Deriving Aerosol Absorption Properties from Solar Ultraviolet Radiation Spectral Measurements at Thessaloniki, Greece
, Athanasios Natsis, Nikolaos Siomos, Theano Drosoglou and Alkiviadis F. Bais
Round 1
Reviewer 1 Report
The manuscript titled “Deriving aerosol absorption properties from solar ultraviolet radiation spectral measurements at Thessaloniki, Greece”, by Fountoulakis et al., presents systematic measurements of the aerosol single scattering albedo and absorption aerosol optical depth derived from a double monochromator Brewer spectrophotometer in the period 1998 – 2017. The main uncertainty factors are quantified. Observed variations in these optical parameters are related to the economical crisis in Greece. The main conclusions are well supported by the data analysis and only several issues require further explanation by the authors.
Firstly, the database must be explained in a more clear fashion. Along the manuscript, different years are employed for different studies, sometimes providing an explanation (problem in Brewer data, limitation of SZA, availability of CIMEL data) but other times without it. It is necessary to explain which data allow the different studies and provide further details about how they are selected and filtered (AOD and SZA limits)
Secondly, the definition of AAOD must be revised (lines 149-151). It is rather worriying that the AAOD is calculated as SSA * AOD instead of the standard (1-SSA)*AOD. Hopefully it is a mistake in the text and the rest of the work shouldn’t be completely revised
Finally, the references must refer to the original work that proposed the idea or reached the conclusion, not to one of the authors that has mentioned it. For instance, the idea of using the 1.5 level AERONET data to enrich the database is from Bond et al (10.1002/jgrd.50171), not Kazadzis [35] or Rapptis [52]. Please revise references thoroughly
Adding to these, some parts of the manuscript needed clarification, included as notes in the attached file. Taking all this into account, my recommendation is accept after minor revision.
Comments for author File: 
Comments.pdf
Author Response
We want to acknowledge Anonymous Reviewer#1 for his helpful comments. We have replied to each of his general comments, as well as his specific comments in the manuscript (pdf file). A version of the new manuscript with tracked changes will be also uploaded in order to help the reviewer find the changes in the document.
Reviewer comment
Firstly, the database must be explained in a more clear fashion. Along the manuscript, different years are employed for different studies, sometimes providing an explanation (problem in Brewer data, limitation of SZA, availability of CIMEL data) but other times without it. It is necessary to explain which data allow the different studies and provide further details about how they are selected and filtered (AOD and SZA limits)
Reply
As the reviewer suggested, a new paragraph has been added at the end of Section 2, where all the relative information is summarized.
Reviewer comment
Secondly, the definition of AAOD must be revised (lines 149-151). It is rather worriying that the AAOD is calculated as SSA * AOD instead of the standard (1-SSA)*AOD. Hopefully it is a mistake in the text and the rest of the work shouldn’t be completely revised
Reply
Thankfully it was a typo and has been revised
Reviewer comment
Finally, the references must refer to the original work that proposed the idea or reached the conclusion, not to one of the authors that has mentioned it. For instance, the idea of using the 1.5 level AERONET data to enrich the database is from Bond et al (10.1002/jgrd.50171), not Kazadzis [35] or Rapptis [52]. Please revise references thoroughly
Reply
Corrected
Reviewer comment
Adding to these, some parts of the manuscript needed clarification, included as notes in the attached file.
Reply
There is a detailed reply for each of the reviewer’s comments in the attached document.
Author Response File: Author Response.pdf
Reviewer 2 Report
Minor deficiencies found. Please incorporate and fix all of them.
Comments for author File: Comments.pdf
Author Response
All the corrections suggested by the reviewer have been directly applied on the new manuscript.
Reviewer 3 Report
The authors used spectral UV measurements in Thessaloniki, Grace to calculate the aerosol single scattering albedo (SSA) and absorption optical depth (AAOD) from 1998 to 2017. The uncertainty factors of this retrieval are analyzed. Because the aerosol absorption is stronger in UV compared to visible bands, the adoption of visible SSA values in models leads to significant biases in the computed UV radiation. The daily and seasonal variabilities of SSA and AAOD and potential causes are also discussed. This study is a significant contribution to narrow the gap in our understanding of the aerosol radiative effects in UV bands. I recommend this manuscript to be accepted after a minor revision.
