the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 24 Apr 2023
Ground-to-UAV, laser-based, multi-species emissions quantification at long standoff distances
Eleanor M. Waxman
Eli Hoenig
Daniel Hesselius
Christopher Chaote
Ian Coddington
Nathan R. Newbury
Abstract. Determination of trace gas emissions from sources is critical for understanding and regulating air quality and climate change. Here, we demonstrate a method for rapid quantification of the emission rate of multiple gases from simple and complex sources using a mass-balance approach with a spatially scannable open-path sensor – in this case, an open-path dual-comb spectrometer. The open-path spectrometer measures the total column density of gases between the spectrometer and a retroreflector mounted on an unmanned aerial vehicle (UAV). By measuring slant columns at multiple UAV altitudes downwind of a source (or sink), the total emission rate can be rapidly determined without the need for an atmospheric dispersion model. Here, we demonstrate this technique using controlled releases of CH4 and C2H2. We show an emission rate determination to within 50 % of the known flux with a single 10-minute flight and within 10 % of the known flux after 10 flights. Furthermore, we estimate a detection limit for CH4 emissions to be 0.03 g CH4/s. This detection limit is approximately the same as the emissions from 25 head of beef cattle and is less than the average emissions from a small oil field pneumatic controller. Other gases including CO2, NH3, HDO, ethane, formaldehyde (HCHO), CO, and N2O can be measured by simply changing the dual-comb spectrometer.
-
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
(2719 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2719 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Kevin C. Cossel et al.
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-691', Anonymous Referee #1, 28 Apr 2023
The work by Cossel at al. “Ground-to-UAV, laser-based, multi-species emissions quantifications at long standoff distances” is well written, is a relevant contribution, and the topic discussed is within the scope of AMT. I recommend publication after minor revisions. However, the title “Ground-to-UAV, laser-based, quantifications of CH4 and C2H2 at long standoff distances” would more adequately describe what is presented (as the authors correctly state, the method is certainly expandable to a variety of species, nevertheless this is an extrapolation, while the title should as accurately as possible depict what actually is delivered).
Comments:
Not much detail is provided on the quality of the recorded interferograms and of spectra derived from these. It would be interesting to give the reader a feeling of the level of degradation introduced by applying DCS in the open field with UAV-borne retroreflector. Specifically, I would be interested to learn how variable the interferograms are due to variable coupling efficiency. The spectral envelope appears surprisingly structured (fig 1a): are these undulations and the overall spectral intensity level variable from spectrum to spectrum?
The background concentrations are established in different ways for CH4 and C2H2. What is the advantage of introducing an additional separate sensor for measuring the background? (1) The use of two different sensors will always introduce some level of bias (I understand that the CRDS was calibrated properly, but I would expect a calibration bias of the CH4 band intensity reported in HITRAN 2008 in the range of 1 … 2%). (2) The UAV needs to climb up beyond the plume signal for a useful flux measurement anyway, and in a complex terrain covered with different sources, I would expect the CH4 concentration measured at a higher altitude to provide a more reliable background value than a measurement taken near ground.
As shown in Figure 5, while the CH4 result is based on reasonably sound statistics (10 release results, 4 background results), there are only two C2H2 release results (and 4 background results). Why there are only 4 C2H2 background results? Could not all flights performed without C2H2 release (so 8 flights) be used for deriving C2H2 background values? Why have only two C2H2 release experiments been conducted?
I am not sure to fully understand the discussion of plume dynamics (lines 270 - 275). Is this “plume centroid offset” an elongation along the horizontal or along the vertical? If I understand correctly, it is interpreted as a horizontal elongation resulting in an uncertainty in d. I would imagine that both horizontal and vertical variations during the measurement process are important, as the concentrations used for evaluating the integral along the altitude coordinate (equation 6) are actually not measured simultaneously, but during ~3 min (as I would estimate from fig 1a), so while the undulating plume is passing by. For further quantification of this source of uncertainty, performing a longer measurement with the UAV resting at an altitude corresponding to the average plume height would be informative.
Minor technical comments / corrections:
In my feeling, it would be useful to provide (in a short appendix) some more detail on the “four different days” mentioned in section 3.2.1. Please provide the actual dates, some relevant details on the location (character of area, surface roughness), and some information on meteorological conditions during flights (average wind speed, variability of wind direction).
Line 241 typo “will lead to lead directly to”
Line 277 “ … a flux of 0.22 g/s.” -> “ … a CH4 flux of 0.22 g/s.” (as measurement noise level is species dependent)
Citation: https://doi.org/10.5194/egusphere-2023-691-RC1 - AC1: 'Reply to RC1 and RC2', Kevin Cossel, 15 Sep 2023
-
RC2: 'Comment on egusphere-2023-691', Anonymous Referee #2, 10 Jul 2023
Review Cossel et al., Ground to UAV… 2023-691.
This is a well put together and concisely written article that demonstrates the use case for this type of methane and ethane emission measurement. It is a shame that the controlled release to demonstrate the use is so limited in scope as it would have been fantastic to understand the limitations in terms of limit of detection and how uncertainties trend with increased emission rates. However, as these experiments are now complete, I expect I will have to hope for an expanded experiment for this to be demonstrated (If there are more data from these experiments, please can this be included!).
In my opinion, this paper can be published with relatively minor corrections / clarifications. Most of the concerns regard full quantification of the uncertainty and lack of detail around steps in the experimental process.
- The experimental set up for the controlled release.
L120 onwards.
There is scant description surrounding the controlled release. There needs to be sufficient detail in the methodology so that this experiment could be repeated. My recommendation would be to have a SI with the description of the controlled release in detail, potentially including photos, set-up, details of the equipment used, gas compositions, uncertainties inherent in the set up (and how they may propagate uncertainties through to the measurement results). Understanding how controlled releases are set up is an important component of being able to understand how any type of flux measurements are verified.
- Characterization of the background
L146. From an operational perspective when using this system outside of a controlled release environment are there issues with assuming that the upwind pathlength is equal to the static measurement and therefore correlating the upwind line to that point measurement? I’m wondering if you could comment on the sensitivity of that pathlength to “unaccounted for methane” and how much methane could be present above baseline without the appearance of an apparent plume. This concerns me a little in settings where multiple plumes may be present and interfering with measurements or just unknown plumes from other sources.
- Influence of multiple plumes?
Is this method suitable for multiple plumes in a single field? I’m guessing that it might be, but it would be great to have that explicitly stated one way or another and what procedure would have to be followed to allow fluxes with overlapping plumes be separated (or joined) to give an overall flux.
- General details on equipment used.
Can all items of equipment used for measurement or quantification be defined throughout the paper, there are mass flow controllers, anemometers etc without provenance.
- Explanation of the rationale behind the controlled release set up and fluxes used.
Ideally the controlled release would have had a number of releases at different rates under different wind conditions. This really only gives us an understanding of the capability of the set up under this specific set of conditions. I would like to understand the reasoning for only testing under such limited scenario and would recommend the authors to consider taking part in something along the lines of Adam Brandt’s group tests of instruments at the next possible opportunity (e.g. https://doi.org/10.1525/elementa.2022.00080).
Minor
Fig 2. Can the averaging time be expanded so that the minimum in the Allen variance can be seen in the CH4 measurement precision.
L121. Requires explanation as to why this emission rate and location is representative of a real emission.
The conclusions feel rather light, it would be good to understand where the group feels the strength of this method is compared to the other technologies on the market.
Citation: https://doi.org/10.5194/egusphere-2023-691-RC2 - AC1: 'Reply to RC1 and RC2', Kevin Cossel, 15 Sep 2023
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-691', Anonymous Referee #1, 28 Apr 2023
The work by Cossel at al. “Ground-to-UAV, laser-based, multi-species emissions quantifications at long standoff distances” is well written, is a relevant contribution, and the topic discussed is within the scope of AMT. I recommend publication after minor revisions. However, the title “Ground-to-UAV, laser-based, quantifications of CH4 and C2H2 at long standoff distances” would more adequately describe what is presented (as the authors correctly state, the method is certainly expandable to a variety of species, nevertheless this is an extrapolation, while the title should as accurately as possible depict what actually is delivered).
Comments:
Not much detail is provided on the quality of the recorded interferograms and of spectra derived from these. It would be interesting to give the reader a feeling of the level of degradation introduced by applying DCS in the open field with UAV-borne retroreflector. Specifically, I would be interested to learn how variable the interferograms are due to variable coupling efficiency. The spectral envelope appears surprisingly structured (fig 1a): are these undulations and the overall spectral intensity level variable from spectrum to spectrum?
The background concentrations are established in different ways for CH4 and C2H2. What is the advantage of introducing an additional separate sensor for measuring the background? (1) The use of two different sensors will always introduce some level of bias (I understand that the CRDS was calibrated properly, but I would expect a calibration bias of the CH4 band intensity reported in HITRAN 2008 in the range of 1 … 2%). (2) The UAV needs to climb up beyond the plume signal for a useful flux measurement anyway, and in a complex terrain covered with different sources, I would expect the CH4 concentration measured at a higher altitude to provide a more reliable background value than a measurement taken near ground.
As shown in Figure 5, while the CH4 result is based on reasonably sound statistics (10 release results, 4 background results), there are only two C2H2 release results (and 4 background results). Why there are only 4 C2H2 background results? Could not all flights performed without C2H2 release (so 8 flights) be used for deriving C2H2 background values? Why have only two C2H2 release experiments been conducted?
