the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
On the influence of underlying elevation data on Sentinel-5 Precursor satellite methane retrievals over Greenland
- 1Institute of Environmental Physics (IUP), University of Bremen FB1, Bremen, Germany
- 2Netherlands Institute for Space Research, SRON, Leiden, the Netherlands
- 1Institute of Environmental Physics (IUP), University of Bremen FB1, Bremen, Germany
- 2Netherlands Institute for Space Research, SRON, Leiden, the Netherlands
Abstract. The Sentinel-5 Precursor (S5P) mission was launched on October 2017 and has since provided data with high spatio-temporal resolution using its remote sensing instrument, the TROPOspheric Monitoring Instrument (TROPOMI). The latter is a nadir viewing passive grating imaging spectrometer. The mathematical inversion of the TROPOMI data yields retrievals of different trace gas and aerosol data products. The column-averaged dry air mole fraction of methane (XCH4) is the product of interest to this study. The daily global coverage of the atmospheric methane mole fraction data enables the analysis of the methane distribution and variation on large scales and also to estimate surface emissions. The spatio-temporal high-resolution satellite data are potentially particularly valuable in remote regions, such as the Arctic, where few ground stations and in-situ measurements are available. In addition to the operational Copernicus S5P total-column averaged dry air mole fraction methane data product developed by SRON, the scientific TROPOMI/WFMD algorithm data product v1.5 (WFMD product) was generated at the Institute of Environmental Physics at the University of Bremen. In this study we focus on the assessment of both S5P XCH4 data products over Greenland and find that spatial maps of both products show distinct features along the coast lines. Anomalies up to and exceeding 100 ppb are observed and stand out in comparison to the otherwise smooth changes in the methane distribution. These features are more pronounced for the operational product compared to the WFMD product. The spatial patterns correlate with the difference of the GMTED2010 digital elevation model (DEM) used in the retrievals to a more recent topography data set indicating that inaccuracies in the assumed surface elevation are the origin of the observed features. These correlations are stronger for the WFMD product. In order to evaluate the impact of the topography dataset on the retrieval we reprocess the WFMD product with updated elevation data. We find a significant reduction of the localized features when GMTED2010 is replaced by recent topography data over Greenland based on ICESat-2 data. This study shows the importance of the chosen topography data on retrieved dry air mole fractions. Use of a precise and up-to-date as possible DEM is advised for all S5P data products as well as for future missions which rely on DEM as input data. A modification based on this study is planned to be introduced to the next version of the WFMD data product.
Jonas Hachmeister et al.
Status: closed
-
RC1: 'Comment on amt-2022-102', Anonymous Referee #1, 16 May 2022
General comments
Hachmeister et al. present an analysis of TROPOMI methane column data to identify the causes of retrieval artifacts over Greenland. They investigate both the SRON operational methane retrieval and the University of Bremen WFMD retrieval. They find that the systematic errors over Greenland are largely explained by errors in the surface altitude datasets underpinning the two methane column retrieval algorithms, particularly for the WFMD retrieval. Applying a simple linear correction to the methane columns on the basis of surface height is sufficient to address the problem. The paper is a useful contribution and a good match for AMT. I recommend publication subject to minor revisions addressing the points below.
Specific comments
- L .52-54: Why give the old spatial resolution upfront and current resolution parenthetically? Consider reversing to give current specs first.
- L. 61: Recommend defining the GMTED2010 acronym here.
- L. 76: You use WFM-DOAS here but WFMD elsewhere.
- L. 93-96: Bias of a few percent could be significant for inference of regional emissions, no?
- L. 135: Recommend mentioning/defining the TROPOMI quality filter somewhere in Section 2.
- L. 141-142: What do you mean by “this effect merely shifts the reference point of the anomaly”. As written it’s not clear what the “effect” is. I think you mean that the selection of reference area defines the reference point for the anomaly, is that right?
- L. 147: Wouldn’t removing observations with quality flag > 0.1 remove most data? Should it not be < 0.1? (I’m not familiar with ICESat-2 data conventions.)
- L. 149: Are these weights inversely proportional to the errors? Using the errors as weights directly seems like it would more strongly weight higher-uncertainty data.
- L. 150: Consider adding an equation here to unambiguously describe the approach.
- L. 168-169: This phrasing seems to imply that the differences between ICESat-2 and GMTED2010 are due to ice sheet dynamics, but isn't 100-200 m too extreme for that to be the case? I don't know much about ice sheets, but from Section 2 the error seems mostly related to the radar altimetry.
