the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 06 Jul 2023
A physically-based correction for stray light in Brewer spectrophotometer data analysis
Henri Diémoz
C. Thomas McElroy
Abstract. Brewer ozone spectrophotometers have become an integral part of the global ground-based ozone monitoring network collecting data since the early 1980s. The double monochromator Brewer version (MkIII) was introduced in 1992. With the Brewer hardware being so robust, both single and double monochromator instruments are still in use. The main difference between the single Brewers and the double Brewers is the much lower stray light in the double instrument. Laser scans estimate the rejection level of the single Brewers to be 10-4.5 while the doubles improve this to 10-8, virtually eliminating the effects of stray light. For a typical single monochromator Brewer, stray light leads to an underestimation of ozone of approximately 1 % at 1000 DU ozone SCD and can exceed 5 % at 2000 DU, while underestimation of sulphur dioxide reaches 30 DU when no sulphur dioxide is present. This is because even a small additional stray light at shorter wavelengths significantly reduces the slant ozone at large values. An algorithm for stray light correction based on the physics of the instrument response to stray light which adds light from longer wavelengths to shorter ones has been developed. The simple assumption is that count rates measured at any wavelength have a contribution from stray light from longer, and thus brighter, wavelengths because of the ozone cross-section gradient leading to a rapid change in intensity as a function of wavelength. Using the longest measured wavelength (320 nm) as a proxy for the overall brightness provides an estimate of this contribution. The sole parameter, in the order of 0.2 to 0.6 %, that describes the percentage of light at the longest wavelength to be subtracted from all channels is determined by matching ozone calculations from the single and the double monochromator Brewers. Removing this additional count rate from the signal mathematically before deriving ozone corrects for the extra photons scattering within the instrument that produces the stray light effect. Analysing historical data from co-located single and double monochromator Brewers can provide an estimate of how the stray light contribution changes over time in an instrument. The corrected count rates of the measured wavelengths can also be used to improve other calculations: the sulphur dioxide column, the aerosol optical depth, the effective temperature of the ozone layer or any other products. Also presented, is an initial analysis of signs consistent with the stray light effect in the double monochromator Brewers. A multi-platform code to correct the count rates for stray light and saving the corrected values in a new B-file for use with any existing Brewer data analysis software is available to the global Brewer user community at https://zenodo.org/record/8097039 (Savastiouk and Diémoz, 2023).
-
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
(3989 KB)
-
Supplement
(2395 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(3989 KB) - Metadata XML
-
Supplement
(2395 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Vladimir Savastiouk et al.
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1461', Anonymous Referee #1, 18 Jul 2023
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1461/egusphere-2023-1461-RC1-supplement.pdf
- AC1: 'Reply on RC1', Vladimir Savastiouk, 31 Aug 2023
-
RC2: 'Comment on egusphere-2023-1461', Anonymous Referee #2, 19 Jul 2023
This manuscript presents a highly innovative method of correcting stray light derived errors in total column amounts of ozone and sulfur dioxide measured by the single Brewer spectrophotometers. Different from the certain correction methods, the presented has very clear basis of physics consideration and more feasible application for the Brewer ozone spectrophotometer community. Furthermore, this physics-based method is also validated by theoretical simulations. The manuscript has a very significant application value for the single Brewer spectrophotometer. All the authors of the manuscript are well-known scientists who have long been engaged in works of Brewer spectrophotometer. This manuscript is recommended to be published with small revisions. There are following two comments:
1）Line # 278, "The ozone ETC, in simulation S3, strictly agree with the one observed without stray light in the model (2583) “---What is the meaning of ‘model (2853)’?
2）Given the significant role of travelling standard # 017 in history, a recommendation of reprocessing the historical data measured by those single Brewer spectrophotometers which had been periodically calibrated by #017 had better been added in "3.2.5 Changes in stray light with time" when Table IV is shown.
Citation: https://doi.org/10.5194/egusphere-2023-1461-RC2 - AC2: 'Reply on RC2', Vladimir Savastiouk, 31 Aug 2023
-
RC3: 'Comment on egusphere-2023-1461', Anonymous Referee #3, 11 Aug 2023
Review of Savastiouk et al. , A physically-based correction for stray light in Brewer spectrophotometer data analysis
General comments
The submitted manuscript describes a new method for correcting stray light effects for single monochromator Brewer spectrophotometers.
The authors propose a very simple parameterization which they justify using calculations from a raditive transfer model and multiple examples of observational data, with quite impressive results. The new algorithm has been implemented in software available from the authors.
The topic is a very important one for the atmospheric measurement community, the manuscript is clearly written and in my opinion is suitable for publication in AMT with only minor revisions.
