Fiducial Reference Measurement for Greenhouse Gases (FRM4GHG)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper by Sha et al on “Fiducial Reference Measurement for Greenhouse Gases 2 (FRM4GHG)” summarises the results from a multi-year campaign of ground-based measurements of total column XCO2, XCH4, and XCO measurements using a suite of instrumentation measuring from the short to the long wave infra-red depending on technology.

Amongst the suit of instruments deployed over multiple years in Sodankyla are the high spectral resolution TCCON instruments (HR125), which are understood to serve as a reference together with a view co-located and simultaneous launches of Aircore in-situ ballon-based profile measurements deployed over each of the three years of the campaign, and with the potential to link the TCCON measurements as a tracable “reference” to the WMO standard.

Using TCCON measurements as a reference (and presenting the approach on how to deploy this standard over a network suing a “traveling standard”) the paper analyses the other suit of instrumentation deployed with respect to their performance over the cause of the campaign. The paper also discusses the details of establishing the traveling standard and the conceptual framework on how to establish the FRM standard. Given the extend and wide scope for the paper, I focus in this review on the campaign aspects and the associated results. I can recommend the parts on the traveling standard and the FRM status for publication without further qualifications.

The campaign, its analysis and the results presented here are of high importance to the current space-based suite of GHG satellite measurements, already relying for their product validation on many of those instruments (in particular TCCON, COCCON and NDACC). Moreover the results are of high relevance to future operational space-based GHG measurement systems like e.g. CO2M, which will have to rely on a large amount of well placed (and therefore ideally cheap, flexible, robust but accurate) ground-based FRMs, in order to allow for the continuous validation, verification and monitoring of operationally assimilated GHG observations in future GHG monitoring and verification support capacities, e.g. for the Copernicus Atmosphere Service and the WMO GHG watch (G3W).

The paper is overall very well written and the results are presented in a clear way. I can therefore recommend the paper for publication in Remote Sensing providing the authors can respond to the following general and specific comments in particular on the Campaign results.

General comments:

1)      The paper is the follow up to a previous paper by Sha et al, 2020, making use of at least part of the data of the three years campaign presented here. The lessons learnt and context of the results presented in the current paper with respect to the previous publication are not clearly stated. It seems to me that one important result from the previous work was the identification of the fact that there were problems identified with the reference TCCON measurements. The correction and improvement of the latter, and indeed the establishment of the latter as a true reference (via Aircore, see below) in the eyes of the reviewer would serve as an important underlying motivation (next to the extended time-series and other aspects) for the presented analysis. If this is so it would be important to make this clearer (both in the introduction but also in other relevant parts of the paper).

2)      The results presented here seem to question the approach using Aircore measurements to establish TCCON (or, in case, other measurements like the COCCON TS) as a potential reference (FRM) with a defined accuracy and error-margin on their accuracy, since the statistic is way too small (although for understandable, practical reasons) and deprives the mean bias performance of (in this case) TCCON w.r.t. Aircore of any meaning. Especially when considering that the variation of observed biases w.r.t. Aircore presented here is of the same order (or even pretty much the same) as the variation of biases between the other instruments and TCCON. While there is still much to be said to nevertheless (from an instrument design, performance and measurement principle point of view) consider the TCCON instrument as a “reference”, I don’t see how this can be established using the presented Aircore measurements. This qualification should not surprise anyone considering the large challenges controlled Aircore flights seem to when used for the purpose of calibrating or validating ground-based remote sensing measurements of GHGs.   

Specific comments:

Section 1 or the beginning of Section 2 should provide a brief overview of the scope and results of the first paper in order to bring the current paper in better perspective, in particular wrt the motivation for modifications to the (HR125 reference) instrumentation.

Section 2.2.1 Here the modifications to the TCCON instrument (correcting non-linearity) are introduced. Throughout the paper this data is then called TCCONmod. This leaves the reader with the question if TCCON measurements principally have to be corrected in this way, and (if yes) if this is done throughout the network in a consistent way.

Section 2. Retrieval software and approaches are referenced and discussed for most instruments but not explicitly of the TCCON instrument it seems, for which GGG is referenced but no outline on its principle approach and assumptions is provided (like e.g. for PROFAST).

