the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Jun 2023
Global retrieval of stratospheric and tropospheric BrO columns from OMPS-NM onboard the Suomi-NPP satellite
Abstract. Quantifying the global bromine monoxide (BrO) budget is essential to understand ozone chemistry better. In particular, the tropospheric BrO budget has not been well characterized. Here, we retrieve nearly a decade (February 2012–July 2021) of stratospheric and tropospheric BrO vertical columns from the Ozone Mapping and Profiling Suite Nadir Mapper (OMPS-NM) onboard the Suomi National Polar-orbiting Partnership (Suomi-NPP) satellite. To address the mismatch between a priori profiles and column retrievals in the stratosphere-troposphere separation, for each OMPS-NM pixel, we save two types of BrO vertical profiles and use the appropriate one based on whether a tropospheric enhancement is detected. Total ozone columns observed from OMPS-NM are used to identify tropospheric BrO enhancements. We demonstrate good agreement for both the stratosphere (r = 0.81–0.83) and the troposphere (r = 0.50–0.69) by comparing monthly mean BrO vertical columns from OMPS-NM with ground-based observations from three stations (Lauder, Utqiag ̇vik, and Harestua). The OMPS-NM BrO retrievals successfully capture tropospheric enhancements not only in the polar but also in the extrapolar regions (the Rann of Kutch and the Great Salt Lake). We also estimate random uncertainties in the retrievals pixel by pixel, which can assist in quantitative applications of the OMPS-NM BrO dataset. Our BrO retrieval algorithm is designed for cross-sensor applications and can be adapted to other space-borne ultraviolet spectrometers, contributing to the creation of continuous long-term satellite BrO observation records.
-
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
(9927 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(9927 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1163', Anonymous Referee #1, 08 Jul 2023
Review of “Global retrieval of stratospheric and tropospheric BrO columns from OMPS-NM onboard the Suomi-NPP satellite”
In this manuscript, the authors describe a new BrO data product using OMPS-NM measurements. The product includes stratospheric and tropospheric columns, uses a complex stratosphere – troposphere separation algorithm, applies pixel specific airmass factors and provides detailed uncertainty estimates. The topic of the manuscript fits well into the AMT scope, the product described is of interest to the atmospheric chemistry community and the paper is well written and includes detailed descriptions of the algorithms used. I therefore recommend it for publication in AMT.
I have however several questions, comments and suggestions to the manuscript and the algorithm, which the authors should address before the manuscript is accepted:
1) Reference sector correction
What the authors call reference sector correction in fact includes two corrections: a) the addition of the modelled BrO offset necessary when using a radiance background spectrum and b) a latitudinal correction of the slant columns based on the model. The latter correction is critical as the way I understand it, it forces the baseline of the measurements on the model values. Therefore, I think the stratospheric BrO product is lo a large extent just reproducing the model values. This is in contrast to the statements in the manuscript claiming that both stratospheric and tropospheric column are retrieved from the measurements.
Please a) discuss this point and b) include an example of the correction for one full orbit, for example one of those shown in Fig. 3.
2) CAM-chem climatology
The algorithm described heavily relies on the CAM-chem climatology for the stratospheric columns (see above), the separation between stratospheric and tropospheric signals and for the airmass factors. However, a) it is not clear how well the climatology represents the real atmospheric BrO field and b) tropospheric BrO enhancements are very dynamic events, and their magnitude and location cannot be reflected by a static monthly climatology. This has important implications for the airmass factors which probably are often not correct as the monthly mean profiles are neither a good representation of BrO events, nor of background conditions. The algorithm foresees the use of “flattened” profiles in case the measurements with low BrO columns, but the opposite case (high BrO in a region where the climatology does not expect a BrO event) is not treated separately.
Please a) discuss the impacts of using a climatology as input for the airmass factors and b) include a figure comparing the modelled climatological tropospheric columns in comparison to the measurements, for example for the orbits shown in Fig. 3.
3) Polar vortex
The authors discuss the well known problem of using O3 columns as proxy for the stratospheric BrO columns and state, that their method “still preserves the overall spatial pattern of the stratospheric field”. I have not understood why that should be the case. If we have ozone depletion (and possibly stratospheric BrO enhancement) within the vortex, the relationship between O3 and BrO will be different for vortex and non-vortex air masses, leading to large scatter in the O3 – BrO plot. If all the in vortex values are removed and filled with surrounding (out of vortex) values, the stratospheric BrO column will not be correct.
Please explain why your method is less affected by this problem than that of previous studies.
4) Uncertainties
Although a detailed discussion of uncertainties is given, I’m somewhat confused by what to expect from the data product. Are the uncertainties given for individual pixels? Has each pixel in the product an uncertainty value? Will uncertainties be smaller in monthly averages? If so, why are the DSCD uncertainties shown in Figs. 12 and 13 for monthly averages comparable to the median uncertainty quoted for an individual pixel? How is the uncertainty of having a high surface BrO event in the data at a location where the CAM-chem climatology has background conditions taken into account? Can the product be used on a daily basis (Figs. 3 and 4 suggest that this is the case) or should it better be used on a monthly basis?
Please add error bars to the satellite data in Fig. 7 and a paragraph on data usage.
5) High values over the ocean
In Fig. 11, large BrO columns are shown over much of the NH oceans, and as far as I can see, the largest BrO columns in that month are not found in the Arctic but somewhere over the Pacific. Do you think this is realistic? How do these findings compare to other satellite products and independent measurements? Are these high columns already visible in the slant columns or are the introduced by the CAMS-chem based airmass factors?
6) Comparison to other satellite data
Why is there no comparison to other satellite data? The data set is advertised as extension of the OMI afternoon BrO time series, and I think it would be good to include some kind of comparison between BrO observations from the two platforms, even if it is just a visual side-by-side comparison of a monthly average.
7) Profile flattening
This procedure seems arbitrary to me. I do not see why such a profile should give reasonable tropospheric BrO columns or realistic airmass factors. Please justify your approach.
8) Coastal artefacts
In Fig. 10, there are many localised spots of suspiciously high tropospheric BrO along the Antarctic coast but also at the sea ice edge. As this is a longterm average, it is clear that the high values are from too small airmass factors, and this is confirmed by figure (e). In my opinion, this should not be part of the product as it clearly is an artefact. I expect similar artefacts in the daily images also in the NH, for example close to Spitsbergen. My suggestion is to at least include a flag for those retrievals having very small airmass factors, and to increase the uncertainty estimates for such pixels.
9) Large AMF above ocean
I’m surprised by the large values of the airmass factor over much of the oceans. Values above two indicate the presence of a significant fraction of the BrO in the free troposphere. Do you think this is realistic?
10) Data availability
I understand that the product is not yet released, but to my knowledge, the data has to be available in some form for the manuscript to be published in AMT. Maybe just add that it is available on request?
Citation: https://doi.org/10.5194/egusphere-2023-1163-RC1 - AC1: 'Reply on RC1', Heesung Chong, 17 Nov 2023
-
RC2: 'Comment on egusphere-2023-1163', Anonymous Referee #2, 08 Aug 2023
In the paper titled "Global Retrieval of Stratospheric and Tropospheric BrO Columns from OMPS-NM on the Suomi-NPP Satellite," Chong and colleagues introduce a new BrO column product derived from the OMPS-NM satellite instrument. The authors focus on presenting a long-term time series of tropospheric columns while effectively distinguishing between stratospheric and tropospheric contributions. The paper's strength lies in its innovative approach, combining the strengths of two state-of-the-art methods to separate stratospheric and tropospheric columns. Furthermore, the long operational lifespan of the OMPS-NM instruments lends a distinct advantage to this product, promising continuous data acquisition over an extended period which allows long-term time series of this chemically important tracer. This BrO column product makes a valuable addition to existing satellite-derived BrO products, with potential benefits for bromine chemistry research and is therefore in good alignment with the scope of AMT. I strongly recommend considering this paper for publication after addressing the noted corrections and points.
General comments:
GC1) The manuscript appears to primarily target polar applications, with retrieval considerations and examples predominantly centered around polar regions. For instance, the effective application of the flattening technique to polar hot-spots raises questions about its performance in non-polar tropospheric enhancements, confer also GC2. Moreover, the wavelength criterion selection (Page 39, line 6-7) implies a concentration on polar applications, aiming to achieve optimal results under high latitude and SZA conditions. While this focus is justified, I suggest to explicitly mention in the abstract, introduction, and potentially the title. Consider either highlighting the emphasis on polar BrO retrieval or substantiating why polar regions pose the greatest challenge.
GC2) Even though it is maybe a bit overcomplicated, I generally like the authors approach for the separation of tropospheric and stratospheric columns as it combines the two pathways taken in previous studies. However, there are several steps where I can see potential issues:
- How does step (i), flattening the tropospheric model profile (page 17, Figure 1), perform with non-polar tropospheric enhancements? Is it specifically tailored for this scenario? For instance, if an extensive area of BrO tropospheric enhancements emerges in the equatorial Pacific due to an unusual climate change event or a significant volcanic plume, how would step (i) handle this situation? It appears that step (i) might overlook such occurrences, possibly leading to their inclusion in V_strat_0 (which may not be problematic by itself). However, a concern arises regarding the effectiveness of the O3-BrO relation-based separation (Figure 4c) under such circumstances.
- Page 19: In step (v) the authors suggest to use the model profile as input for AMF calculation under no hotspot conditions and in difference the “flattened” model profile as input for the AMF calculation under hotspot conditions. The validity of this approach does not occur to me as I have the following concerns/points.
- How confident are you, that this model profile is representative for the tropospheric profile under “no-hotspot” conditions? Please shortly address this in the manuscript.
- For hotspot cases, the authors opt for the flattened tropospheric profile. While not explicitly stated, it seems this choice stems from the model profile being decidedly unsuitable for such scenarios. However, employing a flattened profile could potentially worsen the situation. This selection could yield a lower and possibly excessively low tropospheric AMF (A_trop_select, denominator in equation 12). Considering the high S_trop (numerator in equation 12), this might lead to an excessively high V_trop. This concern is particularly relevant when contrasting with non-hotspot pixels where a non-flattened profile is used for tropospheric AMF calculation. Consequently, I have three queries/suggestions regarding the treatment of hotspot cases:
- Is there a justification for favoring a flattened profile over the model profile as an assumption?
- Given that AMF carries a larger share of random uncertainties, could it be more accurate and practical to assume a distinct "hot-spot" profile shape (e.g., employing a ground-level profile shape as in Fig. 3e for polar hot spots, and a different assumption for tropical cases)?
- Is this factor considered in the error propagation of the AMF? Given its potential significance, it would be beneficial to explicitly acknowledge this and its impact.
- Page 19, line 13: Like Sihler et al. (2012), step (iii) employs the assumption of a constant O3/BrO relation to quantify tropospheric enhancements. However, the authors note that this assumption becomes problematic in polar vortex scenarios, potentially introducing bias to your data. Is this data utilized in the final product? If yes, provide rationale for its inclusion and discuss implications for the data quality flag.
GC3) The paper would benefit from a comparison with other satellite studies. This does not need to be thorough, but differences and agreements should be adressed both with regard to the polar observations as well as the BrO from Rann of Kutch.
Specific Comments
Abstract:
The abstract could benefit from a clearer articulation of the paper’s primary goal and focus. Consider adding a succinct sentence at the beginning of line 4 to outline the central theme of the study. This could lead into the subsequent statement, "To address this concern and improve upon the current methods, our study introduces..." This adjustment would help provide a smoother transition into the specific achievements and advancements discussed.
Introduction:
Page 3, lines 10-29: The provided overview of the broad variety of separation schemes used in the literature is commendably thorough and informative. Nonetheless, its level of detail seems too thorough for an introduction. In the introduction, the focus should be on the paper's new method, and a concise acknowledgment of various approaches would suffice to put the paper’s method into perspective
Given the absence of a designated "methods" section to accommodate such content as a subsection, I understand the authors' predicament in determining its placement. To address this, a practical solution could be integrating it as a subsubsubsection within Page 14, subsubsection 2.2.4.
Page 3 line 35-page 4 line 5: The paragraph's primary emphasis, as underscored by the authors first lines, lies in the substantial potential of OMPS to provide an extensive and enduring time-series well into the 2030s. This aspect should take precedence, and should be highlighted in the paragraphs first sentence, shifting the focus imediately to this critical attribute. Accordingly, I suggest starting with the assertion about OMPS's long time-series capability, and subsequently incorporating the initial sentence, "OMPS-NM instruments ... decommissioning of TROPOMI (Nowlan et al., 2023)," at the paragraph's conclusion.
Page 4, line 16-18 and fig. 1: There are two confusing aspects:
- The use of the numeration (1) – (4) in reference to fig. 1, lets the reader look for the numbers 1-4 in fig. 1. However, the corresponding fields are noted with (A)-(D). Please use the same symbols in text and figure
- When reading “highlighted in blue in fig. 1”, the first look in figure 1 will be to the fields which have a blue background “OMPS-NM L1B product” etc. As Figure 1 contains a lot of information it is difficult to find the highlighted fields. Either specify in the text that they are “encircled/framed in blue” or reconsider the coloring within fig. 1 to avoid this confusion. Also consider to increase the size of the border line to highlight the 4 fields. Furthermore, consider to place the A-D always at the same location w.r.t. the fields they refer to (either all on the top left or top right).
Page 4, lines 16-18 and Fig. 1: This section presents two points of confusion:
- The utilization of the numerals (1) – (4) in reference to Fig. 1 prompts readers to search for numbers 1-4 within the figure. However, the corresponding elements are actually labeled as (A)-(D). To enhance clarity, it's advisable to employ consistent symbols both in the text and the figure.
- When the text mentions "highlighted in blue in Fig. 1," readers instinctively turn their attention to fields with a blue background, such as "OMPS-NM L1B product," within Figure 1. Since the figure contains extensive information, locating the highlighted elements becomes challenging. To address this, you could specify in the text that the relevant fields are "encircled/framed in blue." Alternatively, reconsider the color scheme within Fig. 1 to alleviate this confusion. Additionally, consider enhancing the border line's size to accentuate the four designated fields. Furthermore, for consistency, contemplate consistently placing the labels (A)-(D) at the same relative position with respect to the corresponding fields (either all at the top left or top right).
By harmonizing symbols and refining visual cues, these adjustments can substantially improve the reader's comprehension.
Page 4 and 6: The mention of 2 times “retrieval” in headline 2 and 2.2 is redundant, I suggest to move all the subsubsections in 2.2 up by one rank in hierarchy (e.g. 2.2.1-> 2.1, etc.).
Page 6, line 6: The phrasing currently implies that Beirle et al., 2017; Nowlan et al., 2023 originated the super-gauss concept. I recommend revising it to "super Gaussian and adopt the approach outlined in Beirle et al., 2017; Nowlan et al., 2023."
Page 6, line 23: The authors specify their utilization of a 20° latitude portion within a single orbit for the computation of the earthshine reference spectrum. With an assumed along-track pixel footprint of 50 km, this approach implies that each across-track reference spectrum would be derived from 40-50 individual spectra. Notably, other investigations involving 2D CCD satellites like OMI and TROPOMI adopt larger sectors (e.g., Seo et al., 2018: 150°E – 240°W, 30°S-30°N for BrO; Theys et al., 2017: 120-160°W, 10°N-10°S for SO2). Please Justify the rationale behind employing a relatively compact reference sector and argue why such a low statistic is deemed satisfactory for your study.
Page 8, Table 1: Regarding the SCD retrieval: All (to the reviewers knowledge) recent other BrO DOAS and “DOAS-like” spectral retrievals include OClO in their spectral retrieval (e.g. Suleiman et al., 2019; Herrmanns et al., 2022; They et al., 2011; Sihler et al., 2012; Seo et al., 2019). Please justify your choice not to include it especially with respect to the potential spectral interferences (see overlapping absorption peak at 344nm).
A similar argument can be made for SO2 although it was only implemented in the most recent publications (Suleiman et al., 2019 (Proposed to be implemented); Herrmanns et al., 2022; Sihler et al., 2012; Seo et al., 2019). Please explain why you have not chosed to include it and how strong you estimate for spectra affected by SO2 (e.g. strong pollution emitter or volcanoes) as well as how substantial this impact is on the global data-set.
Page 8, Table 1: I suggest to highlight the trace gas absorption spectra in the list. For example by horizontal lines. Also add “the parameter are listed in their order of appearance in eq. 2”.
Page 9, Figure 2: It would be beneficial to also include the residual spectrum in this plot as it gives information on potential residual structures originating from absorbers which are not accounted for.
Page 13: In other studies (such as Seo et al. 2019) a uniform background of 3.5x1013 moleculesc cm-1 is used (based on Richter et al., 2002). Include how your background correction S_R typically is with respect to this value.
Richter et al., 2002: Richter, A., Wittrock, F., Ladstatter-Weissenmayer, A., and Burrows, J. P.: GOME measurements of stratospheric and tropospheric BrO, in: Remote Sensing of Trace Constituents in the Lower Stratosphere, Troposphere and the Earth’s Surface: Global Observations, Air Pollution and the Atmospheric Correction, edited by: Burrows, J. P. and Takeucki, N., Adv. Space Res., 11, 1667–1672, 2002.
Page 13 line 17-31: It would improve the readability to include the names “S_R” and “S_B”, when talking about these quantities in the text.
Page 37 line 20: Please remove "including volcanic plumes". Volcanic application is not mentioned at all in the result section and as the major volcanic constituent "SO2" is not accounted for in the spectral fitting, this statement is questionable.
Page 38, line 13-18, concerning the “modeled stratospheric BrO DeltaSCD”: I assume the “modeled Delta SCDs” in line 17 is the same as the “modeled stratospheric BrO Delta SCD”. Name both the same, and consider to mention this in the first sentence of its explanation (line 13).
Page 38, Line 15-17: From your explanation, it looks like the “modeled stratospheric BrO Delta SCD”, which is subtracted by is defined in a way that its mean is zero at 0-10°N and non-zero elsewhere. Thus the “Delta SCD bias” will then be the complete retrieved SCD subtracted by zero at the equator and non-zero
Page 38-39 and figure A1, regarding the correlation with O3:
- How have you combined the different O3 SCDs to one O3 SCD? Did you follow the formula proposed by Pukite and Wagner (2016) eq. 16? If so, please add a reference.
- I do not see the benefit of looking at two O3 absorptions at 243 and 273K, as there is only O3 absorption. If you do not gain any benefit from using the two, then I suggest to skip this.
- Should you chose to keep the distinction between 243 and 273K, how did you avoid a cross correlation between the two spectra (which are very similar)? Have you orthogonalized the O3 absorption spectrum at 273K w.r.t the one at 243K? Please state this in the text.
Page 39 line 8: The choice of a percentile seems suboptimal to me and introduced strong data-selection biases and I would urge to change this. For instance, the pixel at high VZA will have a higher SCD compared to the nadir looking pixel and they will be selected thus more frequently. Additionally, tropospheric enhancements will be more dominant. If you want to select for high latitude and high SZA, then why not use latitude and SZA as a selection criterium?
Additionally, the paper is about a global product of BrO. Please justify why you have not also looked if the fit is also performing well at other regions (cf. Seo et al., 2019, who performed a retrieval interval mapping for several cases and also for an equatorial region).
There is one paper about long-term BrO time-series:
Bougoudis, I., Blechschmidt, A.-M., Richter, A., Seo, S., Burrows, J. P., Theys, N., and Rinke, A.: Long-term time series of Arctic tropospheric BrO derived from UV–VIS satellite remote sensing and its relation to first-year sea ice, Atmos. Chem. Phys., 20, 11869–11892, https://doi.org/10.5194/acp-20-11869-2020, 2020.
I suggest to include this paper in your introduction. You can frame this as an advantage of the OMPS data-set who in difference to Bougoudis et al., 2020 does not require a complicated inter-calibration of the time-series of the different sensors.
Technical comments:
Page 2, Line 30: As the authors are very thorough in giving a complete list of relevant references in the introduction, I would here also strive for completeness and include the other two studies of BrO from GOME-2: Hörmann et al., 2013 and Sihler et al., 2012 (already included in the references)
Page 3, Line 14: also here I would complete the list of references who used an area as an estimate for the stratospheric correction and include Hörmann et al., 2013 to the list of Wagner et al., 2001; Hörmann et al., 2016).
