Recommendation: Major

General comments

Overall, this is a well-written manuscript which provides an analysis of upper mantle anisotropy along the CIFALPS2 temporary array. Investigating upper mantle anisotropy through shear wave splitting is crucial for comparison with other seismological observables reported for the CIFALPS2 array. The strength of this work lies in its utilization of both conventional shear wave splitting technique (such as tangential energy minimization) and the splitting intensity method. This study adds one more piece of knowledge to mantle deformation beneath the Western Alps. This is quite good organized work with good figures.

Here are my comments to enhance the manuscript and address some unclear points:

Line 71

Which criteria were considered by selecting the events? It could be good to mention the criteria here.

Lines 80-83

In addition to plotting all individual splitting measurements at the piercing point of incident SKS ray at 150 km depth, it would be beneficial to plot the back-azimuthal variations of the collected FPD and DT parameters. This could help in interpreting potential complex anisotropy beneath the station and surrounding. These plots can be generated separately for each station or for groups of stations within a 100 km radius. Any notable patterns in back-azimuthal variations should be included in the supplementary material or directly in the manuscript if they provide significant insights. Furthermore, the back azimuths of null measurements should be marked in the figure of back-azimuthal variations of FPD.

Lines 101-103

The precise number of good and fair measurements should be provided quantitatively for clarity. While Figure 2 may offer a visual representation, but providing the exact number will be helpful for the reader.

Lines 106-110

There are too many null measurements more than good or fair quality splitting measurements. That is why, it is necessary to plot all back azimuths of the null measurements of the back-azimuthal distribution of FPD parameters to detailed analyses, as I noted above in the comment for lines 80-83. In a simple anisotropy case, it is expected that the null measurements come only from the slow and fast axis, however, the distributed variations of null back-azimuths may indicate a more complex structure beneath the region. So, it would be beneficial to plot null back azimuths on the graph of back azimuth versus FPD.

for TableS2 in Supplementary Material,

6^th Column (FPDerr)

• I am confused by the presence of negative signs in certain FPDerr values (e.g., for stations CE03A, CI15A). The error for calculated FPD should be equal for both negative and positive part. Furthermore, the significantly high deviations in the average FPD (for example, approximately 80 degrees error for station CI42A) suggest that the number of available SI measurements may be insufficient.

7^th Column (TDerr)

• In other columns, dots are used to separate decimals, but in the TDerr column, commas are used instead. Is there any special meaning to using commas in this column? Addition to this, given error for time delays are notably high (for example, 1.3sec and 1.7sec TD error reported for the station CI16A and CI42A, respectively).

for Splitting Intensity measurements,

Considering that the deviations of station-averaged FPD and DT obtained by fitting a sine curve on back-azimuthal variations of individual SI measurements are looking quite large, the reliability of these measurements should be provided to the reader. So, how many individual SI measurements were used to derive the splitting parameters (FPD and DT) of the station? It should be indicated in the additional column (maybe the 8^th column) for the reader to follow. Furthermore, the root mean square (RMS) calculated from the difference between the fitted sinusoidal curve and each individual SI measurement will be an important parameter for assessing the reliability of the obtained splitting parameters (FPD and DT). This should also be included in the TableS2.

FIGURES

Figure 1

The Ligurian Mountains and Bresse Graben should be highlighted in the first figure, particularly for readers who are not familiar with the study region, as these locations are mentioned in the text but not in the figure.

Figure 2

Using smaller “a” and “b” labels to refer to the subfigures can enhance the aesthetics of the figure.

Figure 3

Here, if certain regions mentioned in the discussion (such as FPF, Ligurian Alps, Po Plain) are indicated above the topographic cross-section in the figure, the reader can follow the results and discussion much more easily.

Other Questions

• What is the amount of the difference between the average splitting parameters (FPDs and TDs) obtained from SWS and SI measurements? If there are any systematic differences between the average anisotropy parameters obtained from SWS and SI techniques, this might be worth adding to the discussion. (The following articles may be helpful, Kong et al., 2015; Monteiller and Chevrot, 2010)

• What is the averaging method (misfit surface stacking or basic circular arithmetic mean) by calculating the station-average FPD from individual FPDs obtained from the tangential energy minimization technique? This should be mentioned in the Data and Methods section.

In this paper, the authors use seismic data from the CIFALPS2 network to measure shear-wave splitting, thus shedding light on mantle dynamics in the Western Alps. This paper as a whole is well written, shows interesting results and provides reasonable explanations, and the figures are intuitive, thus meeting the interest of SE. However, I still have some major considerations for the analysis of shear-wave splitting measurements.

1. the authors assume that the anisotropic source of the measurements is at 150 km and present all the measurements at the pierce points at this depth. Given that subsequent interpretations are based on this, I might suggest that the authors first perform an estimation of the depth of the anisotropic source, e.g., based on the general use of the method proposed by Liu & Gao (2011). https://doi.org/10.1785/0120100258
2. Similarly, the authors assumed single-layered anisotropy. Although the actual measured FPDs more or less deviate from surface faults or block boundaries, it is still uncertain whether crustal/lithospheric anisotropy contributes to this, so I would suggest that the authors make a comparison with related studies of crustal anisotropy. SI measurements should be another way in which this can be differentiated, according to Silver & Long (2011). From Fig. S2, monolayer anisotropy is the preferred interpretation and the authors should emphasize it further. https://doi.org/10.1111/j.1365-246X.2010.04927.x

Some other comments:

Lines 17-19: Please rephrase this sentence.

Fig. 1: Please mark the key block names here. Also please add the scale of the latitude axis, this will help the reader to determine the position (same for the other diagrams).

Lines 45-57: The authors provide an overview of previous studies of lithospheric structure in the region, but for readers unfamiliar with the region, an introduction to the tectonic settings may be missing. In addition, pending scientific issues need further elaboration.

Data and Methods: I suggest that the authors show shear-wave splitting measurements with different data quality under different regions in the supplementary material.

Fig. 3: R-values are difficult to discern from the graphs. Or maybe use a transparent background colour instead?

Fig. 4: Lack of explanation for the dots.

Lines 191-193: Please rephrase this sentence.