Specific comments:
Line 93: “SZA”, solar zenith angle? Please define here.
Line 448-449: The increase of the AAOD with wavelength? Or with decreasing wavelength?
Line 457: with or within?
Line 458-459: “as most winter measurements are not included”. Is this because of the low SZA or AOD? Or something else?
Line 465: “SZA smaller than 70”. Does this introduce biases because the hour of day is different in different seasons?
Line 487: “SSA is greater in winter than in summer”. Not the case according to Fig. 2b.
Line 496-498: Are you suggesting the high SSA and large AOD contribute to the large AAOD in August or just pointing out the coincidence? Could you be more specific about the meteorological conditions?
Line 536-546: Are there any large scale contribution from other area? Are there some estimates or previous studies that can be used as a reference?
Line 551: SZAs between 50 to 70: could you translate that to local solar time? I am wondering if this choice can introduce some biases in diurnal cycle between seasons?
Line 560-561: “the AAOD” decreases by ~0.02”: not apparent from Fig. 6.
Fig. 6: The legend labels used in (d) should be explained here in addition to the definition in the text.
Author Response
We want to acknowledge Anonymous Reviewer#3 for his helpful comments. We have replied to each of them in this document. A version of the new manuscript with tracked changes will be also uploaded in order to help the reviewer find the changes in the document.
Reviewer comment
Line 93: “SZA”, solar zenith angle? Please define here.
Reply
Done
Reviewer comment
Line 448-449: The increase of the AAOD with wavelength? Or with decreasing wavelength?
Reply
The word “decreasing has been added (line 505 in the document with track-changes)
Reviewer comment
Line 457: with or within?
Reply
The correct is “with”. Remained as it was (line 513)
Reviewer comment
Line 458-459: “as most winter measurements are not included”. Is this because of the low SZA or AOD? Or something else?
Reply
It is due to the SZA limitation. It is now clarified in the document (line 515).
Reviewer comment
Line 465: “SZA smaller than 70”. Does this introduce biases because the hour of day is different in different seasons?
Reply
Indeed there is a bias, which however is small and does not affect the results significantly. However, for the sake of accuracy we re-analyzed the monthly climatological values using measurements only for hours 10 – 14, for which measurements are available throughout the year. Figure 3 has been also updated in the new version of the manuscript.
Reviewer comment
Line 487: “SSA is greater in winter than in summer”. Not the case according to Fig. 2b.
Reply
Corrected
Reviewer comment
Line 496-498: Are you suggesting the high SSA and large AOD contribute to the large AAOD in August or just pointing out the coincidence? Could you be more specific about the meteorological conditions?
Reply
The high AAOD in August is due to the high AOD. SSA is higher than in winter which means that the aerosols are less absorbing. The sentence has changed so that the meaning is clearer. A detailed discussion about the meteorological conditions being responsible for the summer maximum AOD is out of the scope of this study and is already provided in the cited literature, which has been updated.
Reviewer comment
Line 536-546: Are there any large scale contribution from other area? Are there some estimates or previous studies that can be used as a reference?
Reply
There is contribution from other areas. Relative discussion has been added, as well as the corresponding literature where more details can be found.
Reviewer comment
Line 551: SZAs between 50 to 70: could you translate that to local solar time? I am wondering if this choice can introduce some biases in diurnal cycle between seasons?
Reply
Depending on season, the particular range of SZAs corresponds to different hours of the day. So, yes, it introduces some (small) bias and affects the annual cycle of the presented quantities. However, since in Figure 6 we investigate and discuss the long term changes of the presented quantities, this effect is insignificant. This bias does not practically affect the year-to-year variability (for the annual averages or for individual months) or the long term changes presented here.
Reviewer comment
Line 560-561: “the AAOD” decreases by ~0.02”: not apparent from Fig. 6.
Reply
We think that the decrease of ~0.02 is obvious in the graph. The annual averages for 1998 -2000 are ~0.02 units above the averages for 2004 – 2005.
Reviewer comment
Fig. 6: The legend labels used in (d) should be explained here in addition to the definition in the text.
Reply
This already is clearly explained in the figure caption:
“(d) depicts the difference of the annual average anomalies of the global UV irradiance at 320 nm relative to 1998 from the measurements (yellow), as well as from the simulations using the measured SSA (green) and a default SSA of 0.85 (blue). “