Page 4 line 22: I believe you mean “local solar time”?
Page 13 line 17: The comm in “(i.e., […])” is not needed.
Page 13 line 20: add “separately” after “across-track position”
Page 38 footnote: somewhere there is a bracket missing or one too many.
Citation: https://doi.org/10.5194/egusphere-2023-1163-RC2 - AC2: 'Reply on RC2', Heesung Chong, 17 Nov 2023
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1163', Anonymous Referee #1, 08 Jul 2023
Review of “Global retrieval of stratospheric and tropospheric BrO columns from OMPS-NM onboard the Suomi-NPP satellite”
In this manuscript, the authors describe a new BrO data product using OMPS-NM measurements. The product includes stratospheric and tropospheric columns, uses a complex stratosphere – troposphere separation algorithm, applies pixel specific airmass factors and provides detailed uncertainty estimates. The topic of the manuscript fits well into the AMT scope, the product described is of interest to the atmospheric chemistry community and the paper is well written and includes detailed descriptions of the algorithms used. I therefore recommend it for publication in AMT.
I have however several questions, comments and suggestions to the manuscript and the algorithm, which the authors should address before the manuscript is accepted:
1) Reference sector correction
What the authors call reference sector correction in fact includes two corrections: a) the addition of the modelled BrO offset necessary when using a radiance background spectrum and b) a latitudinal correction of the slant columns based on the model. The latter correction is critical as the way I understand it, it forces the baseline of the measurements on the model values. Therefore, I think the stratospheric BrO product is lo a large extent just reproducing the model values. This is in contrast to the statements in the manuscript claiming that both stratospheric and tropospheric column are retrieved from the measurements.
Please a) discuss this point and b) include an example of the correction for one full orbit, for example one of those shown in Fig. 3.
2) CAM-chem climatology
The algorithm described heavily relies on the CAM-chem climatology for the stratospheric columns (see above), the separation between stratospheric and tropospheric signals and for the airmass factors. However, a) it is not clear how well the climatology represents the real atmospheric BrO field and b) tropospheric BrO enhancements are very dynamic events, and their magnitude and location cannot be reflected by a static monthly climatology. This has important implications for the airmass factors which probably are often not correct as the monthly mean profiles are neither a good representation of BrO events, nor of background conditions. The algorithm foresees the use of “flattened” profiles in case the measurements with low BrO columns, but the opposite case (high BrO in a region where the climatology does not expect a BrO event) is not treated separately.
Please a) discuss the impacts of using a climatology as input for the airmass factors and b) include a figure comparing the modelled climatological tropospheric columns in comparison to the measurements, for example for the orbits shown in Fig. 3.
3) Polar vortex
The authors discuss the well known problem of using O3 columns as proxy for the stratospheric BrO columns and state, that their method “still preserves the overall spatial pattern of the stratospheric field”. I have not understood why that should be the case. If we have ozone depletion (and possibly stratospheric BrO enhancement) within the vortex, the relationship between O3 and BrO will be different for vortex and non-vortex air masses, leading to large scatter in the O3 – BrO plot. If all the in vortex values are removed and filled with surrounding (out of vortex) values, the stratospheric BrO column will not be correct.
Please explain why your method is less affected by this problem than that of previous studies.
4) Uncertainties
Although a detailed discussion of uncertainties is given, I’m somewhat confused by what to expect from the data product. Are the uncertainties given for individual pixels? Has each pixel in the product an uncertainty value? Will uncertainties be smaller in monthly averages? If so, why are the DSCD uncertainties shown in Figs. 12 and 13 for monthly averages comparable to the median uncertainty quoted for an individual pixel? How is the uncertainty of having a high surface BrO event in the data at a location where the CAM-chem climatology has background conditions taken into account? Can the product be used on a daily basis (Figs. 3 and 4 suggest that this is the case) or should it better be used on a monthly basis?
Please add error bars to the satellite data in Fig. 7 and a paragraph on data usage.
5) High values over the ocean
In Fig. 11, large BrO columns are shown over much of the NH oceans, and as far as I can see, the largest BrO columns in that month are not found in the Arctic but somewhere over the Pacific. Do you think this is realistic? How do these findings compare to other satellite products and independent measurements? Are these high columns already visible in the slant columns or are the introduced by the CAMS-chem based airmass factors?
6) Comparison to other satellite data
Why is there no comparison to other satellite data? The data set is advertised as extension of the OMI afternoon BrO time series, and I think it would be good to include some kind of comparison between BrO observations from the two platforms, even if it is just a visual side-by-side comparison of a monthly average.
7) Profile flattening
This procedure seems arbitrary to me. I do not see why such a profile should give reasonable tropospheric BrO columns or realistic airmass factors. Please justify your approach.
8) Coastal artefacts
In Fig. 10, there are many localised spots of suspiciously high tropospheric BrO along the Antarctic coast but also at the sea ice edge. As this is a longterm average, it is clear that the high values are from too small airmass factors, and this is confirmed by figure (e). In my opinion, this should not be part of the product as it clearly is an artefact. I expect similar artefacts in the daily images also in the NH, for example close to Spitsbergen. My suggestion is to at least include a flag for those retrievals having very small airmass factors, and to increase the uncertainty estimates for such pixels.
9) Large AMF above ocean
I’m surprised by the large values of the airmass factor over much of the oceans. Values above two indicate the presence of a significant fraction of the BrO in the free troposphere. Do you think this is realistic?
10) Data availability
I understand that the product is not yet released, but to my knowledge, the data has to be available in some form for the manuscript to be published in AMT. Maybe just add that it is available on request?
Citation: https://doi.org/10.5194/egusphere-2023-1163-RC1 - AC1: 'Reply on RC1', Heesung Chong, 17 Nov 2023
-
RC2: 'Comment on egusphere-2023-1163', Anonymous Referee #2, 08 Aug 2023
In the paper titled "Global Retrieval of Stratospheric and Tropospheric BrO Columns from OMPS-NM on the Suomi-NPP Satellite," Chong and colleagues introduce a new BrO column product derived from the OMPS-NM satellite instrument. The authors focus on presenting a long-term time series of tropospheric columns while effectively distinguishing between stratospheric and tropospheric contributions. The paper's strength lies in its innovative approach, combining the strengths of two state-of-the-art methods to separate stratospheric and tropospheric columns. Furthermore, the long operational lifespan of the OMPS-NM instruments lends a distinct advantage to this product, promising continuous data acquisition over an extended period which allows long-term time series of this chemically important tracer. This BrO column product makes a valuable addition to existing satellite-derived BrO products, with potential benefits for bromine chemistry research and is therefore in good alignment with the scope of AMT. I strongly recommend considering this paper for publication after addressing the noted corrections and points.
General comments:
GC1) The manuscript appears to primarily target polar applications, with retrieval considerations and examples predominantly centered around polar regions. For instance, the effective application of the flattening technique to polar hot-spots raises questions about its performance in non-polar tropospheric enhancements, confer also GC2. Moreover, the wavelength criterion selection (Page 39, line 6-7) implies a concentration on polar applications, aiming to achieve optimal results under high latitude and SZA conditions. While this focus is justified, I suggest to explicitly mention in the abstract, introduction, and potentially the title. Consider either highlighting the emphasis on polar BrO retrieval or substantiating why polar regions pose the greatest challenge.
GC2) Even though it is maybe a bit overcomplicated, I generally like the authors approach for the separation of tropospheric and stratospheric columns as it combines the two pathways taken in previous studies. However, there are several steps where I can see potential issues:
- How does step (i), flattening the tropospheric model profile (page 17, Figure 1), perform with non-polar tropospheric enhancements? Is it specifically tailored for this scenario? For instance, if an extensive area of BrO tropospheric enhancements emerges in the equatorial Pacific due to an unusual climate change event or a significant volcanic plume, how would step (i) handle this situation? It appears that step (i) might overlook such occurrences, possibly leading to their inclusion in V_strat_0 (which may not be problematic by itself). However, a concern arises regarding the effectiveness of the O3-BrO relation-based separation (Figure 4c) under such circumstances.
- Page 19: In step (v) the authors suggest to use the model profile as input for AMF calculation under no hotspot conditions and in difference the “flattened” model profile as input for the AMF calculation under hotspot conditions. The validity of this approach does not occur to me as I have the following concerns/points.
- How confident are you, that this model profile is representative for the tropospheric profile under “no-hotspot” conditions? Please shortly address this in the manuscript.
- For hotspot cases, the authors opt for the flattened tropospheric profile. While not explicitly stated, it seems this choice stems from the model profile being decidedly unsuitable for such scenarios. However, employing a flattened profile could potentially worsen the situation. This selection could yield a lower and possibly excessively low tropospheric AMF (A_trop_select, denominator in equation 12). Considering the high S_trop (numerator in equation 12), this might lead to an excessively high V_trop. This concern is particularly relevant when contrasting with non-hotspot pixels where a non-flattened profile is used for tropospheric AMF calculation. Consequently, I have three queries/suggestions regarding the treatment of hotspot cases:
- Is there a justification for favoring a flattened profile over the model profile as an assumption?
- Given that AMF carries a larger share of random uncertainties, could it be more accurate and practical to assume a distinct "hot-spot" profile shape (e.g., employing a ground-level profile shape as in Fig. 3e for polar hot spots, and a different assumption for tropical cases)?
- Is this factor considered in the error propagation of the AMF? Given its potential significance, it would be beneficial to explicitly acknowledge this and its impact.
- Page 19, line 13: Like Sihler et al. (2012), step (iii) employs the assumption of a constant O3/BrO relation to quantify tropospheric enhancements. However, the authors note that this assumption becomes problematic in polar vortex scenarios, potentially introducing bias to your data. Is this data utilized in the final product? If yes, provide rationale for its inclusion and discuss implications for the data quality flag.
GC3) The paper would benefit from a comparison with other satellite studies. This does not need to be thorough, but differences and agreements should be adressed both with regard to the polar observations as well as the BrO from Rann of Kutch.
Specific Comments
Abstract:
The abstract could benefit from a clearer articulation of the paper’s primary goal and focus. Consider adding a succinct sentence at the beginning of line 4 to outline the central theme of the study. This could lead into the subsequent statement, "To address this concern and improve upon the current methods, our study introduces..." This adjustment would help provide a smoother transition into the specific achievements and advancements discussed.
Introduction:
Page 3, lines 10-29: The provided overview of the broad variety of separation schemes used in the literature is commendably thorough and informative. Nonetheless, its level of detail seems too thorough for an introduction. In the introduction, the focus should be on the paper's new method, and a concise acknowledgment of various approaches would suffice to put the paper’s method into perspective
Given the absence of a designated "methods" section to accommodate such content as a subsection, I understand the authors' predicament in determining its placement. To address this, a practical solution could be integrating it as a subsubsubsection within Page 14, subsubsection 2.2.4.
Page 3 line 35-page 4 line 5: The paragraph's primary emphasis, as underscored by the authors first lines, lies in the substantial potential of OMPS to provide an extensive and enduring time-series well into the 2030s. This aspect should take precedence, and should be highlighted in the paragraphs first sentence, shifting the focus imediately to this critical attribute. Accordingly, I suggest starting with the assertion about OMPS's long time-series capability, and subsequently incorporating the initial sentence, "OMPS-NM instruments ... decommissioning of TROPOMI (Nowlan et al., 2023)," at the paragraph's conclusion.
Page 4, line 16-18 and fig. 1: There are two confusing aspects:
- The use of the numeration (1) – (4) in reference to fig. 1, lets the reader look for the numbers 1-4 in fig. 1. However, the corresponding fields are noted with (A)-(D). Please use the same symbols in text and figure
- When reading “highlighted in blue in fig. 1”, the first look in figure 1 will be to the fields which have a blue background “OMPS-NM L1B product” etc. As Figure 1 contains a lot of information it is difficult to find the highlighted fields. Either specify in the text that they are “encircled/framed in blue” or reconsider the coloring within fig. 1 to avoid this confusion. Also consider to increase the size of the border line to highlight the 4 fields. Furthermore, consider to place the A-D always at the same location w.r.t. the fields they refer to (either all on the top left or top right).
Page 4, lines 16-18 and Fig. 1: This section presents two points of confusion:
- The utilization of the numerals (1) – (4) in reference to Fig. 1 prompts readers to search for numbers 1-4 within the figure. However, the corresponding elements are actually labeled as (A)-(D). To enhance clarity, it's advisable to employ consistent symbols both in the text and the figure.
- When the text mentions "highlighted in blue in Fig. 1," readers instinctively turn their attention to fields with a blue background, such as "OMPS-NM L1B product," within Figure 1. Since the figure contains extensive information, locating the highlighted elements becomes challenging. To address this, you could specify in the text that the relevant fields are "encircled/framed in blue." Alternatively, reconsider the color scheme within Fig. 1 to alleviate this confusion. Additionally, consider enhancing the border line's size to accentuate the four designated fields. Furthermore, for consistency, contemplate consistently placing the labels (A)-(D) at the same relative position with respect to the corresponding fields (either all at the top left or top right).
By harmonizing symbols and refining visual cues, these adjustments can substantially improve the reader's comprehension.
Page 4 and 6: The mention of 2 times “retrieval” in headline 2 and 2.2 is redundant, I suggest to move all the subsubsections in 2.2 up by one rank in hierarchy (e.g. 2.2.1-> 2.1, etc.).
Page 6, line 6: The phrasing currently implies that Beirle et al., 2017; Nowlan et al., 2023 originated the super-gauss concept. I recommend revising it to "super Gaussian and adopt the approach outlined in Beirle et al., 2017; Nowlan et al., 2023."
Page 6, line 23: The authors specify their utilization of a 20° latitude portion within a single orbit for the computation of the earthshine reference spectrum. With an assumed along-track pixel footprint of 50 km, this approach implies that each across-track reference spectrum would be derived from 40-50 individual spectra. Notably, other investigations involving 2D CCD satellites like OMI and TROPOMI adopt larger sectors (e.g., Seo et al., 2018: 150°E – 240°W, 30°S-30°N for BrO; Theys et al., 2017: 120-160°W, 10°N-10°S for SO2). Please Justify the rationale behind employing a relatively compact reference sector and argue why such a low statistic is deemed satisfactory for your study.
Page 8, Table 1: Regarding the SCD retrieval: All (to the reviewers knowledge) recent other BrO DOAS and “DOAS-like” spectral retrievals include OClO in their spectral retrieval (e.g. Suleiman et al., 2019; Herrmanns et al., 2022; They et al., 2011; Sihler et al., 2012; Seo et al., 2019). Please justify your choice not to include it especially with respect to the potential spectral interferences (see overlapping absorption peak at 344nm).
A similar argument can be made for SO2 although it was only implemented in the most recent publications (Suleiman et al., 2019 (Proposed to be implemented); Herrmanns et al., 2022; Sihler et al., 2012; Seo et al., 2019). Please explain why you have not chosed to include it and how strong you estimate for spectra affected by SO2 (e.g. strong pollution emitter or volcanoes) as well as how substantial this impact is on the global data-set.
Page 8, Table 1: I suggest to highlight the trace gas absorption spectra in the list. For example by horizontal lines. Also add “the parameter are listed in their order of appearance in eq. 2”.
