General comments:

In general, this study evaluated sub-daily and daily precipitation data from a WRF simulation over CONUS against NCEP and PRISM datasets. This paper is well-written and logically flows well. The findings and caveats in WRF simulations are comparable to earlier studies. I have one major comment and several minor comments and hope the authors can address them.

• The second paragraph in Summary and Discussion is way too dense and hard to read. I would recommend the authors split it into two or even three paragraphs and reorganize it to increase readability.

Specific comments:

• L65: Please simply use the regional climate model instead of RCM since this is the only time that the abbreviation was used in this manuscript.
• Figure 1: can the authors please 1) highlight the domain for this study, 2) add a topography layer, and 3) add the NCA region boundaries and names to this map for better illustration purposes?
• L163: The selection of the 0.25mm threshold seems random. Please justify it.
• L193-198: can the authors please explain why WRF improves less on capturing extreme precipitation values in the NGP and SGP regions?
• L201: Please justify the selection of 1mm and add references if any.
• L233: Should it be “For example, it shows wet biases during winter and spring, but a mix of wet and dry biases during summer and fall?”

Review of “Assessment of WRF (v 4.2.1) dynamically downscaled precipitation on subdaily and daily timescales over CONUS” by Srivastavia et al, 2023.  Submitted to GMD. 

This paper compares the results from a 12km WRF dynamically downscaled climate experiment driven with ERA5 boundary conditions with ERA5 itself, Stage-IV precipitation (for 3-hourly fields) and PRISM (for daily fields).  This paper examines the diurnal cycle, seasonal cycle, precipitation frequency, and precipitation intensity.  This paper was well written and easy to follow. This paper should be accepted for publication after minor revisions suggested below.

Overall Comments:

• I suggest the authors also create a daily product from the Stage-IV 3-hourly product to compare with the PRISM data, ERA5, and WRF-ERA5 or provide a discussion of how Stage-IV data compares with PRISM from other papers. 
• It is unclear what the authors mean by “operational purposes” in their paper.  Perhaps WRF climate simulations would be used to help inform stakeholders about future precipitation/water availability- but such climate simulations that these WRF simulations represent are not “operational” in any traditional sense. 
• It was unclear to me if you included the zeros in your mean precipitation plots for 3-hrly precipitation and daily precipitation means.  That should be made clear because that does influence if the plots are really “average precipitation intensity” or just “average precipitation” - you make some statements about precipitation intensity vs frequency, but if you average with the zeros you don’t really capture how much it rains when it rains.
• The utility of the paper to end-users of this data would be improved if discussions of how much bias is tolerable for a model to be useful and if the WRF/ERA5 data are within that range. 
• Percent biases (maybe added to supplemental) could be really informative
• Mean plots are often too "blue" . I suggest using a more-non linear scale to highlight more differences.

Minor comments:

L28: update to “processes facilitates the study of future changes”

L34-35: sentence a little unclear as written perhaps: “biases are generally not consistent across the variables, regions, and seasons of interest” or just “biases vary with variable, region, and season.

L46: no the needed in front of historical so “we evaluate historical”

L83: How might the inclusion of urban surfaces influence precipitation in these simulations?

L89-94: I think this would read better if there was no “the” in front of Stage VI. 

Section 2.1.1 - How well do Stage IV and PRISM perform in mountains?

Section 2.1.2 - I think a small statement about what “solar noon” means would help contextualize this work for people who do not work with diurnal cycle data. 

L110: What do you mean “variability is assumed to be the same in 20 year period”?  Is this saying you picked a 20-year period similar to the SageIV for the PRISM data because then internal variability wouldn’t play a role?  I think this needs to be expanded slightly for clarity.

L138: You say here ERA5 against two different reference datasets - but so far you have only done Stage IV so that is confusing. 

L144: You mention MCS and the Great Plains, but why is there too much precip in the southeast?

Section 3.2 - I am curious how stage IV and PRISM compare (see note at top).

L159-160: sentence a little confusing I think “averaged precipitation peak are improved in the downscaled WRF simulations compared to ERA5” is a bit more clear. 

Figure 6 - what is the giant red dot in the Stage IV product?  I think these figures would be improved if you masked out the red dot and you re-scaled precipitation … all this blue makes things hard to see. 

L174-177: Starting with “The maximum values …” the comments about CMORPH seem really random and an aside given that you are not talking about CMORPH …

Fig 7: Do you include zeros in this average?  If so it might also be good to look at the precipitation rate when it rains.

Fig 11/12 - should this be PRISM in the left and column?