Dear authors,

Thank you for submitting your revised manuscript. I have now heard back from two of the reviewers of the original version. While both reviewers agree that the manuscript is improved, we find that further revisions are necessary before it is suitable to publication in AMT. I would encourage you to submit a further revised version of the study, addressing the following comments:

Reviewer 1:

The revised manuscript has made significant improvements. A few remaining questions for clarification:

Line 86-90: Are cloud properties retrieved on its native MODIS/VIIRS resolution and then converted into CERES footprint or the radiances are converted into CERES footprint resolution first and then applied with cloud retrieval algorithm? If the latter, mention radiance then cloud property.

Line 164-167: The FM3 and FM5 response to LW seems to explain the NPP LW radiance smaller compared to Aqua LW radiance. But SW responses of FM3 and FM5 seem contradictory to the observed radiance differences.

Figure 6: Based on my understanding, VIIRS-like cloud would only affect the choice of ADM, not the generation of NB radiance for NPP footprint size and the conversion from NB to BB radiance. If this is the case, the right branch of the procedure for NPP with VIIRS-like retrieval is not accurate. Do step one and step two affected by cloud?

Reviewer 2:

Overall comment:
The manuscript has much improved from the first version I reviewed. However I still have a few concerns that I would like the authors to address.

Major comments:
Beginning of Section 2: you guys still haven’t explained why you didn’t use the scattering angle differences in your pixel matching algorithm. It’s the scattering angle that really matters when comparing VIIRS and MODIS. When you do individual angle thresholds the differences can add up constructively and you end up with a large scattering angle difference, larger than you’d want to have in any case.
Please evaluate the scattering angle differences between MODIS and VIIRS in your selected footprints. Differences in individual zenith/azimuth angles can add up constructively and create significant scattering angle differences.

Figure 3 appears to contain an artifact such that it makes statements made in the paper invalid. Either correct the figure or explain the discrepancy / source of banding.
Figure 3: Can you please explain the source of the dark blue band of lower insolation around 60 deg. N? That almost looks like some kind of an artifact. Based on that banding artifact, your discussion on Page 8 that TOA fluxes track insolation differences is not valid. The TOA fluxes do not display any banding at 60 deg. N.

Minor comments:
Line 14: “… the imagers sharing the spacecraft with CERES-NPP… .”
Line 91: reference for the IGBP dataset?
Line 98: Do you mean “VIIRS”? If not, please expand what VIRS stands for.
Line 200: “… to be applicable to the VIIRS observations….”
Line 207: “…. was generated …. “
Line 245: “ … except for small differences … “
Line 365: “… the imagers sharing the spacecraft … “ They are mounted together, not flying in a constellation
Line 386: “ … five years ….” no dash

Please let me know if you have any questions.

Best wishes,

Andrew

Dear authors,

Thank you for your revision. After some follow-up discussion with a previous Reviewer, they are concerned that you did not explain the high-latitude artefact in Figure 3 sufficiently. Following on from their initial review comments, could you please check whether this is a plotting error, and if it is not, go in to more detail about this artefact? This is a major sticking point of theirs before they can assess the revision in detail, as they feel that if the artefact cannot be explained then the later analysis may not hold up.

Best wishes,

Andrew

Dear authors,

I have had another offline discussion with the reviewer in which they provided me with a full report; after talking it through they are satisfied with the high-latitude issue we went through before, but have one more point which requires explanation.

The major concern is as follows: line 200 states that a 0.003 (or 1.02%) difference is caused by calibration, with a reference to Figure 1, although this does not seem to be directly supported by the contents of Figure 1.
The suggestion is to add some statistics to the panels of Figure 1 (and Figure 2 as well). For example, mean offset, RMSE, and fit coefficients (i.e. y=ax+b). Presumably b will come out to 0 and a will come out to 1.0102 or close, or the mean bias will be around 0.003, which will then support your argument. (Or possibly the other direction dependent on whether Aqua or NPP is taken as the reference.) If not, then further discussion or amending the statement is required.

Minor comments:
line 1: please expand NPP, first use of the acronym
line 25: “...when compared...”
line 41: “...data product crucial to...”
line 147: “ … 2 degrees and 3 degrees respectively. ...”
line 188: “ … for April 2013. Insolation for NPP...”
line 189: “… is greater than...”
line 192: strike “solar”. The word insolation means “solar radiation”. You shouldn’t double up.
Line 196: “… We then compare...”
line 212: “ … mainly affect cloud detections...”
line 187, 213, 532, and everywhere else: remove “solar” when referring to “insolation”
Fig 3. caption: should read “… mean insolation…”

Best wishes,

Andrew

Dear authors,

I am pleased to accept this latest version of your manuscript for publication in AMT, as I judge that you have addressed reviewer comments and concerns well. Thank you for your contribution to the journal.

Best wishes,

Andrew