- L. 182: Recommend using “r” or “rho” for the Pearson correlation coefficient, not “p”, because “p” can easily be mistaken for the p-value of the regression. This confused me on my initial review of the figures.
- Section 4.5: Is there a reason not to show maps of height-corrected WFMD v1.5 ΔXCH4 (or XCH4) and height-corrected operational ΔXCH4 (or XCH4)? Perhaps these could be added to Fig. 10 or Fig. 12, or made into a separate new Figure. I understand the paper already has quite a few figures and you show how the scatter plots improve from the linear height correction – but I was surprised not to see how the final methane maps improve post-correction.
- L. 210 & 222-229: Going back to my question about ice sheet changes over time – can you say more about what causes the GMTED2010 data to be outdated? It would be useful to know what fraction of the surface altitude errors are due to ice sheet dynamics vs. altimetry errors.
Technical corrections
- L. 14: “difference of GMTED2010… in the retrievals *compared* to a more recent…”
- L. 35: not clear what “and the surface extent” means here.
- L. 57: duplicate “a”.
- L. 92: should “led” be “lead” (present tense)?
- L. 155: *affect
- L. 161: Ice shield or ice sheet? Same question elsewhere in the manuscript where you use the term “ice shield”.
- L. 206: should be *were instead of “where”.
- L. 212: you say “region 2” here but elsewhere “region two” – should use consistent terminology.
- AC1: 'Reply on RC1', Jonas Hachmeister, 10 Jun 2022
-
RC2: 'Comment on amt-2022-102', Anonymous Referee #2, 19 May 2022
General comments
Hachmeister et al. present a topography correction for S5P TROPOMI methane retrievals, and show evidence of significant improvements of the product due to this correction. Their focus region is Greenland where the coasts have significant altitude variability in small (sub-TROPOMI-pixel level) spatial scales and also temporal variability due to the melting of the ice sheet. They propose to replace an obsolete elevation model GMTED2010 with elevation derived from ICESat-2 data. Methane anomalies over Greenland are shown to be correlated with both altitude and altitude difference (defined as old minus new elevation model), highlighting the sensitivity of the retrieval to surface pressure (and thus altitude). The spurious methane anomalies over Greenland coasts largely disappear when accounting for the updated elevation model.
This paper highlights the significance of updated input data for space-based retrievals of atmospheric composition, and is well-suited for the scope of AMT. The study is presented in a concise manner and the message it delivers is important for the atmospheric retrieval community. As for applied studies using satellite data (such as greenhouse gases source-sink analysis), it is important to reduce systematic errors in order to improve the reliability of the results derived using these data. I recommend publishing the paper in AMT after considering my minor comments and corrections below (in addition to those from Reviewer 1).
Specific comments
Title: Have you considered including "TROPOMI" in the title? I'm suggesting it for an improved visibility through search engines etc.
Introduction: This section is completely missing the motivation for the need to address the elevation (or surface pressure) sensitivity of the retrieval and thus an improved elevation model. Since this is the content of the paper, I propose to introduce the topic in the introduction. Other high-latitude retrieval challenges have been mentioned (dark surfaces); perhaps also mention the elevation sensitivity there (I would also recommend mentioning the solar zenith angle limitations at high latitudes), and then add a paragraph, perhaps after the 3rd paragraph in introduction, about what you are addressing in this paper, along with relevant background on GMTED2010 (complementing the request by Reviewer 1 here). Applicable text has already been written in several other parts of the manuscript.
Sect. 2.1.2 (and also 2.1.1 as applicable): I suggest to add information on the filtering (quality-screening) of the data, in particular because in e.g. Fig. 10 caption you refer to an updated quality filtering. You also most likely quality-screen the data before gridding so it is important to mention the qa_value criteria in 2.1.1 also.
Sect. 2.1.2: This is more of a question than a comment or suggestion: could steep elevation changes (especially at high latitudes where the SZA are large) also have an effect on the retrievals through casting shadows? Likely this is much less significant; I was just looking at Fig. 2 where one can see different XCH4 anomalies in the northern coast of Greenland compared to elsewhere in the coast.
Sect. 3.1: For the calculation of the 7-day methane anomaly, could you please specify how you do the gridding; is it only based on the centre coordinates of each pixel?
Sect. 4.5 and Conclusions: I assume that the "preliminary version of the updated WFMD product" is indeed a preliminary reprocessing of the WFMD retrieval (i.e. considers also the updated reference spectra corresponding to the updated elevation information) and not limited to postprocessing corrections based on the linear relationship shown in the paper. Could you please specify this part in the paper?
Conclusions: Is the updated DEM recommended also for the retrievals of other atmospheric gases? Please specify.
