to help set up the “why” for your study? After reading the paper again, it seems you didn’t set out to study pingos, per say, but you found those pingos at the ground features under the aerial hot spots. If you present these ground features as part of the results, I don’t think you need to describe why you chose pingos to study, but instead, it will be clear why you sampled them.
Line 262: “these sites had no obvious geologic….” What sites are being referred to here: wetland 1, 2 and 3? If so, is there a hot spot at wetland 1? It’s not obvious on the map. And wetland 3 is very close to swiss cheese seeps, correct? Seems like a potential geologic source.
Line 268: Can you add in a sentence after “….Wetland 3.”? Please consider adding in: “Our data is consistent with knowledge that wetlands produce significant methane from microbial degradation. This carbon is also probably recent in age versus geologic methane.” And you can give some citations for that.
Line 269: For the lack of signal at site 9, were you downwind of the wetland?
Line 273: Can you give reference for this?
Line 288: What is the evidence for thermogenic methane at this site? Is there any reason to think that the seep could be thermogenic? -53 from the keeling intercept still seems quite low for thermogenic. It just seems that this is most likely an oxidation signal. But as you say, it is still possible there could be thermogenic.
Line 294: I believe the location of Lake 1 is the same as “swiss cheese” that has been visited by the GSC before. Are there reports that report the bubble signature isotope value? It might be informative in your discussion of your values.
References:
Lapham, L.L., Dallimore, S., Magen, C., Henderson, L.C., Powers, L., Gonsior, M., Clark, B., Cote, M., Fraser, P., and Orcutt, B.N. (2021). Microbial greenhouse gas dynamics associated with warming coastal permafrost, western Canadian Arctic. Frontiers in Earth Sciences https://doi.org/10.3389/feart.2020.582103.
Citation: https://doi.org/10.5194/egusphere-2022-549-CC1 - AC1: 'Reply on CC1', Daniel Wesley, 11 Apr 2023
-
RC1: 'Comment on egusphere-2022-549', Anonymous Referee #1, 13 Dec 2022
This manuscript aims to determine the contribution of biogenic and thermogenic methane (CH4) to CH4 fluxes from the Mackenzie River Delta (MRD) into the atmosphere. Therefore, the authors collected surface air samples from several sampling sites in the MRD and analysed their CH4 concentration and the δ13C value of CH4. To differentiate between thermogenic and biogenic CH4 they used two thresholds, assuming that CH4 with a carbon stable isotope value of > -50‰ is of thermogenic origin and CH4 with a value < -70‰ of biogenic origin. Values between -50‰ and -70‰ would indicate a mixture of both sources. The main conclusions of the manuscript are based on this assumption, which is, however, a substantial oversimplification. There are numerous studies demonstrating carbon stable isotope signatures of biogenic CH4 of > -70‰, in particular CH4 from acetoclastic methanogenesis (see e.g. Bréas et al. (2001), Chanton & Smith (1993), Conrad (2005)), including from permafrost affected wetlands (Nakagawa et al., 2002). Furthermore, CH4 emitted from highly heterogeneous wetlands as the one studied here are affected by microbial CH4 oxidation, which causes the carbon stable isotope signature of released CH4 to increase, also to values above -50‰. There are many studies about the impact of CH4 oxidation in northern wetlands on CH4 fluxes and the carbon stable isotope signatures of released CH4 (e.g. Happell et al. (1994), Vaughn et al. (2016)), but the effect of CH4 oxidation on carbon stable isotope values of CH4 is mentioned only very briefly. Since the carbon stable isotope values of released methane may vary strongly, e.g. due to different CH4 production pathways, CH4 transport and CH4 oxidation, carbon stable isotope values between -42‰ and -88‰, as presented in this manuscript, may be explained by biogenic sources alone and are also reported for northern wetlands not affected by fluxes of thermogenic methane. Hence, I do not see that carbon stable isotope values of released methane alone provide robust information to answering the central research question of this manuscript, the contribution of biogenic and thermogenic methane to methane release in the MRD. To give substantial information on this question, further data are needed, e.g. the δD signatures of CH4, its 14C age, or the concentration of further hydrocarbons.
Furthermore, methods of gas sampling and analysis and calculation of the source δ13C value should be described in more detail. What was the gas flow while flushing the Synflex tube, how often was it flushed with the air sample to ensure that no contaminations remained? How was gas collected with the LI-7810, how often and at which positions? Why were gas samples collected in the Synflex tube, if they were not analysed for d13C of CH4? How far is ‘as close as possible’? Please clearly describe which samples were collected for which analysis. Particularly for the Keeling-plots and the calculation of the δ13C source values it should be clearly explained from which collected sample the CH4 concentrations and δ13C values were analysed.
Finally I suggest restructuring the Results and Discussion section. In the current version of the manuscript a substantial part of the results are presented (or repeated) in the Discussion.
Specific comments:
L11: To my understanding, CH4 is released but not produced from thermogenic sources. Please clarify
L30: What means ‘conductive for biogenic CH4 production’? Please clarify
L56f: This assumption is an oversimplification (see above)
L85: Do mixed sources contain other CH4 than biogenic and thermogenic? Please clarify.
L159f: The sampling of surface gas with the aluminium tubing is unclear to me. How was the tube filled and how it was possible to analyse discrete samples from this tube? Please explain in more detail.
L 177: value not ratio
L 275f: This might just indicate a higher contribution of CH4 oxidation in summer than in winter, when the surface soil is frozen.
L282f: Methane oxidation in permafrost-affected wetlands is most important in the ice-free summer. High CH4 oxidation might even cause the lack of ebullition and explain the high δ13C value of CH4.
L306: It is unclear, which data indicate the multiple sources of CH4. Please clarify.
L311 f: The second part of this sentence is unclear.
L318f: What are ‘eddy covariance hotspot locations’ and which data of this study verify these?
Cited literature:
Bréas O, Guillou C, Reniero F, Wada E (2001) The Global Methane Cycle: Isotopes and Mixing Ratios, Sources and Sinks. Isotopes in Environmental and Health Studies, 37, 257-379.
Chanton JP, Smith LK (1993) Seasonal variations in the isotopic composition of methane associated with aquatic macrophytes. In: Biogeochemistry of Global Change, Radiatively Active Trace Gases. (ed Oremland RS) pp 618-632. London, Chapman & Hall Inc.
Conrad R (2005) Quantification of methanogenic pathways using stable carbon isotopic signatures: a review and a proposal. Organic Geochemistry, 36, 739–752.
Happell JD, Chanton JP, Showers WS (1994) The influence of methane oxidation on the stable isotopic composition of methane emitted from Florida swamp forests. Geochimica et Cosmochimica Acta, 58, 4377-4388.
Nakagawa F, Yoshida N, Nojiri Y, Makarov VN (2002) Production of methane from alasses in eastern Siberia: Implications from its C-14 and stable isotopic compositions. Global Biogeochemical Cycles, 16.
Vaughn LJS, Conrad ME, Bill M, Torn MS (2016) Isotopic insights into methane production, oxidation, and emissions in Arctic polygon tundra. Global Change Biology, 22, 3487-3502.
Citation: https://doi.org/10.5194/egusphere-2022-549-RC1 - AC2: 'Reply on RC1', Daniel Wesley, 11 Apr 2023
-
RC2: 'Comment on egusphere-2022-549', Anonymous Referee #2, 21 Feb 2023
Review of Characterization of atmospheric methane release in the outer Mackenzie River Delta from biogenic and thermogenic sources
Synopsis:
The authors present results of ground-based measurements of atmospheric methane concentrations and 13C isotope signatures made in the Mackenzie River Delta, NWT, CN in 2019 and 2021 near locations which had anomalously high fluxes during airborne eddy covariance surveys in 2012 and 2013. They interpret their findings to mean that these previously determined “hotspot” areas were generated either by thermogenic CH4 sources, biogenic CH4 sources, or mixtures of both sources. They conclude that their study only provided a snapshot and that more robust methods are needed to more confidently confirm the origins of the CH4 hotspots identified by the airborne surveys.
The authors’ dataset is indeed rare, is collected in a climate-sensitive environment, and could be valuable to the scientific community. However, their incomplete methods and lack of description regarding some of the assumptions they make about their sampling approach greatly limit the interpretation of their results. I believe this study will make an important contribution, but not before the authors fully articulate some critical shortcomings about their sampling and whether or not it can actually be used to interpret previously-discovered methane emission hotspots in the Mackenzie River Delta.
General Comments:
This study highlights the challenges of relating remotely sensed patterns to those which can be observed from the ground. A proper ground-truthing of remotely sensed signals requires careful consideration of the limitations of both approaches. Aside from a comment that it is impractical to sample the airborne-derived hotspots from the ground, there was little discussion regarding how representative their ground sampling was and whether it could be used to interpret the airborne-derived hotspots at all. The significant spatial disparity between the previous airborne survey and the present ground survey, the minimum spatial resolution at which the airborne survey can be confidently interpreted, and the limited description/presentation of the ground-based survey’s methods all raise considerable uncertainty in the study’s findings. At a minimum, this study should more explicitly (and quantitatively) describe their findings in space relative to the remotely-observed hotspots. For example, can the walking transect data be represented in figure 1? Were samples collected up-wind or down-wind of the hotspots? ...etc.
