the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Jan 2024
Accounting for Aerosols Effect in GHGSat Methane Retrieval
Abstract. GHGSat comprises a constellation of high spatial and spectral resolution satellites, specializing in monitoring methane emissions at 1.65 μm. This study investigates the ability to accurately retrieve both the methane mixing ratio enhancement (ΔXCH4) and aerosol optical depth (XAOD) simultaneously from simulated GHGSat observations that incorporate angle-dependent scattering information. Results indicate that the polarity of ΔXCH4 when neglecting aerosols changes from negative to positive as surface albedo increases, which is consistent with previous studies. Biases in ΔXCH4 are most pronounced when XAOD is not simultaneously retrieved, ranging from −3.0 % to 6.3 % with a 0.1 AOD, a 60° solar zenith angle, and a 0.2 surface albedo for the nadir-only retrieval. Using multiple satellite viewing angles during the GHGSat observation sequence with a scattering angle ranging from 100° to 140°, the study shows that the mean bias and standard deviation of XCH4 are within 0.2 % and 2.7 % relative to the background. The correlation between simultaneously retrieved ΔXCH4 and XAOD shifts from positive to negative as surface albedo increases and aerosol asymmetry factor decreases, signifying a transition of the dominating aerosol effect from aerosol-only scattering to aerosol-surface multiple scattering. The variety of scattering angle ranges has little impact on the performance of the multi-angle viewing method. This study improves the understanding of the aerosol impact on the GHGSat ΔXCH4 retrieval and provides guidance for improving future GHGSat-like point-source imagers.
- Preprint
(1164 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2023-2958', Anonymous Referee #1, 20 Feb 2024
In this paper, numerical investigations are carried out on the effects of aerosol retrievals on methane remote sensing with GHGSat. The authors first used a radiative transfer model and the multi-angle viewing method to generate “GHGSat measurements”, then the aerosols and methane were retrieved using a retrieval method. If the retrieval method is just right, we can clearly see how the aerosols affect the methane retrieval as there are small uncertainties in the "measurements". This work provides valuable information for methane measurements with the new instrument, so this study certainly falls within the scope of AMT and the results could be of great importance to the scientific community.
However, as shown in my initial review, the authors should first clearly review and discuss the applicability of the interrogation method, as this would significantly affect the conclusion. The specific comments are listed below:
(1) The authors try to show that the retrieval of aerosols is important for methane retrieval. However, in the simulations, many parameters were fixed in the determination of aerosols. In real cases, these parameters are generally unknown (e.g. surface albedo). Therefore, the retrieval of aerosols in real measurements may lead to larger uncertainties in the methane remote sensing than shown in this manuscript.
(2) As my first review has shown, the authors should go into more detail on the applicability of the retrieval algorithm. Is the retrieval algorithm correct for the real cases? If it is not correct, the improvement in retrieval accuracy when adding aerosol retrieval does not necessarily mean that aerosol retrieval is important, and it may also be because the algorithm is wrong. Of course, the authors used their retrieval accuracy to show the applicability of the algorithm. However, their simulations and retrievals were simplified by many fixed parameters and assumptions. I suggest that the authors perform a sensitivity analysis for more complex parameters. If the authors can prove that the retrieval is accurate in differtent and more complex simulation configurations, the retrieval algorithm is credible.
(3) The numerical simulations in this paper are not able to represent the real retrievals because the real retrievals are more complex, so the conclusion of this paper is not necessarily correct. In particular, the aerosol retrieval may introduce other uncertainties, but these are eliminated by other fixed parameters.
To summarise, this work could provide valuable information for methane measurements with the new instrument if the authors can improve the applicability of the retrieval algorithm.Citation: https://doi.org/10.5194/egusphere-2023-2958-RC1 - AC2: 'Reply on RC1', Qiurun Yu, 29 Mar 2024
-
RC2: 'Comment on egusphere-2023-2958', Anonymous Referee #3, 23 Feb 2024
The manuscript investigates the retrieval of GHGSat-like methane retrievals in the presence of a simulated aerosol layer. Performance of the retrieval is investigated in four different configurations: Δ𝑋𝐶𝐻4 only in nadir and multi-angle viewing methods, and simultaneous Δ𝑋𝐶𝐻4 and 𝐴𝑂𝐷 in both viewing approaches. Bias and standard deviation in the retrieval are investigated under varying satellite viewing angle, aerosol SSA, asymmetry factor (g), and surface albedo. Some aspects of the work do feel simplistic, as described by Referee #1. However, the results are clearly described, and the interaction with aerosols is important to understand in the satellite methane community. I believe the work is suitable for publication, if the following comments are addressed:
Regarding the notation “𝑋𝐴𝑂𝐷” used in the text, this choice seems unusual. The numerical plots already simply use “AOD”, which I would encourage.