Technical corrections
Line 19: as up-to-date as possible
Line 21: introduced in
Line 36: spatio-temporal
Line 55: the Earth's
Line 55: SWIR wavelengths
Line 66 (also elsewhere): the word 'data' is plural so please change 'data has' --> 'data have' (and correspondingly also elsewhere in the text)
Lines 75-77: WFM-DOAS; please harmonise with the rest of the manuscript (either use WFM-DOAS or WFMD systematically)
Lines 95-96: remove the spaces between number and % symbol; 1 % --> 1%
Line 101: just --> only
Line 136: 7-days --> 7 days
Sect. 3.2 and Figs 6 & 7: Ambiguous use of p.
Line 155: don't effect --> do not affect
Line 195: is no longer --> are no longer
Line 206: where --> were
Appendix A is not referred to in the main text, please add.
Line 252: don't --> do not
Line 253: Fig 7 --> Fig. 7
Line 262: due melting --> due to the melting
Figs. 6 & 7: please include 'a' also for the top row figures
- AC2: 'Reply on RC2', Jonas Hachmeister, 10 Jun 2022
Status: closed
-
RC1: 'Comment on amt-2022-102', Anonymous Referee #1, 16 May 2022
General comments
Hachmeister et al. present an analysis of TROPOMI methane column data to identify the causes of retrieval artifacts over Greenland. They investigate both the SRON operational methane retrieval and the University of Bremen WFMD retrieval. They find that the systematic errors over Greenland are largely explained by errors in the surface altitude datasets underpinning the two methane column retrieval algorithms, particularly for the WFMD retrieval. Applying a simple linear correction to the methane columns on the basis of surface height is sufficient to address the problem. The paper is a useful contribution and a good match for AMT. I recommend publication subject to minor revisions addressing the points below.
Specific comments
- L .52-54: Why give the old spatial resolution upfront and current resolution parenthetically? Consider reversing to give current specs first.
- L. 61: Recommend defining the GMTED2010 acronym here.
- L. 76: You use WFM-DOAS here but WFMD elsewhere.
- L. 93-96: Bias of a few percent could be significant for inference of regional emissions, no?
- L. 135: Recommend mentioning/defining the TROPOMI quality filter somewhere in Section 2.
- L. 141-142: What do you mean by “this effect merely shifts the reference point of the anomaly”. As written it’s not clear what the “effect” is. I think you mean that the selection of reference area defines the reference point for the anomaly, is that right?
- L. 147: Wouldn’t removing observations with quality flag > 0.1 remove most data? Should it not be < 0.1? (I’m not familiar with ICESat-2 data conventions.)
- L. 149: Are these weights inversely proportional to the errors? Using the errors as weights directly seems like it would more strongly weight higher-uncertainty data.
- L. 150: Consider adding an equation here to unambiguously describe the approach.
- L. 168-169: This phrasing seems to imply that the differences between ICESat-2 and GMTED2010 are due to ice sheet dynamics, but isn't 100-200 m too extreme for that to be the case? I don't know much about ice sheets, but from Section 2 the error seems mostly related to the radar altimetry.
- L. 182: Recommend using “r” or “rho” for the Pearson correlation coefficient, not “p”, because “p” can easily be mistaken for the p-value of the regression. This confused me on my initial review of the figures.
- Section 4.5: Is there a reason not to show maps of height-corrected WFMD v1.5 ΔXCH4 (or XCH4) and height-corrected operational ΔXCH4 (or XCH4)? Perhaps these could be added to Fig. 10 or Fig. 12, or made into a separate new Figure. I understand the paper already has quite a few figures and you show how the scatter plots improve from the linear height correction – but I was surprised not to see how the final methane maps improve post-correction.
- L. 210 & 222-229: Going back to my question about ice sheet changes over time – can you say more about what causes the GMTED2010 data to be outdated? It would be useful to know what fraction of the surface altitude errors are due to ice sheet dynamics vs. altimetry errors.
Technical corrections
- L. 14: “difference of GMTED2010… in the retrievals *compared* to a more recent…”
- L. 35: not clear what “and the surface extent” means here.
- L. 57: duplicate “a”.
- L. 92: should “led” be “lead” (present tense)?
- L. 155: *affect
- L. 161: Ice shield or ice sheet? Same question elsewhere in the manuscript where you use the term “ice shield”.
- L. 206: should be *were instead of “where”.
- L. 212: you say “region 2” here but elsewhere “region two” – should use consistent terminology.
