A study of synthetic <sup>13</sup>CH<sub>4</sub> retrievals from TROPOMI and Sentinel-5/UVNS

Malina, Edward; Hu, Haili; Landgraf, Jochen; Veihelmann, Ben

doi:https://doi.org/10.5194/amt-12-6273-2019

Articles | Volume 12, issue 12

https://doi.org/10.5194/amt-12-6273-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Special issue:

TROPOMI on Sentinel-5 Precursor: first year in operation (AMT/ACP...

https://doi.org/10.5194/amt-12-6273-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 12, issue 12

Research article

|

29 Nov 2019

Research article |

| 29 Nov 2019

A study of synthetic ¹³CH₄ retrievals from TROPOMI and Sentinel-5/UVNS

Edward Malina, Haili Hu, Jochen Landgraf, and Ben Veihelmann

Download

Final revised paper (published on 29 Nov 2019)
Preprint (discussion started on 21 Jan 2019)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

SC1: 'Precision requirement is not adequate for scientific interpretation', Thomas Röckmann, 25 Jan 2019
- AC1: 'Response to Professor Roeckmann', Edward Malina, 12 Mar 2019
  - AC4: 'Further response', Edward Malina, 16 Sep 2019
RC1: 'Review comments on “A study of synthetic 13CH4 retrievals from TROPOMI and Sentinel 5/UVNS Part 1: non scattering atmosphere” by Malina et al', Anonymous Referee #1, 18 Mar 2019
- AC2: 'Response to Review comment 1.', Edward Malina, 31 May 2019
RC2: 'A study of synthetic 13CH4 retrievals from TROPOMI and Sentinel 5/UVNS Part 1: non scattering atmosphere', Anonymous Referee #2, 29 Mar 2019
- AC3: 'Response to Review comment 2.', Edward Malina, 31 May 2019

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Edward Malina on behalf of the Authors (31 May 2019) Author's response Manuscript

ED: Referee Nomination & Report Request started (12 Jun 2019) by Andreas Hofzumahaus

RR by Anonymous Referee #1 (28 Jun 2019)

Suggestions for revision or reasons for rejection

I would like to suggest to add some more descriptions and discussions regarding the comparison between non-scattering atmosphere and scattering atmosphere.

1. Section 6
What are blank land areas in Figs. 16, 17 and 18? Are there no retrievals available? Or are there no difference between non-scattering and scattering atmosphere? In the cases of SWIR1 and SWIR3, the authors say that notable exceptions (pronounced spike) in systematic error over the Hudson Bay in Canada. Besides the Hudson Bay, similar or larger differences appear on other places, such as over tropical Pacific, and the Gulf in the Middle East. In the case of SWIR1+SWIE3, the authors say “some of the localized systematic bias (e.g., the Hudson Bay) are no longer apparent”. However, one positive spike (>30 permill) appears over the Hudson Bay, while negative spots appear in the first two cases there.

Why the global mean biases (differences between non-scattered and scattered) appear to be negative for both SWIR1 and SWIR3? Is the global mean bias negative also in SWIR1+SWIR3?

2. Discussion
The authors state in the discussion (P35, L9-10) that “given that the bias is largely globally consistent, it will not be difficult to apply a correction factor and remove this (systematic) bias’. In reality, it would be difficult to remove systematic bias because “systematic errors are difficult to quantify (Houweling et al., 2014) (P35, L9).”
Figure 15 shows that the errors in 13CH4 along with aerosol optical depth (AOD) are in Gaussian distribution (no correlation). In reality, there is uncertainty also in AOD. Moreover, other complications come into the retrievals. That is why the systematic error is difficult to be removed. “given that the bias is largely globally consistent” seems too ideal. It would be helpful to mention possible challenges with the real atmosphere.

Hide

RR by John Worden (19 Aug 2019)

Suggestions for revision or reasons for rejection

For the purpose of a hopefully more useful review I am identifying myself (John Worden)

The paper is generally well written and I understood all the major points and conclusions and consequently I do not have many suggested changes. That said, I do suggest that they try a similar approach to how we estimate the HDO/H2O ratio from TIR radiances (e.g. Worden et al. 2006, Schneider et al., 2012, Lacour et al., 2012) in order to potentially reduce the uncertainties of their delta CH4 retrieval. If such an approach is not feasible in the NIR (although it looks like it should be) then it would be useful to discuss why so that the rationale is documented. Alternatively, add language in the discussion that the TIR based HDO/H2O retrieval approach could be another way for reducing uncertainties in deltaCH4 retrievals.

Just to be sure, from what I understood of their retrieval, the study authors are retrieving the 13CH4 column independently of 12CH4 (although the two are retrieved jointly) such that the constraint for 13CH4 is independent of the 12CH4 retrieval. However, we actually do not scientifically “care” about the 13CH4 column but instead want to know about the ratio of 13CH4 to 12CH4. Our approach for the TIR based HDO/H2O estimate is to therefore jointly estimate the HDO and H2O amounts (similar to the above) but constraining the ratio via cross terms in the constraint matrix. Without this approach the uncertainties in the HDO/H2O ratio were much larger (I think about a factor of 10) than what we report which is about 15-20 per mille (delta-d varies by about 300 per mille so this is an acceptable error). Adding this cross-term is fairly straightforward although I am not yet sure how it fits with the Tikhonov constraint used here but might be as simple as adding a covariance term reflecting prior information about deltaCH4 (e.g. the range is between ~-70 per mille to -20 per mille), see Equation 21 in Worden et al. 2006 “Tropospheric Emission Spectrometer observations of the tropospheric HDO/H2O ratio: Estimation approach and characterization”. Without this additional constraint, the errors jump up to ~150 per mille and make the data essentially unusable.

Other Comments

The errors is on the order of 1ppb… is that about a 5% uncertainty? (e.g. abundance is ~0.01*2000ppb?). Some language here would be nice as I was initially confused that you could obtain 1 ppb uncertainties when the uncertainty of a total column 12CH4 retrieval is ~10-15 ppb.

The Jacobians are given as delta Radiance / delta CH4, this was confusing to me at first as delta is also the notation used for the isotopologue, perhaps change to a partial derivative notation instead?

Hide

ED: Publish subject to minor revisions (review by editor) (03 Sep 2019) by Andreas Hofzumahaus

ED: Publish subject to minor revisions (review by editor) (10 Sep 2019) by Andreas Hofzumahaus

AR by Edward Malina on behalf of the Authors (16 Sep 2019) Author's response Manuscript

ED: Publish subject to technical corrections (27 Sep 2019) by Andreas Hofzumahaus

AR by Edward Malina on behalf of the Authors (24 Oct 2019) Author's response Manuscript

Short summary

We present a feasibility study on retrieving ¹²CH₄ and ¹³CH₄ using the recently launched TROPOMI on the Copernicus Sentinel 5P satellite and the future UVNS instrument on Sentinel 5. The ratio of ¹²CH₄ and ¹³CH₄ can be used to calculate the δ¹³C value, which has been shown to be able to distinguish between biological and non-biological sources of methane. We show that Sentinel 5/UVNS may be used to distinguish between methane source types, while Sentinel 5P/TROPOMI is subject to large biases.

A study of synthetic 13CH4 retrievals from TROPOMI and Sentinel-5/UVNS

Download

Interactive discussion

Peer-review completion

A study of synthetic ¹³CH₄ retrievals from TROPOMI and Sentinel-5/UVNS