Section3: Referring back to my general comment here, it would be helpful to briefly outline at the beginning of Section 3 in how far the results presented in the following are conceptually (i.e beyond the extended data-series) different or complementary or going beyond the results presented in Sha 2017. Is it predominantly because the identified "reference" being the TCCON HR125 has been corrected? Is the retrieval procedure/software different, or other aspects?

Section 3.1.1, l 500f: “A time window of 3 hour around the AirCore measurement time was used as the coincidence limit.” How do we know what this entails? The results seem to indicate its not a good choice. Have station/aircore footprint analysis been performed at the time or for the launch times?

Figure 2: Why have XCO2 measurements been omitted from Figure 2? It would be very important to add them from my point of view.

Fig2/Fig3: The number of Aircore measurements seem to be insufficient to conclude on the TCCON bias. Taking the scatter and distribution for XCH4 and XCH4 from Fig2 as a guidance, the accuracy for XCO2 can be anywhere between -0.5 and 0.5 ppm. So a stated mean of -0.051 ppm seems to me not to be a reliable number (here adding the XCO2 results to Fig 2 could help).

Section 3.2.1, l.544f. “It is understood and happens due to different instrument sensitivities (AKs) and the a priori difference from the true atmospheric state.” Are these differences in the a priori used by GGG and PROFAST? Or is this an effect on the side of the EM27 chain only. This is not clear. If this is described elsewhere at least a reference should be provided.

Figure 4 and general: The label TCCONmod needs better contextualisation (see before). Although briefly explained it poses the question to which the results presented here can be generalised to other measurements using TCCON data.

Section 3.5: It is not clear if the TS instrument (being an EM27) has been calibrated by the COCCON Karlsruhe reference or only by the Karlsruhe TCCON station and involving Aicore at the Karlsruhe side, or using HCL and Xair again or all of the three. 

Comments on the Quality of English Language

The word “container” is misspelled as ‘contained’ or similar at various points in the paper.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

It is  valuable for introducing the two ground-based network of greenhouse gases, four instruments are compared with TCCON measurement, and more detailed results are discussed, however some comments need pay attention. 

1.In L76-77, the keywords are more and redundant, for example, Fiducial Reference Measurements, the greenhouse gases, should aviod repeat with title words. Please delete it.

2.Please provide the XCO2 curve retrieved from AirCore and TCCON in figure 2.

3. In whole manuscript, It is better to use determination coefficient (R2) instead of correlation coefficient(R), and the R2 should maintain two decimals.

4. In L840, L1000 and L1016, as for the sentence of "Error! Reference source not found", please give the right literature.

5. The paper is too long and like a study report, it will be better to reorganize and compress it.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

1.        While the paper presents the experimental results in great detail, the discussion section could be further expanded to thoroughly analyze the significance and implications of the findings. Clarify how the results compare to previous studies and highlight the key novelties and limitations of the study.

2.        Provide a more in-depth introduction outlining the need for fiducial reference measurements for greenhouse gases, the limitations of existing approaches, and how the proposed study aims to overcome those limitations.

3.        Enhance the description of the instrumentation and methodologies used, particularly for those not widely known or documented. Ensure clarity in the adaptations made to achieve FRM status and how they contribute to the accuracy and precision of the measurements.

4.        Include a more comprehensive discussion of the uncertainties associated with the measurements and their propagation throughout the data analysis. Clearly explain the methods used to estimate these uncertainties and their potential impact on the final results.

5.        Consider revising the figures and tables to make them more readable and self-contained. Provide detailed captions explaining the data presented, including statistical measures and uncertainties where appropriate. Ensure the labels and legends are clear and consistent.For example, Figures 1 and 2. Provide more details on the statistical methods used to analyze the data, including error estimates and uncertainty budgets.Discuss the significance of the observed biases and correlations between instruments.

6.        Conduct a more comprehensive literature review to properly situate the study within the context of previous research. Cite the most relevant and recent studies to support your findings and motivate the significance of the work.

7.        Draw clear and concise conclusions from the findings, emphasizing the main outcomes and implications. Suggest future directions for research to build upon the current work and address any limitations.

8.        Check for consistency in formatting throughout the manuscript, including citations, section headings, and in-text references. Ensure that the writing is clear, concise, and logical, facilitating the reader's understanding of the work.

 

Comments on the Quality of English Language

Minor editing of English language required.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf