Response to reviewer's comments:

The authors are thankful to the reviewer for constructive comments and suggestions. The authors’ responses are in italics.

Review of the manuscript “Path-CVP (pCVP) – Polarimetric radar data snapshot along the predefined path based on Columnar Vertical Profiles” submitted to AMT by Bukovcic and Krause.

This manuscript presents a novel display technique for radar data called path-CVP (pCVP). This is a way to observe the current status of radar polarimetric variables along a specified path which enable one to see microphysical processes and precipitation characteristics at desired locations. The authors describe the how the method is applied and present a number of examples that show the application and describe the benefits of the novel approach. The advantages of such a technique are clearly discussed, as well as the limitations.

The document is logically structured, well written, and the content is relevant and useful for practical purposes. I have only a few minor comments and suggestions.

Thank you for the constructive comments and suggestions.

Comments:

Lines 35-36: “range-height indicator (RHI) scan based QVP (R-QVPs, Allabakash et al. 2019, RSVP Blanke et al. 2023)

Blanke, A., A. J. Heymsfield, M. Moser and S. Trömel, 2023: Evaluation of polarimetric ice microphysical retrievals with OLYMPEX campaign data, Atmos. Meas. Tech., 16(8), 2089–2106, https://doi.org/10.5194/amt-16-2089-2023.

The reference is added to the text and the References section. In addition, a short description of the method is also included in the text as: “The RHI sector vertical profile (RSVP; Blanke et al., 2023) represents the average of the RHI azimuthal sectors, each 22 degrees wide, providing noise-reduced quasi-vertical profiles of polarimetric variables. This technique can track the research aircraft within the sector covered by the RHIs and enable joint analysis with fixed, vertically pointing ground-based devices (e.g., micro rain radars).”

Line 43 (and others): The authors mention several times the accuracy being proportional to N^0.5 but there is no explanation for this. Please provide a brief explanation.

The references Doviak and Zrnic (1993), Melnikov (2004) and Ryzhkov et al. (2016) are added to the text - the concept is introduced in former, and its application described in latter. The authors also added “see Section 2 for additional details”. The brief explanation about the N^0.5 origin is added to the Section 2.

Lines 129-130: “where the standard deviation of all polarimetric variables is directly proportional to λ^1/2 (λ is the radar wavelength), and inversely proportional to N^1/2”

I think here a reference is necessary (concerning the wavelength), since this is not a trivial concept.

The reviewer may have overlooked the references at the end of the Section 2, where the wavelength concept is introduced. The authors stated: “The averaging reduces statistical errors of the radar estimates, where the standard deviation of all polarimetric variables is directly proportional to λ^1/2 (λ is the radar wavelength), and inversely proportional to N^1/2, where N is the total number of points used for averaging (within a 400 m cylinder slice). Hence, the statistical accuracy of all radar variables is smaller for shorter wavelengths and for a greater number of points used for averaging. For more details, see Ryzhkov et al. (2016) and Ryzhkov and Zrnic (2019)”. The authors also added the references Doviak and Zrnic 1993, and Melnikov 2004, and a brief description about the concept as follows: “For example, the standard deviations of Z, ZDR, and ρ_hv estimates are given by the following equations (Doviak and Zrnic 1993; Melnikov 2004; Ryzhkov and Zrnic 2019):

SD(Z) = 3.24/(σ_vnM)^0.5    (dB),                                                                                           (1)

SD(ZDR) = 4.62(1-ρ_hv²)^0.5/(σ_vnM)^0.5    (dB),                                                                      (2)

SD(ρ_hv) = 0.53(1-ρ_hv²)/ (σ_vnM)^0.5,                                                                                       (3)

where σ_vn = 4σ_vT_s/λ is the normalized spectrum width, σ_v is the Doppler spectrum width (in m/sec), λ is the wavelength (in m), T_s is the pulse repetition period (in sec), and M is the number of samples, and the equations are valid for 0.04 < σ_vn < 0.60. In the S-band WSR-88D radar operations, Ts = 3.1 x 10^-3 (long PRT) and M = 16 (surveillance scan). The typical value of σ_v = 3 m/s produces SD(ZDR) = 0.68 in the melting layer (for ρ_hv = 0.94). In the case of horizontally uniform melting layer, the 360 degrees azimuthal averaging (N=360) reduces the statistical error of ZDR by a factor of 360^0.5, producing a standard deviation value of 0.036 dB for ZDR (Ryzhkov et al. 2016)”.

Line 155: suggest the change “...in the lower 0.5 km AGL, the reduction in ρhv throughout the column up to 3 km AGL…”

Incorporated.

Figures 2b, 3b and 4b have the colorbar reversed (from high to low values). Better to present this with increasing values.

The authors reversed the colorbar in 2b, 3b, and 4b to the increasing values.

Line 198: “… from the WSR-88D KTLX”

Done.

line 208: “...layer of increased values of Z…” (remove “the”)

The “the” is removed.

Lines 244-245: “An example of using pCVP for this purpose is in Fig. 8, …” This seems a little misleading because Fig. 8 shows a road map.

The authors restructured the sentence as: “An example of the path portion of pCVP is in Fig. 8, highlighted (orange) on the map as the route from an arbitrary point 1 (e.g., Dunkirk, NY) to point 2 (the 490 exit to Rochester) along highway I-90”.

Line 257: “...is an artefact…”

Done.

Line 295: “...very few direct wind speed measurements on the ground along the tornado path.”

Incorporated.

Lines 355-356: “The local microphysical processes are more present in the 2 km version, ” The formulation “more present” is a bit odd…

I suggest “The local microphysical processes are clearly visible…”

Done.

In all the examples the radar moments shown are Z, Zdr and ρhv. Could the authors comment on the possibility of using pCVP for displaying Kdp?

KDP is derived from PhiDP, hence, not directly measured by radar. That is the sole reason for not displaying KDP using pCVP technique. The following paragraph is added to the section 4:

“The authors only displayed the directly measured radar moments Z, ZDR, and ρhv in the pCVP technique throughout the paper. The user can also display other radar variables (direct or derived, e.g., spectrum width or KDP) similarly to Z, ZDR, and ρhv, including all quantitative retrievals. For example, the retrieved visibility and snowfall rate from pCVP KDP and Z, σ_e(KDP, Z), and S(KDP, Z) (Bukovčić et al. 2020; 2021), can be advantageous to road crews along the LES routes (Fig. 9), highlighting the route portions with low visibility and potentially high accumulations”.

What about quantitative retrievals, can these be applied to pCVPs and what is the expected outcome of doing it?

Quantitative retrievals can be applied to pCVPs. The authors added the explanation to the text – see the response to the previous reviewer’s question.

The authors are thankful to the reviewer for constructive comments and suggestions. The authors’ responses are in italics.

This manuscript presents a novel radar data processing and visualization technique referred to as Path-CVP. The method enables the depiction of vertical distributions of radar-observed parameters along arbitrary paths. It holds considerable potential for comparison with aircraft and radiosonde observations, as well as for operational applications in transportation systems that follow fixed routes, such as aviation, highways, and railways.

The authors effectively demonstrate the utility of the proposed method through multiple well-chosen and accessible case studies. The manuscript is clearly structured, and the content is highly valuable. There are only minor comments.

Thank you for the constructive comments and suggestions.

Specific Comments

• L93: Does the Cressman interpolation utilize distance-weighted coefficients, or is it implemented in a linear fashion? Clarification would be appreciated.

The authors added the reference for the Cressman technique, and the following text: “All of the data within the locations’ horizontal radius is selected when interpolating the data to the CVP height locations. The Cressman weights for each data point are computed using distance on the vertical axis, with a maximum (vertical) distance of 200 m. The data are then combined as a weighted sum.”

Reference: https://journals.ametsoc.org/view/journals/mwre/87/10/1520-0493_1959_087_0367_aooas_2_0_co_2.xml

• L129–131: Are there additional dependencies on parameters such as Doppler spectrum width, pulse repetition time (PRT), or ρhv? As the authors note in the Discussion, selecting an appropriate CVP radius is crucial for balancing accuracy and the spatial scale of the observed phenomena. Therefore, I suggest that the manuscript more clearly describe the factors that influence the accuracy of pCVP-derived variables—even if these are already discussed in previous literature. Including a concise summary within this manuscript would improve clarity and completeness.　

The authors added the equations for standard deviations of Z, ZDR, and ρhv (SD(Z), SD(ZDR), and SD(ρhv)), a clarification on the dependencies and brief explanations, and appropriate references, as follows:

“For example, the standard deviations of Z, ZDR, and ρ_hv estimates are given by the following equations (Doviak and Zrnic 1993; Melnikov 2004; Ryzhkov and Zrnic 2019):

SD(Z) = 3.24/(σ_vnM)^0.5    (dB),                                                                                      (1)

SD(ZDR) = 4.62(1-ρ_hv²)^0.5/(σ_vnM)^0.5    (dB),                                                                 (2)

SD(ρ_hv) = 0.53(1-ρ_hv²)/ (σ_vnM)^0.5,                                                                                  (3)

where σ_vn = 4σ_vT_s/λ is the normalized spectrum width, σ_v is the Doppler spectrum width (in m/sec), λ is the wavelength (in m), T_s is the pulse repetition period (in sec), and M is the number of samples, and the equations are valid for 0.04 < σ_vn < 0.60. In the S-band WSR-88D radar operations, Ts = 3.1 x 10^-3 (long PRT) and M = 16 (surveillance scan). The typical value of σ_v = 3 m/s produces SD(ZDR) = 0.68 in the melting layer (for ρ_hv = 0.94). In the case of horizontally uniform melting layer, the 360 degrees azimuthal averaging (N=360) reduces the statistical error of ZDR by a factor of 360^0.5, producing a standard deviation value of 0.036 dB for ZDR (Ryzhkov et al. 2016).”