Page 9, Figure 2: It would be beneficial to also include the residual spectrum in this plot as it gives information on potential residual structures originating from absorbers which are not accounted for.
Page 13: In other studies (such as Seo et al. 2019) a uniform background of 3.5x1013 moleculesc cm-1 is used (based on Richter et al., 2002). Include how your background correction S_R typically is with respect to this value.
Richter et al., 2002: Richter, A., Wittrock, F., Ladstatter-Weissenmayer, A., and Burrows, J. P.: GOME measurements of stratospheric and tropospheric BrO, in: Remote Sensing of Trace Constituents in the Lower Stratosphere, Troposphere and the Earth’s Surface: Global Observations, Air Pollution and the Atmospheric Correction, edited by: Burrows, J. P. and Takeucki, N., Adv. Space Res., 11, 1667–1672, 2002.
Page 13 line 17-31: It would improve the readability to include the names “S_R” and “S_B”, when talking about these quantities in the text.
Page 37 line 20: Please remove "including volcanic plumes". Volcanic application is not mentioned at all in the result section and as the major volcanic constituent "SO2" is not accounted for in the spectral fitting, this statement is questionable.
Page 38, line 13-18, concerning the “modeled stratospheric BrO DeltaSCD”: I assume the “modeled Delta SCDs” in line 17 is the same as the “modeled stratospheric BrO Delta SCD”. Name both the same, and consider to mention this in the first sentence of its explanation (line 13).
Page 38, Line 15-17: From your explanation, it looks like the “modeled stratospheric BrO Delta SCD”, which is subtracted by is defined in a way that its mean is zero at 0-10°N and non-zero elsewhere. Thus the “Delta SCD bias” will then be the complete retrieved SCD subtracted by zero at the equator and non-zero
Page 38-39 and figure A1, regarding the correlation with O3:
- How have you combined the different O3 SCDs to one O3 SCD? Did you follow the formula proposed by Pukite and Wagner (2016) eq. 16? If so, please add a reference.
- I do not see the benefit of looking at two O3 absorptions at 243 and 273K, as there is only O3 absorption. If you do not gain any benefit from using the two, then I suggest to skip this.
- Should you chose to keep the distinction between 243 and 273K, how did you avoid a cross correlation between the two spectra (which are very similar)? Have you orthogonalized the O3 absorption spectrum at 273K w.r.t the one at 243K? Please state this in the text.
Page 39 line 8: The choice of a percentile seems suboptimal to me and introduced strong data-selection biases and I would urge to change this. For instance, the pixel at high VZA will have a higher SCD compared to the nadir looking pixel and they will be selected thus more frequently. Additionally, tropospheric enhancements will be more dominant. If you want to select for high latitude and high SZA, then why not use latitude and SZA as a selection criterium?
Additionally, the paper is about a global product of BrO. Please justify why you have not also looked if the fit is also performing well at other regions (cf. Seo et al., 2019, who performed a retrieval interval mapping for several cases and also for an equatorial region).
There is one paper about long-term BrO time-series:
Bougoudis, I., Blechschmidt, A.-M., Richter, A., Seo, S., Burrows, J. P., Theys, N., and Rinke, A.: Long-term time series of Arctic tropospheric BrO derived from UV–VIS satellite remote sensing and its relation to first-year sea ice, Atmos. Chem. Phys., 20, 11869–11892, https://doi.org/10.5194/acp-20-11869-2020, 2020.
I suggest to include this paper in your introduction. You can frame this as an advantage of the OMPS data-set who in difference to Bougoudis et al., 2020 does not require a complicated inter-calibration of the time-series of the different sensors.
Technical comments:
Page 2, Line 30: As the authors are very thorough in giving a complete list of relevant references in the introduction, I would here also strive for completeness and include the other two studies of BrO from GOME-2: Hörmann et al., 2013 and Sihler et al., 2012 (already included in the references)
Page 3, Line 14: also here I would complete the list of references who used an area as an estimate for the stratospheric correction and include Hörmann et al., 2013 to the list of Wagner et al., 2001; Hörmann et al., 2016).
Page 4 line 22: I believe you mean “local solar time”?
Page 13 line 17: The comm in “(i.e., […])” is not needed.
Page 13 line 20: add “separately” after “across-track position”
Page 38 footnote: somewhere there is a bracket missing or one too many.
Citation: https://doi.org/10.5194/egusphere-2023-1163-RC2 - AC2: 'Reply on RC2', Heesung Chong, 17 Nov 2023
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 642 | 213 | 31 | 886 | 27 | 25 |
- HTML: 642
- PDF: 213
- XML: 31
- Total: 886
- BibTeX: 27
- EndNote: 25
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Heesung Chong
Gonzalo González Abad
Caroline R. Nowlan
Christopher Chan Miller
Alfonso Saiz-Lopez
Rafael P. Fernandez
Hyeong-Ahn Kwon
Zolal Ayazpour
Huiqun Wang
Amir H. Souri
Xiong Liu
Kelly Chance
Ewan O’Sullivan
Jhoon Kim
Ja-Ho Koo
William R. Simpson
François Hendrick
Richard Querel
Glen Jaross
Colin Seftor
Raid M. Suleiman
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(9927 KB) - Metadata XML