General comment:

This study provides new results on the diurnal variability of ozone; the data figures are very good based on the two nearly co-located ground-based microwave radiometer instruments.  The authors emphasize monthly and seasonal diurnal ozone scaling factors, but they also report finding significant sub-monthly diurnal ozone variations during northern hemisphere winter.  Even so, the original stated goal of the study is to generate a refined diurnal model for the purpose of merging multiple datasets for analyses of long-term ozone time series and for comparisons with model ozone time series.  Thus, I would argue that they are showing that it is best to avoid winter hemisphere data for that purpose.  Winter anomalies in temperature may be more important than those of NOx, but it is not easy to assess that prospect because of the low vertical resolution of the MW ozone profiles.

Specific comments:

Line 73—It would be helpful to learn at this point why there was an overestimate of the ozone diurnal cycle previously.  Also define GROMOS here.

Line 134—Here the authors give two specific overpass times for the MLS measurements, while on line 179 they indicate a more general range of time.  Which is correct?

Line 213—What is the source of the noisy appearance? Gravity waves, perhaps?

This paper uses the harmonized timeseries from two ground-based microwave radiometers along with output from several models to investigate the diurnal cycle of ozone in the stratosphere and mesosphere over Switzerland.  An interesting finding of this paper is that the amplitude of the ozone diurnal cycle has short-term variability.  The paper is well-organized and provides useful data on the diurnal cycle of ozone.  I list some general and specific comments below.

General comments:  

Section 3.3: Tides are mentioned in the abstract and conclusion but could be discussed here as well to more strongly tie this section to the abstract and conclusions. It might also be helpful to show the timeseries for one of the observations (similar to Fig. 8a) but with the daily mean overplotted, to help distinguish changes in the diurnal cycle from changes in daily mean ozone, since both seem to be happening in the time series. 

Specific comments:

Line 125: Are the day and night averaging kernels very different, and if so, how does this affect the results?

Line 151: Is it because the profile is normalized that you cannot apply the kernel, or because it is a monthly average?

Line 168: Why was a free-running perpetual year simulation used for this study instead of nudging?

Section 2.2.2: Please provide some information on the chemical mechanism in BASCOE

Line 261: This might be easier to see if there was a plot of just the magnitude of the diurnal cycles.