Additionally, in Section 3.1 and onward, CVP radii of 2, 5, and 10 km are used. Were these values selected through trial and error, considering the trade-off between the scale of the phenomena and the visibility of their signatures?

Yes, these were selected via considering the trade-off between the scale of the phenomena and the visibility of their signatures. The explanation is provided in the Discussion section, referring to Figures 7, 13, and 14.

• Figure 4: Is the value “−99.00” used to indicate undefined or missing data? If so, this should be explicitly stated in the figure caption or legend. The same clarification applies to other ZDR figures.

The authors added the following sentence in the Figure 3 caption (the first Figure with -99.00 value in the colorbar): “The value -99.00 herein indicates the missing data.”

• Figure 7: Does “Heading 0 deg.” correspond to true north? This may not be immediately clear to all readers and could benefit from clarification.

Yes, we used true north. The heading information used in the plots is taken from the airport runways’ orientation, as noted in publicly available pilot information. Pilot information is denoted in magnetic north, but we have used true north for convenience. Over the distance of the runway, this assumption is within tolerance. The following is added to the Figure 7 caption: “…, where heading 0 degrees represents true north, and 180 degrees true south”.

• L263–264: I agree that negative ZDR values are indicative of conical graupel in LES (lake-effect snow) environments, as discussed by the authors. However, are snowfall cases with positive ZDR not present in this region, or are they considered less relevant? From an operational perspective, should road maintenance authorities focus only on cases exhibiting negative ZDR?

The cases with positive ZDR may be present in the region, usually when a large scale processes are dominant with the greater storm depths (2-2.5 km AGL for LES vs. ~7-8 km AGL for synoptic scale storms), producing so-called synoptic snow. Those situations may be less detrimental regarding the road conditions. The authors presented the case which is more challenging for the radar observations, due to relatively small vertical extent of the LES processes. In addition, some large scale storms with specific conditions (e.g., Nor’easters), may also have a detrimental effect in the wider area.   

• L308: The phrase “a minimum in…” seems imprecise in this context. Could the authors clarify what specific range or domain this minimum refers to? Alternatively, would it be more appropriate to describe the feature using terms such as “low values” or “a decrease in the variable” ?

The authors replaced “a minimum in…” with “a decrease in”.

• L313: Is “BEWR” a typographical error for “BWER” (Bounded Weak Echo Region)?

The authors corrected the typo.

• L381–383: I did not fully understand why RD-QVP is also effective at capturing higher altitudes, as shown in Figures 10b and 10c. My understanding is that RD-QVP differs from QVP in that it incorporates lower elevation angles, allowing vertical profiles to include near-surface layers even close to the radar. Could the authors please provide additional explanation on how RD-QVP enables observation of higher altitudes in these examples? Is this related to the inclusion of data from farther horizontal distances from the radar?

In this context, both RD-QVP and QVP provide the similar information at higher altitudes - we are referring to RD-QVP because of the lower-level elevation angles incorporation, as reviewer noted. pCVP is an offspring of a modified CVP, and it may have higher variable standard deviations at the altitudes above a certain level in the radar vicinity (due to 400 m vertical slice extent used for averaging – hence, a smaller number of averaging points). For the same radius size, QVPs/RD-QVPs will have a higher variable accuracy (due 360 deg averaging) compared to modified CVPs. The authors modified the corresponding paragraph as follows: “One way to mitigate the issue in radar proximity is to use (for example) a 100% larger-radius RDQVP estimate along the affected portion of the route. A sector RDQVP may be utilized for the off-centre locations, with the complete RDQVP estimate for the few closest points, to replace the pCVP processing in such situations. In the radar proximity and for the same averaging radius, RD-QVP variable estimates have lower standard deviations compared to pCVPs, especially at higher altitudes (due to fewer points in pCVPs’ middle-point average from a 400 m vertical slice; see Fig. 1a) ”.   

• While the manuscript employs Cressman interpolation, would the use of a median filter also be effective in reducing noise when applying smaller CVP radii?

The authors added the following to the Discussion: “In addition, the use of a median filter for the data within the CVP radius for each height would likely remove much of the vertical variation found within the CVP. Smoother data in the CVP plots would make the identification of melting and freezing layers more difficult. For smaller CVP radii, it might be a reasonable tradeoff if the current Cressman method creates unwanted levels of noise.”