Assessment of the Representativeness of MODIS Aerosol Optical Depth Products at Different Temporal Scales Using Global AERONET Measurements

Round 1

Reviewer 1 Report

Dear Authors,

 

I have now read your manuscript and I think it is a moderately useful study for the readers of Remote Sensing and the scientific community. From a quick look of the presented results seem quite convincing, even if I cannot completely judge them based on the clarifications dues on the methodology (see comment 1) and a systematic clarification and simplification of the text (see comment 2). So, before having the possibility to fully judge the quality of this manuscript and possibly express a recommendation to publish your study, there are two major points that have to be addressed.

1) The Method of AERONET/MODIS data colocation is not really clear to me. While I signal a whole lot of minor comments on the text and language throughout the whole manuscript (see comment 2), unclear description of the methodology requires a particular attention. If the methodology is not clear, it is difficult to judge the results of a study. More details of my concerns with this section are in the attached annotated PDF but please pay a particular attention at the description of the spatiotemporal matching described in support of Fig. 2a, for exemple: how the spatial matching is done (the "3x3" thing is very obscure for this reader)? How the daily matches are done? Another specific concern I have for the methodology you use, is that you include MODIS data with QF=1 ("marginal"). What is the impact of including these low-quality data?

2) While the manuscript's text is generally of an acceptable quality, I have found a very large amount of clumsy or unclear sentences, and proposed modifications in the attached annotated PDF. Most important, the manuscript is very verbose and involute in some sections. It is strongly suggested to simplify and reduce the text, where possible, and try to be a lot more synthetic. Synthesis is very important in scientific communication. Also the nomenclature of the different compared products and averages is very very (and innecessarily) complicated. All these "aHAOD_T", "aHAOD_A" etc must be simplified: please consider to explicitly indicate some information and not incorporate all into acronyms.  

 

Due to the aggregate points 1 and 2, I have made, in this review round, only a quick view of the Results and Discussions section and, once resolved points 1 and 2, I'll also read these sections with care.

 

Please amend the manuscript according to my recommendation and I'll be willing to re-review the revised version in the second round.

Sincerily,

Reviewer #1

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

In this manuscript, the authors have carried out an exhaustive assessment of Satellite (MODIS Terra and Aqua) derived AODs at different temporal scales (hourly, daily, and monthly) using ground based AERONET measurements. For this study they have analyzed over 17 years of global and co-located Satellite and in situ data. By carrying out an extensive error analysis, the authors have found several interesting results, which can help understand the reliability of the Satellite AOD products and their associated uncertainties.

Author Response

Thanks for your encouraging comments.

Reviewer 3 Report

This paper investigates the performance of MODIS aerosol data by comparing with ground based AERONET data over 2005-2017 covering more than 1000 sites globally.  The work is extensive and elaborative. I appreciate authors efforts to provide such detailed information and presentation which may be useful for future workers who need to use MODIS aerosol measurements.

Their conclusions can be summarized as daily satellite data is better than weekly data and monthly data. The paper also provides ways of how to manage monthly data since there is both temporal and spatial limitation due to the nature of satellite observations. 

However, I think by using AERONET data as the basis is also a problem. Since AERONET data are also limited in spatial, temporal, and accuracy.  The conclusion about the quality ranking of hourly, daily, and monthly can be understood and derived perhaps by some simpler and straight methods.

Some typos exist for example at

line 400 Wei, Li [44] ??

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx