The authors evaluate the diurnal cycle characteristics of several satellite/gauge data sets and reanalysis data sets over three continental-scale regions in the Northern Hemisphere against gauge data, and more generally against each other globally.  The computations and conclusions are complete and largely convincing, but the text is in need of more consistency and completeness.  I have not commented on simple grammatical issues, but someone needs to do a sweep through the entire manuscript to address this problem.

High-level:

1. Satellite data set descriptions: Section 2.2 needs to be reworked to give the same information for each data set – pointing at Table 2 from the start can shorten the descriptions a bit?  As well, the descriptions are not necessarily accurate:

• IMERG now starts in 1998, even if the dataset started in 2000 at the time that you downloaded it. Also, V07 propagation vectors are computed from multiple reanalysis and forecast fields (but not winds, and certainly not IR).  “IR estimates” is really “estimates based on the global collection of GEO-IR Tb’s”.
• TMPA should be described in the past tense, since it is no longer active. Here I’d drop sensor and GEO satellites, since it’s in Table 2.  Also, I think you should justify using the more-approximate TMPA-RT instead of the research-oriented TMPA, which includes gauge, as the other satellite products do.  I would strongly suggest TMPA.
• If you’re going to go into such detail about morphing for CMORPH, you need a comment that this approach was the starting point for IMERG’s morphing, although they have since diverged.
• For MSWEP, I don’t think GridSat provides actual precip estimates. What is the algorithm?  In L.298, to what is CDF matching applied?  It is not unique for MSWEP to apply undercatch correction – TMPA, IMERG, and GPCP all do that – but I believe the particular scheme used is unique.

1. Table 2: Several corrections are needed:

• IMERG V07 Final Run.
• IMERG starts in 1998 (and state if you only used it starting in 2000).
• The first 4 satellite products all have the full complement of GEO-IR satellites because they all use the CPC 4 km Merged IR dataset: GMS/MTSat/Himawari, GOES-E, GOES-W, Meteosat.

1. TMPA-RT, not TRMM: There are multiple TRMM products, so you can’t use “TRMM” as short-hand for the product.
2. Acronyms vs. written-out names: The manuscript varies between using acronyms, some of which I’m not sure were defined, and written-out phrases.  Please take a pass through the text, tables, and figures and enforce consistency.
3. Scaling to daily totals and subdaily detail: On L.934-936 you haven’t convinced me that scaling to daily totals will somehow warp the subdaily detail, including the diurnal cycle.
4. Data availability: The editor gets to weigh in here on what the HESS practice is.  My experience is that you should give pointers to the actual datasets that you used.  Giving general addresses like https://cdc.dwd.de/ leave the reader floundering around chasing links and potentially grabbing something other than the data set you actually used.

Additional:

1. significant: “Significant” is used qualitatively, and should be changed to some synonym like “large” or “major”.
2. “comprehensive”: I think you should drop using this word and let the reader decide (perhaps) that this study is comprehensive.  In particular, the initial wording makes it sound like all Northern Hemisphere land is covered, when it’s really CONUS, western Europe, and East Asia.
3. L.50-51: the phrase “necessary moisture” leaves hanging the question “necessary for what?”
4. L.53: I’d change “low-level inversion” to “inversion at the top of the PBL” since you’ve introduced the PBL.
5. L.69 and elsewhere: “precipitation type” is a pretty vague term, subject to reader mis-interpretation (like “intense” vs. “showers” or “solid” vs. “liquid”). It’s more like “storm system type” or something similar.
6. L.102: I’d drop the very incomplete “GPM-based” and “microwave-morphing-based” descriptors and let the following paragraphs introduce the differences.
7. L.106: The more-complete statement is “… (e.g., radiative emission by liquid at lower levels, radiative scattering by ice at upper levels, and cloud-top temperatures) rather …”
8. L.175-176: You need to be more explicit here; presumably this is a volume-preserving averaging, particularly for splitting the CMFD cells across analysis grid boxes? What happens at the edges of data coverage?  Is there a minimum number of stations (or CMFD grid boxes) to register a value?
9. L.276: I would suggest “Project (GPCP) Pentad (5-day) Product is used … technique in two steps: first using”
10. L.290-290: I think “thereby improving …” is extraneous.
11. L.518: I think something is missing in “Second, GSMAP is applied exclusively over land …”
12. L.520: I’m pretty sure that “temporally fixed monthly ratio” will lose most readers. Please slow down and be a little more descriptive.
13. L.633: I think “sub-daily temporal disaggregation” will lose readers unless you introduce the concept in the description back in Section 2.2.
14. L.655: “… observation is depicted in (e)-(h).”
15. L.659-659: This also applies to IR, particularly for TMPA.
16. L.668-669: The current phrasing misleadingly implies that there’s some kind of algorithmic linkage among the reanalyses. You’re just giving the various offsets.
17. L.698-699: You’re listing rows, not columns.
18. L.700: “backward and forward masking in Fig. 6” is really confusing. I think you just mean “the masking depicted in Fig. 6”.
19. L.761-2: I think “In CMORPH” should come after “precipitation maximum” for clarity.
20. L.929: Here as well, IR contributes to the delay, not just PMW ice scattering.
21. L.659-659: This also applies to IR, particularly for TMPA.
22. L.668-669: The current phrasing misleadingly implies that there’s some kind of algorithmic linkage among the reanalyses. You’re just giving the various offsets.
23. L.698-699: You’re listing rows, not columns.
24. L.700: “backward and forward masking in Fig. 6” is really confusing. I think you just mean “the masking depicted in Fig. 6”.
25. L.761-2: I think “In CMORPH” should come after “precipitation maximum” for clarity.
26. L.929: Here as well, IR contributes to the delay, not just PMW ice scattering.

Please find my review comments in the attached document.