An appendix has been included describing the working principles of the Brewer as needed to follow the rest of the paper, which is very helpful.
At times the language does verge into something more like an advertisement (eg lines 530-538) rather than a sober and objective scientific paper.
Lots of interesting discussion has been included and the authors clearly have spent a lot of time thinking deeply about Brewers, but at times some of this discussion is probably to the detriment of the overall focus and impact.
The arguments presented seem quite sound to me and convincing, but before the method is widely adopted in the global Brewer network, I would hope to see some careful comparisons of the results obtained compared to those from other methods, in particular the Redondas et al. technique which I believe is now standard in EUBREWNET.
A final general comment is that the authors seem to intermingle the discussion of the algorithm itself with their own software implementation of it, which I found confusing at some points. In principle, the algorithm and the software are distinct. They seem to refer to both as "PHYCS", although this name does not appear in the abstract.
My comments are mostly very minor.
Specific comments
Line 5 You say "virtually eliminating" but then only shortly later (line 21) you say you can see it. In practice your method relies on assuming the comparison MKIII has no stray light it seems to me?
Line 8 Add a word such as "contribution" or "effect" after "a small"
Line 9 Please reword – stray light doesn't reduce the "slant ozone" itself of course.
Line 15 The reader can't know what "matching ozone calculations from the single and double mono-chromator Brewers" means, please re-word for better clarity, ie side-by-side comparison
Lines 18-21 I suggest removing these lines from the abstract. Lines 18-22 describe further work, and 20-21 is a very tentative result.
Line 29 Why do you specifically mention NDACC here? I count only eight Brewer stations on the NDACC website! Wouldn’t WMO-GAW or EUBREWNET be more appropriate?
Line 52 Only true for the northern hemisphere, and not Antarctica.
Line 57 Remove "the" in "the space"
Line 67 Is "PHYCS" the algorithm or the software?
Line 67 How do the authors want "PHYCS" to be pronounced?! ("Ficks"?)
Line 71 I don't think you mean "peculiar" here.
Lines 78-79 Wouldn't this advantage also be the case for other available stray-light correction methods?
Table 2 For consistency, Hobart should be in Australia – for the other sites you have given the name of the country.
Table 2 I suggest putting the stations in the table in order of latitude or magnitude of latitude, so that the reader can more easily grasp the "vastly different observing conditions" which you claim.
Lines 106-113 Personally I found the construction of this run of sentences quite repetitive, all beginning the same way, "Additionally,", "Indeed,", "Additionally" and "In fact,".
Figure 1 The figure is quite important for the paper but it seems strangely orphaned, without much description of where it has come from. Is there a reference? (It sounds like it was work done twenty years ago but never written up?)
Line 142 Shouldn't what you have called here I320 be written as Id (320) to follow the notation of equation (1)?
Line 144 At this point, alpha is not a constant, but a function of the wavelength and presumably other parameters to describe the atmosphere. The hypothesis coming next is that alpha in fact doesn't depend on the wavelength or the conditions.
Lines 146-149 I find this confusing because here you seem to be saying there is one value of alpha for slits one to five but shortly afterwards you introduce beta for slit 1 and later alpha only actually seems to ever be used for slit 2 anyway?
Line 170 Add "in practice" or similar wording after "In most cases"
Lines 170-172 I would say an absolute calibration can be performed wherever you like but the results might not be very good, so add "to the desired uncertainty" or words like that here.
Line 200 Change to either "in the tropics" or "close to the equator"
Lines 2011-2014 But wouldn't other problems also present the same symptoms, for example any change in the calibration?
Line 220 "first" not "fist"
Figure 2 You should label the dashed lines in both (a) and (b) clearly with the slit numbers, this would make the text much easier to follow.
Line 276-278 I would tone the language down a little bit here, ie "impressive", "especially", "perfectly" .
Lines 279-282 Yes this is a good point to mention, I was wondering about whether high aerosol optical depth (in the UV) would have an effect, possibly depending on its spectral properties. Is this worth discussing?
Line 288 I don’t think you mean "likely" here?
Lines 336-338 Just to be clear here, you are using the Mk III to derive alpha and then comparing the new results with the same Mk III instrument again?
Line 351 Are you worried that these "simplified assumptions" and "slight discrepancies" have such a big effect?
Line 353 Insert "the" before "year"
Lines 424-470 I'm pleased that you have included the discussion of uncertainties. Everything you say seems quite reasonable to me, but I wouldn't be surprised if later work modified some of the conclusions to some extent.
