A Method to Downscale Satellite Microwave Land-Surface Temperature
Round 1
Reviewer 1 Report
The authors present a method to downscale low-resolution (25x25 km) land surface temperature estimates from MW satellite instruments using high-resolution ancillary information.
Overall the manuscript is well structured and written and the context is easy accessible for readers. The described approach shows some new developments and has the potential to be useful for many applications. There are some points, which could be improved before publication.
Major review points
1. I really like ready this manuscript. The context is well explained, the figures are clear and meaningful and the approach is interesting. However, at the point where I expected an evaluation of the results, the authors show one single result for a random scene. That was quite disappointing, as it is not clear how the approach performs in the real world. Line 433: “However a systematic evaluation of the performance of the MW-derived T is outside of the scope of this study…” Why? I understand the lack of global reference data, but there are at least some data from ground stations and there are data from other satellite data records, which could be used for comparisons.
I would expect at least some more examples, like shown in Fig. 10. This figure shows a scene at 05:00, where the impact of clouds is less than in the afternoon. Why are no further examples presented (at least one more figure showing an afternoon scene with cloud impact and some more summarized in a table) to get a feeling if the shown numbers are robust. At the moment the reader doesn’t know how the presented approach performs and if it is applicable for specific applications. The uncertainty for SEVIRI LST estimates alone is about ± 1.5 K and just from the one shown scene I guess the downscaling adds another 1.6 K uncertainty. If not, please correct me.
This could be a really valuable study, but it could be much better by including an adequate way to estimate the quality of this approach.
Minor review points
Line 161: Why are ERA5 data used? Because they are easier to get? There are several SEVIRI-based cloud products available (e.g., CM SAF CLAAS or COMET data records).
Line 182: This refers also to my major point. Why just one scene? Why not at least one more in the afternoon, which shows the impact of clouds better?
Line 292: Why not more sophisticated from the beginning. What are the reasons to do it in a ‘simple way’? Are much better results expected by using methods that are more sophisticated?
Line 389: Please check sentence grammar.
Line 487: “The downscaling could not be tested under cloudy conditions as no high resolution satellite product is yet available for cloudy scenes.” Please explain!
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
See attached file
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
The paper presents a method to downscale temperature retrievals aimed for CDRs. It is an honest and straightforward paper just reporting a method that the authors have developed. The manuscript is clearly written and organised but the results are not ground-shaking. However, they are interesting enough to merit publication in RS. The method is simple and can be easily replicated, and the product can be of the interest of the community. There are a few typos but those can be certainly corrected in the editing stage.
Author Response
The authors are grateful for this review, that summarizes well the usefulness of this method for the community. Beside the novel methodology the paper also provide insights on the factors impacting the T (clouds, temperature amplitudes) for future downscaling application developments. Hopefully most of the typos will have been corrected in this updated version.
Round 2
Reviewer 2 Report
The authors answered all my concerns, and it can be accepted for publication.