How representative are their discrete samples of the airborne-observed hotspots to which they attempt to attribute their findings? From Figure 1, one can see that the airborne hotspots are several kilometers across. Are the authors attempting to use singular samples from discrete locations to describe CH4 sources from kilometers-wide “hotspots?” The ground-based sampling methods should further discuss the assumptions, strategy, and potential biases involved in their plan. For example, was sampling meant to describe individual features (e.g. pingos) on the ground or the large-area hotspots more generally? How do the authors reconcile these scale differences? If they intended to apply isotopic signatures and flux rates to individual features on the ground, why not use chambers or other methods to explicitly isolate the air from those sources? If they intended on sampling the air to infer something about the surrounding landscape, why not conduct a more formal wind analysis or “footprint” determination and then describe the features within the footprint? It seems that the approaches fall somewhere in between these two objectives, but as a result, fail to robustly describe the environment in which they were made.
The authors have not clearly articulated their assumptions regarding separating background air from the air that contains mixtures of CH4 originating from their hotspots. Can their methods sufficiently distinguish between the two? How do they know that the background air wasn’t already a mixture of other local sources? How do variable wind conditions (velocity and direction) affect their sampling approach? There is little description regarding the direction of the wind and how this could affect their goal to interpret the hotspots.
No reference is made to Figure S1 or Figure S2 in the manuscript text. This was surprising, especially since Figure S2 contains the walking transect data.
Specific Comments:
Line 39: A more recent/updated reference is available for the global CH4 budget. Saunois, M., Stavert, A.R., Poulter, B., Bousquet, P., Canadell, J.G., Jackson, R.B., Raymond, P.A., Dlugokencky, E.J., Houweling, S., Patra, P.K. and Ciais, P., 2020. The global methane budget 2000–2017. Earth system science data, 12(3), pp.1561-1623.
Line 76: Without context, a “hotspot” can have an ambiguous definition. Since 5 mg CH4 m-2 d-1is not a particularly high flux in the context of typical wetland/lake emissions, I think that the authors need to emphasize on what spatial scale this would be considered a high flux or a hotspot. This would also help set the stage for placing the new observations in the proper spatial context.
Lines 81 - 86: Check verb tenses in this paragraph. It currently reads like this work has not happened yet- as one would write in a proposal.
Line 102: Bowen et al. (2008) is plural, so probably change “has” to “have”
Lines 119 - 121: Results should be moved to the results section.
Line 125: What does “focused” mean here? It’s also misspelled. This sentence could use a rewrite to improve clarity.
Line 155: Could sampling only in wetlands bias the observations towards one emission pathway (or source) over another? Do thermogenic emissions only occur in wetlands and/or lakes? Or can they occur in dry areas as well? Some discussion and communication of assumptions around this sampling strategy are warranted (see general comment).
Lines 159 – 164: Is this an “air core?” or something else? This description is somewhat vague. It raises several questions. What is the objective of the “walking transect?” How was the tubing filled? What are “analytical determinations?” Mixing of what?
Line 168: The aerial EC hotspot sites? Or on walking transects, or both? Be more specific about the spatial representativeness of the observations.
Line 175: How were fluxes derived from chamber data? How many vials were analyzed per flux observation? For how long? Linear extrapolation? Was non-linearity observed? If so, how was this handled in data QA/QC? These parameters are typically reported alongside chamber flux data.
Line 221: What is “reasonable confidence?” Can you be more quantitative? Correlation statistic? Otherwise, this is not very informative.
Line 222: Phrases like “substantially enriched” and “relatively close” are not very informative and weaken the impact of the claims being made. Please quantify where possible.
Line 266: How was “background” determined?
Citation: https://doi.org/10.5194/egusphere-2022-549-RC2 - AC3: 'Reply on RC2', Daniel Wesley, 11 Apr 2023
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 547 | 255 | 26 | 828 | 63 | 13 | 10 |
- HTML: 547
- PDF: 255
- XML: 26
- Total: 828
- Supplement: 63
- BibTeX: 13
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Cited
Daniel Wesley
Scott Dallimore
Roger MacLeod
Torsten Sachs
David Risk
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1027 KB) - Metadata XML
-
Supplement
(159 KB) - BibTeX
- EndNote
- Final revised paper