Lines 471 – 493 I am not sure of the value of section 3.2.7. For this method to be applied in the network you need to assume the Mk IIIs have no stray light, don't you?
Lines 486-493 This section seems to be almost thinking aloud, and I suggest deleting it. Even if you could perform this experiment, what specific indications would you be looking for as a signature of stray light, rather than some other issue?
Line 506 Delete "to"
Lines 502-512 This is quite an interesting discussion, but it would be better with some numbers. The situation of a Brewer being calibrated at low latitude but then being operated at high latitudes must have happened a lot over the last thirty years.
Line 513 "A few word but an important addition" .. please re-word
Lines 530-538 This reads like a salesman's pitch! Please re-word to be more specific and informative.
Line 569 "Effortlessly" is excessive
Line 640 I would like to see this point expanded just slightly, to explain the combination of the laboratory cross-section and the particular slit function
Citation: https://doi.org/10.5194/egusphere-2023-1461-RC3 - AC3: 'Reply on RC3', Vladimir Savastiouk, 31 Aug 2023
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1461', Anonymous Referee #1, 18 Jul 2023
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1461/egusphere-2023-1461-RC1-supplement.pdf
- AC1: 'Reply on RC1', Vladimir Savastiouk, 31 Aug 2023
-
RC2: 'Comment on egusphere-2023-1461', Anonymous Referee #2, 19 Jul 2023
This manuscript presents a highly innovative method of correcting stray light derived errors in total column amounts of ozone and sulfur dioxide measured by the single Brewer spectrophotometers. Different from the certain correction methods, the presented has very clear basis of physics consideration and more feasible application for the Brewer ozone spectrophotometer community. Furthermore, this physics-based method is also validated by theoretical simulations. The manuscript has a very significant application value for the single Brewer spectrophotometer. All the authors of the manuscript are well-known scientists who have long been engaged in works of Brewer spectrophotometer. This manuscript is recommended to be published with small revisions. There are following two comments:
1）Line # 278, "The ozone ETC, in simulation S3, strictly agree with the one observed without stray light in the model (2583) “---What is the meaning of ‘model (2853)’?
2）Given the significant role of travelling standard # 017 in history, a recommendation of reprocessing the historical data measured by those single Brewer spectrophotometers which had been periodically calibrated by #017 had better been added in "3.2.5 Changes in stray light with time" when Table IV is shown.
Citation: https://doi.org/10.5194/egusphere-2023-1461-RC2 - AC2: 'Reply on RC2', Vladimir Savastiouk, 31 Aug 2023
-
RC3: 'Comment on egusphere-2023-1461', Anonymous Referee #3, 11 Aug 2023
Review of Savastiouk et al. , A physically-based correction for stray light in Brewer spectrophotometer data analysis
General comments
The submitted manuscript describes a new method for correcting stray light effects for single monochromator Brewer spectrophotometers.
The authors propose a very simple parameterization which they justify using calculations from a raditive transfer model and multiple examples of observational data, with quite impressive results. The new algorithm has been implemented in software available from the authors.
The topic is a very important one for the atmospheric measurement community, the manuscript is clearly written and in my opinion is suitable for publication in AMT with only minor revisions.
An appendix has been included describing the working principles of the Brewer as needed to follow the rest of the paper, which is very helpful.
At times the language does verge into something more like an advertisement (eg lines 530-538) rather than a sober and objective scientific paper.
Lots of interesting discussion has been included and the authors clearly have spent a lot of time thinking deeply about Brewers, but at times some of this discussion is probably to the detriment of the overall focus and impact.
The arguments presented seem quite sound to me and convincing, but before the method is widely adopted in the global Brewer network, I would hope to see some careful comparisons of the results obtained compared to those from other methods, in particular the Redondas et al. technique which I believe is now standard in EUBREWNET.
A final general comment is that the authors seem to intermingle the discussion of the algorithm itself with their own software implementation of it, which I found confusing at some points. In principle, the algorithm and the software are distinct. They seem to refer to both as "PHYCS", although this name does not appear in the abstract.
My comments are mostly very minor.
Specific comments
Line 5 You say "virtually eliminating" but then only shortly later (line 21) you say you can see it. In practice your method relies on assuming the comparison MKIII has no stray light it seems to me?
Line 8 Add a word such as "contribution" or "effect" after "a small"
Line 9 Please reword – stray light doesn't reduce the "slant ozone" itself of course.
Line 15 The reader can't know what "matching ozone calculations from the single and double mono-chromator Brewers" means, please re-word for better clarity, ie side-by-side comparison
Lines 18-21 I suggest removing these lines from the abstract. Lines 18-22 describe further work, and 20-21 is a very tentative result.
Line 29 Why do you specifically mention NDACC here? I count only eight Brewer stations on the NDACC website! Wouldn’t WMO-GAW or EUBREWNET be more appropriate?
Line 52 Only true for the northern hemisphere, and not Antarctica.
Line 57 Remove "the" in "the space"
Line 67 Is "PHYCS" the algorithm or the software?
Line 67 How do the authors want "PHYCS" to be pronounced?! ("Ficks"?)
Line 71 I don't think you mean "peculiar" here.
Lines 78-79 Wouldn't this advantage also be the case for other available stray-light correction methods?
Table 2 For consistency, Hobart should be in Australia – for the other sites you have given the name of the country.
Table 2 I suggest putting the stations in the table in order of latitude or magnitude of latitude, so that the reader can more easily grasp the "vastly different observing conditions" which you claim.
Lines 106-113 Personally I found the construction of this run of sentences quite repetitive, all beginning the same way, "Additionally,", "Indeed,", "Additionally" and "In fact,".
Figure 1 The figure is quite important for the paper but it seems strangely orphaned, without much description of where it has come from. Is there a reference? (It sounds like it was work done twenty years ago but never written up?)
Line 142 Shouldn't what you have called here I320 be written as Id (320) to follow the notation of equation (1)?
Line 144 At this point, alpha is not a constant, but a function of the wavelength and presumably other parameters to describe the atmosphere. The hypothesis coming next is that alpha in fact doesn't depend on the wavelength or the conditions.
Lines 146-149 I find this confusing because here you seem to be saying there is one value of alpha for slits one to five but shortly afterwards you introduce beta for slit 1 and later alpha only actually seems to ever be used for slit 2 anyway?
Line 170 Add "in practice" or similar wording after "In most cases"
Lines 170-172 I would say an absolute calibration can be performed wherever you like but the results might not be very good, so add "to the desired uncertainty" or words like that here.
Line 200 Change to either "in the tropics" or "close to the equator"
Lines 2011-2014 But wouldn't other problems also present the same symptoms, for example any change in the calibration?
Line 220 "first" not "fist"
Figure 2 You should label the dashed lines in both (a) and (b) clearly with the slit numbers, this would make the text much easier to follow.
Line 276-278 I would tone the language down a little bit here, ie "impressive", "especially", "perfectly" .
Lines 279-282 Yes this is a good point to mention, I was wondering about whether high aerosol optical depth (in the UV) would have an effect, possibly depending on its spectral properties. Is this worth discussing?
Line 288 I don’t think you mean "likely" here?
Lines 336-338 Just to be clear here, you are using the Mk III to derive alpha and then comparing the new results with the same Mk III instrument again?
Line 351 Are you worried that these "simplified assumptions" and "slight discrepancies" have such a big effect?
Line 353 Insert "the" before "year"
Lines 424-470 I'm pleased that you have included the discussion of uncertainties. Everything you say seems quite reasonable to me, but I wouldn't be surprised if later work modified some of the conclusions to some extent.
Lines 471 – 493 I am not sure of the value of section 3.2.7. For this method to be applied in the network you need to assume the Mk IIIs have no stray light, don't you?
Lines 486-493 This section seems to be almost thinking aloud, and I suggest deleting it. Even if you could perform this experiment, what specific indications would you be looking for as a signature of stray light, rather than some other issue?
Line 506 Delete "to"
Lines 502-512 This is quite an interesting discussion, but it would be better with some numbers. The situation of a Brewer being calibrated at low latitude but then being operated at high latitudes must have happened a lot over the last thirty years.
Line 513 "A few word but an important addition" .. please re-word
Lines 530-538 This reads like a salesman's pitch! Please re-word to be more specific and informative.
Line 569 "Effortlessly" is excessive
Line 640 I would like to see this point expanded just slightly, to explain the combination of the laboratory cross-section and the particular slit function
Citation: https://doi.org/10.5194/egusphere-2023-1461-RC3 - AC3: 'Reply on RC3', Vladimir Savastiouk, 31 Aug 2023
Peer review completion
Journal article(s) based on this preprint
Vladimir Savastiouk et al.
Data sets
PHYCS Data and software Vladimir Savastiouk and Henri Diémoz https://doi.org/10.5281/zenodo.8097039
Vladimir Savastiouk et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 348 | 116 | 25 | 489 | 32 | 9 | 9 |
- HTML: 348
- PDF: 116
- XML: 25
- Total: 489
- Supplement: 32
- BibTeX: 9
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(3989 KB) - Metadata XML
-
Supplement
(2395 KB) - BibTeX
- EndNote
- Final revised paper