The paper presents a thorough characterization of the observations mode of the EarthCARE CPR. 

In particular it investigates the impact of the different modes on two aspect: 1) Doppler velocity errors , 2) occurrence of second-trip echoes. The paper is very important because it provides a description of the pros and cons associated to the different modes that have been,  are and will be used at different latitudes and in different mission periods.

I have some major comments that I would like the authors address to make the paper more clear and coincise.

1) The authors mention several times that they compute the standard deviation of Doppler velocity in several part of the paper. But this is left vague. The authors need to say across what ensemble such standard deviation is computed. This is very relevant in relation to the impact that what the authors refer to as "natural variability" std means. For instance if I understand correctly in Fig.4 the second row represent the std computed for each pixel in lat-height including all data with reflectivity between -20 dBZ and 0 dBZ (for 1 km Doppler velocities?). In that case the impact of the random component will be marginal (I expect) because in such a wide range of reflectivities there will be a large natural variability. In fact the authors do not present any quantitative intercomparison between the standar deviations of the different modes derived from Fig.4 (am I correct?). In fact I do not see the reason of plotting Fig.4 apart for documenting the behaviour of the mean Doppler velocity as a function of latitude and height (which by the way appears quite puzzling to me, why there are constant updrafts in the upper levels?).  Also statement like in the conclusions (Line 395-397) are not very general since they are applicable to a given ensemble with the relative impact of the natural variability being very different from ensemble to ensemble. 

2) Fig3 and Fig.5: Instead of plotting Fig.3 like it is it would be much better to merge Fig.3 in the top panel of Fig.5 (accounting for the integration length of course). For Fig.5 it would be good again to specify across which domani the std is taken. The intercomparison between the two curves will indeed explain what is the natural variability component (which shoul be kind of independent from the PRF but should depend on he integration length). It would be good to see some attempt to separate the  impact of the random component (to verify whether it follows the expected behaviour).

3) I am not so sure Fig.6 adds much on what already shown in Fig.5.

4) terminology: instead of using "mirror"  frequency and similar expression I would use "second-trip echoes" frequency because the origin can actually be different (not necessarily mirror images)

5) "Cloud overlap risk" in Sect 3.5 : I would highlgiht caveats to the proposed definition. Indeed using a product of probabilities is correct if the two events are independent (which I would not say is the case).

Minor comments:

1)  Fig 9, last row. Cloud and mirror overlap = is this the same as cloud overlap risk? Then use the same name

2) Eq.3: why are you introducing C (C=1 in Table 1) I would delete C from the Eqaution and the TAble.

3) Table 1: sigma_sm=: in the third colum there is an expression that is correct for sigma_sm not the square of it. Also not clear to me what values of V_s is used

4) Line 188: this is due to pulse strecthing as documented in Xu et al., 2025 https://doi.org/10.5194/egusphere-2025-5421 

This paper summarizes the Doppler velocity accuracy, cloud detection altitude, and contamination with second-trip echoes for the EarthCARE/CPR operational mode, providing useful information for users of EarthCARE/CPR data.

On the other hand, a quantitative evaluation of the observed values against theoretical values is required. Specifically, while Figure 3 shows the relationship between theoretical Doppler velocity error and PRF for each radar reflectivity value, a similar plot should be created using actual observation data, and the differences between them should be discussed.

As for other aspects, the revisions listed below should be implemented.

1. 10: The authors should mention that CPR is nadir looking radar
2. 13: “mirror images” should be “2^nd trip echo from mirror images”
3. 65: The authors should mention that CPR is w-band radar and nadir pointing.
4. 68: The authors should add Kobayashi et al. 2002 and Kobayashi et al. 2003.

Kobayashi, S., Kumagai H. , and Kuroiwa H. , 2002: A proposal of pulse-pair operation on a spaceborne cloud-profiling radar in the W band. J. Atmos. Oceanic Technol., 19 , 1294–1306.

Kobayashi, S., H. Kumagai, and T. Iguchi, 2003: Accuracy Evaluation of Doppler Velocity on a Spaceborne Weather Radar through a Random Signal Simulation. J. Atmos. Oceanic Technol., 20, 944–949, 

1. 96: The authors should concretely explain about the MIX-mode, in terms of original planning of CPR operation in one orbit (operating 20-km mode in low latitude and switching to 16-km mode in high latitude).
2. 100: “the CPR orbit altitude” should be “the nominal CPR orbit altitude”
3. 140: “the CPR pulse wavelength” should be “the wavelength of CPR”

Table 1 NB: the authors should explain the “data frame”.

Table 1 theta (definition box): “0.095º from CPR specification”

1. 168: “30 or higher”; please add the explanation what does it means.
2. 293: why the STD (10 km) at -20 dBZ in Fig. 6 is larger than that in Fig. 5?

Fig. 5 (g) (h): the y-axis label should be “Count”.