L9 ‘polarity of Δ𝑋𝐶𝐻4’ – suggest rephrasing as the sign of Δ𝑋𝐶𝐻4 biases
L104 “In Fig. 1b, strong CH4 absorptions are found around 1666 nm, affirming that the DISORT-simulated radiance is adequate for simulating the methane effect.” This conclusion is of importance but seems a little weak. Can it be strengthened with more evidence, or past validation?
L134 “solar and satellite zenith angles” is repeated twice. The second occurrence should be azimuth
Fig 2 It would be preferable to draw the phi2 angle to the projected point (possibly lying on the x-axis), not to the dashed backward viewing line. Often, these figures are labeled with North/South. Additionally, I am now realizing that this figure does not show the same set of angles as Table 1, which seems a little confusing, though not necessarily ‘wrong’ since it is a schematic diagram.
I would like to understand the difference between Fig 3 and Fig 1(c). The main two differences seem to be: the 0.3 FWHM smoothing and the added simulation of the 0.1 AOD aerosol layer. Looking at regions such as 1664-1665 nm, I do not think the coarser resolution explains the quantitative differences of values near ~0.2 compared to ~0.15. Is this then explained by the aerosol layer, or also other factors?
In the title of Figure 3, what is SU in AOD(SU)?
In Figure 4, can the first step (labeled DISORT), be a little descriptive, to distinguish it from DISORT in the forward model step?
L216-217 Note that Fig 6b (phase function for specific g) is not explicitly mentioned, and the link could be clearer to what is discussed (intensity of scattering energy)
L229 I do not necessarily understand the uncertainty simulation here. Is 0.2% the combined magnitude of the white noise and 1/f noise, or for each individually? Is uncertainty of the aerosol optical properties included somehow?
L356 ‘A bunch of’ – Suggest to quantitatively specific the satellite zenith angles considered
While satellite zenith angles were considered, what about solar zenith angle? I believe this was fixed to 60 degrees throughout and wonder how this influences the results.
Figure 8 and others, note there is a discrepancy between the figure (“Mean bias in XCH4”) and caption (Mean bias of ΔXCH4)
Citation: https://doi.org/10.5194/egusphere-2023-2958-RC2 - AC1: 'Reply on RC2', Qiurun Yu, 29 Mar 2024
Status: closed
-
RC1: 'Comment on egusphere-2023-2958', Anonymous Referee #1, 20 Feb 2024
In this paper, numerical investigations are carried out on the effects of aerosol retrievals on methane remote sensing with GHGSat. The authors first used a radiative transfer model and the multi-angle viewing method to generate “GHGSat measurements”, then the aerosols and methane were retrieved using a retrieval method. If the retrieval method is just right, we can clearly see how the aerosols affect the methane retrieval as there are small uncertainties in the "measurements". This work provides valuable information for methane measurements with the new instrument, so this study certainly falls within the scope of AMT and the results could be of great importance to the scientific community.
However, as shown in my initial review, the authors should first clearly review and discuss the applicability of the interrogation method, as this would significantly affect the conclusion. The specific comments are listed below:
(1) The authors try to show that the retrieval of aerosols is important for methane retrieval. However, in the simulations, many parameters were fixed in the determination of aerosols. In real cases, these parameters are generally unknown (e.g. surface albedo). Therefore, the retrieval of aerosols in real measurements may lead to larger uncertainties in the methane remote sensing than shown in this manuscript.
(2) As my first review has shown, the authors should go into more detail on the applicability of the retrieval algorithm. Is the retrieval algorithm correct for the real cases? If it is not correct, the improvement in retrieval accuracy when adding aerosol retrieval does not necessarily mean that aerosol retrieval is important, and it may also be because the algorithm is wrong. Of course, the authors used their retrieval accuracy to show the applicability of the algorithm. However, their simulations and retrievals were simplified by many fixed parameters and assumptions. I suggest that the authors perform a sensitivity analysis for more complex parameters. If the authors can prove that the retrieval is accurate in differtent and more complex simulation configurations, the retrieval algorithm is credible.
