The manuscript presents a PD-BX method to address the shadow effects of catenary pillar and improve the wind speed measurements along the high-speed railway lines. The RMSE of railway anemometer was reduced, providing enhanced accuracy and reliability of wind measurement. The manuscript is well written, and all figures are clear. Please see my comments and suggestions for minor edits below.

General comments:

The CFD model was performed under a standard state at 25 C, and I assume the dry condition. However, under extreme wind conditions, e.g. strong thunderstorms, relative humidity would be high, and heavy rain is also expected. How does authors’ model perform under such extreme conditions? Also, in real world, it requires quick response of wind speed under such extreme conditions. How long does it take for authors’ model from data processing to wind speed results?

Based on the PD-BX method, authors reduced the uncertainty of wind velocities caused by catenary pillar. Beside velocities, anemometer can also detect wind directions. Do catenary pillars lead to uncertainties of wind directions, e.g. shadow effects. If so, can we also use this PD-BX method to correct the wind directions?

The manuscript presents a PD-BX method aimed at mitigating the shadow effects of catenary pillars and improving wind speed measurements along high-speed railway lines. Overall, the quality of this paper is good. However, certain sections of the English require improvement. Additionally, the following issues need to be addressed and clarified:

1. Strong winds are a significant factor impacting high-speed rail operations. This paper does not discuss the measures that various countries have implemented to respond to mitigate the effects of strong winds; could the authors provide more specific information on this? Additionally, there are few mentions of rail safety incidents caused by strong winds. It is recommended that the authors include more detailed information on the hazards associated with these incidents.
2. In the computational fluid dynamics (CFD) analysis presented by the authors, only dry air and standard atmospheric pressure conditions are specified. Comment 1 supplemented the simulation experiments with temperature, humidity and pressure. Please request that the authors to provide additional fluid dynamic parameters for equations (5) and (6).
3. In the experiment, the authors employed machine learning algorithms to reduce errors. Why not consider simple and convenient physical methods to reduce errors? For example, adjusting the relative position between the anemometer and the catenary support structure could be effective.
4. There is a point of confusion in Section 4 regarding the results and analysis. Generally, the critical limit for wind speed is set at 15 m/s. However, in Figures 10 and 11, it appears that this article utilities 13 m/s as the threshold for wind speed alerts. Could the authors clarify the reasoning behind this discrepancy?