- AC1: 'Reply on RC1', Jonas Hachmeister, 10 Jun 2022
-
RC2: 'Comment on amt-2022-102', Anonymous Referee #2, 19 May 2022
General comments
Hachmeister et al. present a topography correction for S5P TROPOMI methane retrievals, and show evidence of significant improvements of the product due to this correction. Their focus region is Greenland where the coasts have significant altitude variability in small (sub-TROPOMI-pixel level) spatial scales and also temporal variability due to the melting of the ice sheet. They propose to replace an obsolete elevation model GMTED2010 with elevation derived from ICESat-2 data. Methane anomalies over Greenland are shown to be correlated with both altitude and altitude difference (defined as old minus new elevation model), highlighting the sensitivity of the retrieval to surface pressure (and thus altitude). The spurious methane anomalies over Greenland coasts largely disappear when accounting for the updated elevation model.
This paper highlights the significance of updated input data for space-based retrievals of atmospheric composition, and is well-suited for the scope of AMT. The study is presented in a concise manner and the message it delivers is important for the atmospheric retrieval community. As for applied studies using satellite data (such as greenhouse gases source-sink analysis), it is important to reduce systematic errors in order to improve the reliability of the results derived using these data. I recommend publishing the paper in AMT after considering my minor comments and corrections below (in addition to those from Reviewer 1).
Specific comments
Title: Have you considered including "TROPOMI" in the title? I'm suggesting it for an improved visibility through search engines etc.
Introduction: This section is completely missing the motivation for the need to address the elevation (or surface pressure) sensitivity of the retrieval and thus an improved elevation model. Since this is the content of the paper, I propose to introduce the topic in the introduction. Other high-latitude retrieval challenges have been mentioned (dark surfaces); perhaps also mention the elevation sensitivity there (I would also recommend mentioning the solar zenith angle limitations at high latitudes), and then add a paragraph, perhaps after the 3rd paragraph in introduction, about what you are addressing in this paper, along with relevant background on GMTED2010 (complementing the request by Reviewer 1 here). Applicable text has already been written in several other parts of the manuscript.
Sect. 2.1.2 (and also 2.1.1 as applicable): I suggest to add information on the filtering (quality-screening) of the data, in particular because in e.g. Fig. 10 caption you refer to an updated quality filtering. You also most likely quality-screen the data before gridding so it is important to mention the qa_value criteria in 2.1.1 also.
Sect. 2.1.2: This is more of a question than a comment or suggestion: could steep elevation changes (especially at high latitudes where the SZA are large) also have an effect on the retrievals through casting shadows? Likely this is much less significant; I was just looking at Fig. 2 where one can see different XCH4 anomalies in the northern coast of Greenland compared to elsewhere in the coast.
Sect. 3.1: For the calculation of the 7-day methane anomaly, could you please specify how you do the gridding; is it only based on the centre coordinates of each pixel?
Sect. 4.5 and Conclusions: I assume that the "preliminary version of the updated WFMD product" is indeed a preliminary reprocessing of the WFMD retrieval (i.e. considers also the updated reference spectra corresponding to the updated elevation information) and not limited to postprocessing corrections based on the linear relationship shown in the paper. Could you please specify this part in the paper?
Conclusions: Is the updated DEM recommended also for the retrievals of other atmospheric gases? Please specify.
Technical corrections
Line 19: as up-to-date as possible
Line 21: introduced in
Line 36: spatio-temporal
Line 55: the Earth's
Line 55: SWIR wavelengths
Line 66 (also elsewhere): the word 'data' is plural so please change 'data has' --> 'data have' (and correspondingly also elsewhere in the text)
Lines 75-77: WFM-DOAS; please harmonise with the rest of the manuscript (either use WFM-DOAS or WFMD systematically)
Lines 95-96: remove the spaces between number and % symbol; 1 % --> 1%
Line 101: just --> only
Line 136: 7-days --> 7 days
Sect. 3.2 and Figs 6 & 7: Ambiguous use of p.
Line 155: don't effect --> do not affect
Line 195: is no longer --> are no longer
Line 206: where --> were
Appendix A is not referred to in the main text, please add.
Line 252: don't --> do not
Line 253: Fig 7 --> Fig. 7
Line 262: due melting --> due to the melting
Figs. 6 & 7: please include 'a' also for the top row figures
- AC2: 'Reply on RC2', Jonas Hachmeister, 10 Jun 2022
Jonas Hachmeister et al.
Jonas Hachmeister et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 342 | 121 | 20 | 483 | 7 | 6 |
- HTML: 342
- PDF: 121
- XML: 20
- Total: 483
- BibTeX: 7
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1