Continuing the MODIS Dark Target Aerosol Time Series with VIIRS

Round 1
Reviewer 1 Report
The article "Continuing the MODIS Dark Target aerosol time 2 series with VIIRS" by Sawyer et al., provides a comprehensive evaluation and comparison of VIIRS AOD product with MODIS and ground based observations. This article is well written and the results discussed are supported by the figures. I think this is a very important paper especially for aerosol science community and climate modeling groups since this provide information about the continuity of a legacy data set from MODIS. I only have very few minor comments that should be incorporated before publication of this manuscript.
lines: 62-74, I think Wei et al. 2019, provided a comprehensive evalutation and comparisons of multi-satellite products including MODIS terra and MODIS aqua. I would recommend to include that citation here. The citation:
de Leeuw, G., Holzer-Popp, T. etc.: Evaluation of seven European aerosol optical depthretrieval algorithms for climate analysis, Remote Sens. Envi-ron., 162, 295–315, 2015.
Wei, J., Peng, Y., Mahmood, R., Sun, L., and Guo, J.: Intercomparison in spatial distributions and temporal trends derived from multi-source satellite aerosol products, Atmos. Chem. Phys., 19, 7183–7207,
https://doi.org/10.5194/acp-19-7183-2019, 2019.
Since this paper is about continuity of AOD product, I wonder if a small discussion (one sentence/or more) could be added in the conclusions about how the AOD trends could be affected when calculated based on combining the two products. This is to provide a word of caution for potential users?
It would be useful if authors could provide a link for the VIIRS level 3 AOD product.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 2 Report
Review
This manuscript describes aerosol optical depth (AOD) obtained from VIIRS on board Suomi-NPP with customized MODIS Dark Target (DT) AOD algorithm. This is the first operational DT AOD algorithm for VIIRS. The DT AOD from VIIRS Suomi-NPP is compared with MODIS Aqua DT AOD and ground-based AOD from Aeronet stations. The agreement between the different AOD products is found to be good enough, so that MODIS AOD time series can be continued with VIIRS AOD when MODIS has to be discontinued. In the future, this enables creating a global climate data record of AOD.
Overall, the topic of the manuscript is timely and the results are promising. I recommend this manuscript for publication, though I have one major comment.
Major comment
My comment is related to Fig. 3, where time series of MODIS and VIIRS AOD as well as the AOD offset of SNPP-Aqua are presented. A seasonal cycle is observed for the SNPP-Aqua offset over land and the origins of this seasonality are investigated in the following sections.
On line 420-422 it is stated “Over ocean, VIIRS averages 0.006 higher than the MODIS Aqua AOD retrieval. This is smaller than the offset between MODIS Terra and MODIS Aqua, and it does not vary significantly by season (Figure 3).” In my opinion Fig. 3a shows a seasonal cycle in SNPP-Aqua offset over ocean that is at least as strong as over land (Fig. 3b), although it does not change sign. Please discuss also the seasonality in Fig. 3a SNPP-Aqua offset over ocean.
Minor comments
L34 Please define “GCOS”
L178-180 “for both MODIS and VIIRS standard products, this data comes from the nearest six-hourly Global Data Assimilation Model (GDAS) re-analysis product [34].” Would using higher temporal (and spatial) resolution re-analysis (e.g. ERA5) improve the quality of the AOD retrieval?
L305-306 “There is no combined Dark Target/Deep Blue AOD product for VIIRS at this time.” Do you know if a combined product is being planned?
L346-347 “Aqua is chosen from the two MODIS instruments because its equatorial crossing time is close to that of Suomi-NPP.” How close are Aqua and Suomi-NPP?
L358 “Figure 2” do you mean “Figure 1b”?
L410-411 Please use same y-axis limits in Fig. 3 panels a and b.
L796 Please comment also on the overestimation compared to Aeronet AOD (slope >1 in Figs. 1 and 2). Collocation seems to explain the difference between MODIS and VIIRS, but is the overestimation compared to Aeronet due to some other process?
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
Thank you for answering all comments!