(3) The numerical simulations in this paper are not able to represent the real retrievals because the real retrievals are more complex, so the conclusion of this paper is not necessarily correct. In particular, the aerosol retrieval may introduce other uncertainties, but these are eliminated by other fixed parameters.
To summarise, this work could provide valuable information for methane measurements with the new instrument if the authors can improve the applicability of the retrieval algorithm.Citation: https://doi.org/10.5194/egusphere-2023-2958-RC1 - AC2: 'Reply on RC1', Qiurun Yu, 29 Mar 2024
-
RC2: 'Comment on egusphere-2023-2958', Anonymous Referee #3, 23 Feb 2024
The manuscript investigates the retrieval of GHGSat-like methane retrievals in the presence of a simulated aerosol layer. Performance of the retrieval is investigated in four different configurations: Δ𝑋𝐶𝐻4 only in nadir and multi-angle viewing methods, and simultaneous Δ𝑋𝐶𝐻4 and 𝐴𝑂𝐷 in both viewing approaches. Bias and standard deviation in the retrieval are investigated under varying satellite viewing angle, aerosol SSA, asymmetry factor (g), and surface albedo. Some aspects of the work do feel simplistic, as described by Referee #1. However, the results are clearly described, and the interaction with aerosols is important to understand in the satellite methane community. I believe the work is suitable for publication, if the following comments are addressed:
Regarding the notation “𝑋𝐴𝑂𝐷” used in the text, this choice seems unusual. The numerical plots already simply use “AOD”, which I would encourage.
L9 ‘polarity of Δ𝑋𝐶𝐻4’ – suggest rephrasing as the sign of Δ𝑋𝐶𝐻4 biases
L104 “In Fig. 1b, strong CH4 absorptions are found around 1666 nm, affirming that the DISORT-simulated radiance is adequate for simulating the methane effect.” This conclusion is of importance but seems a little weak. Can it be strengthened with more evidence, or past validation?
L134 “solar and satellite zenith angles” is repeated twice. The second occurrence should be azimuth
Fig 2 It would be preferable to draw the phi2 angle to the projected point (possibly lying on the x-axis), not to the dashed backward viewing line. Often, these figures are labeled with North/South. Additionally, I am now realizing that this figure does not show the same set of angles as Table 1, which seems a little confusing, though not necessarily ‘wrong’ since it is a schematic diagram.
I would like to understand the difference between Fig 3 and Fig 1(c). The main two differences seem to be: the 0.3 FWHM smoothing and the added simulation of the 0.1 AOD aerosol layer. Looking at regions such as 1664-1665 nm, I do not think the coarser resolution explains the quantitative differences of values near ~0.2 compared to ~0.15. Is this then explained by the aerosol layer, or also other factors?
In the title of Figure 3, what is SU in AOD(SU)?
In Figure 4, can the first step (labeled DISORT), be a little descriptive, to distinguish it from DISORT in the forward model step?
L216-217 Note that Fig 6b (phase function for specific g) is not explicitly mentioned, and the link could be clearer to what is discussed (intensity of scattering energy)
L229 I do not necessarily understand the uncertainty simulation here. Is 0.2% the combined magnitude of the white noise and 1/f noise, or for each individually? Is uncertainty of the aerosol optical properties included somehow?
L356 ‘A bunch of’ – Suggest to quantitatively specific the satellite zenith angles considered
While satellite zenith angles were considered, what about solar zenith angle? I believe this was fixed to 60 degrees throughout and wonder how this influences the results.
Figure 8 and others, note there is a discrepancy between the figure (“Mean bias in XCH4”) and caption (Mean bias of ΔXCH4)
Citation: https://doi.org/10.5194/egusphere-2023-2958-RC2 - AC1: 'Reply on RC2', Qiurun Yu, 29 Mar 2024
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 278 | 72 | 17 | 367 | 9 | 11 |
- HTML: 278
- PDF: 72
- XML: 17
- Total: 367
- BibTeX: 9
- EndNote: 11
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