I am not sure to fully understand the discussion of plume dynamics (lines 270 - 275). Is this “plume centroid offset” an elongation along the horizontal or along the vertical? If I understand correctly, it is interpreted as a horizontal elongation resulting in an uncertainty in d. I would imagine that both horizontal and vertical variations during the measurement process are important, as the concentrations used for evaluating the integral along the altitude coordinate (equation 6) are actually not measured simultaneously, but during ~3 min (as I would estimate from fig 1a), so while the undulating plume is passing by. For further quantification of this source of uncertainty, performing a longer measurement with the UAV resting at an altitude corresponding to the average plume height would be informative.
Minor technical comments / corrections:
In my feeling, it would be useful to provide (in a short appendix) some more detail on the “four different days” mentioned in section 3.2.1. Please provide the actual dates, some relevant details on the location (character of area, surface roughness), and some information on meteorological conditions during flights (average wind speed, variability of wind direction).
Line 241 typo “will lead to lead directly to”
Line 277 “ … a flux of 0.22 g/s.” -> “ … a CH4 flux of 0.22 g/s.” (as measurement noise level is species dependent)
Citation: https://doi.org/10.5194/egusphere-2023-691-RC1 - AC1: 'Reply to RC1 and RC2', Kevin Cossel, 15 Sep 2023
-
RC2: 'Comment on egusphere-2023-691', Anonymous Referee #2, 10 Jul 2023
Review Cossel et al., Ground to UAV… 2023-691.
This is a well put together and concisely written article that demonstrates the use case for this type of methane and ethane emission measurement. It is a shame that the controlled release to demonstrate the use is so limited in scope as it would have been fantastic to understand the limitations in terms of limit of detection and how uncertainties trend with increased emission rates. However, as these experiments are now complete, I expect I will have to hope for an expanded experiment for this to be demonstrated (If there are more data from these experiments, please can this be included!).
In my opinion, this paper can be published with relatively minor corrections / clarifications. Most of the concerns regard full quantification of the uncertainty and lack of detail around steps in the experimental process.
- The experimental set up for the controlled release.
L120 onwards.
There is scant description surrounding the controlled release. There needs to be sufficient detail in the methodology so that this experiment could be repeated. My recommendation would be to have a SI with the description of the controlled release in detail, potentially including photos, set-up, details of the equipment used, gas compositions, uncertainties inherent in the set up (and how they may propagate uncertainties through to the measurement results). Understanding how controlled releases are set up is an important component of being able to understand how any type of flux measurements are verified.
- Characterization of the background
L146. From an operational perspective when using this system outside of a controlled release environment are there issues with assuming that the upwind pathlength is equal to the static measurement and therefore correlating the upwind line to that point measurement? I’m wondering if you could comment on the sensitivity of that pathlength to “unaccounted for methane” and how much methane could be present above baseline without the appearance of an apparent plume. This concerns me a little in settings where multiple plumes may be present and interfering with measurements or just unknown plumes from other sources.
- Influence of multiple plumes?
Is this method suitable for multiple plumes in a single field? I’m guessing that it might be, but it would be great to have that explicitly stated one way or another and what procedure would have to be followed to allow fluxes with overlapping plumes be separated (or joined) to give an overall flux.
- General details on equipment used.
Can all items of equipment used for measurement or quantification be defined throughout the paper, there are mass flow controllers, anemometers etc without provenance.
- Explanation of the rationale behind the controlled release set up and fluxes used.
Ideally the controlled release would have had a number of releases at different rates under different wind conditions. This really only gives us an understanding of the capability of the set up under this specific set of conditions. I would like to understand the reasoning for only testing under such limited scenario and would recommend the authors to consider taking part in something along the lines of Adam Brandt’s group tests of instruments at the next possible opportunity (e.g. https://doi.org/10.1525/elementa.2022.00080).
Minor
Fig 2. Can the averaging time be expanded so that the minimum in the Allen variance can be seen in the CH4 measurement precision.
L121. Requires explanation as to why this emission rate and location is representative of a real emission.
The conclusions feel rather light, it would be good to understand where the group feels the strength of this method is compared to the other technologies on the market.
Citation: https://doi.org/10.5194/egusphere-2023-691-RC2 - AC1: 'Reply to RC1 and RC2', Kevin Cossel, 15 Sep 2023
Peer review completion
Journal article(s) based on this preprint
Kevin C. Cossel et al.
Kevin C. Cossel et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 225 | 117 | 18 | 360 | 12 | 10 |
- HTML: 225
- PDF: 117
- XML: 18
- Total: 360
- BibTeX: 12
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2719 KB) - Metadata XML