see PDF

Although the study addresses an important topic, substantial revisions are required to improve the scientific rigor, validate the methodology, and clarify the novelty of the work.

1. The novelty of the study should be more clearly established in the Introduction. A more comprehensive review of previous studies is needed to identify the limitations and research gaps in the existing literature, followed by a clear explanation of how the present study addresses these limitations and what its specific scientific contributions are.
2. Aerosol sampling was conducted upstream and downstream of the filter using sampling ports located on the duct cross-section. However, the manuscript does not demonstrate whether the aerosol number concentration was sufficiently uniform across the cross-section. If the concentration distribution was non-uniform, significant measurement uncertainty could have been introduced. Additional evaluation or justification is therefore required.
3. The downstream sampling location appears to have been determined by applying the dilution characteristics previously obtained using NO as a tracer gas. However, the diffusion coefficient of gaseous NO differs substantially from that of aerosol particles. The validity of directly applying the NO-based dilution characteristics to aerosol measurements should be critically discussed. If necessary, the measured data should be corrected or the associated uncertainty should be quantified.
4. The authors assumed that air infiltration through window and door leakage was negligible because a new vehicle was used. This assumption requires further justification. Previous studies on infiltration through building envelopes and leakage paths in indoor air quality research may provide useful methodologies or evidence to support the validity of this assumption.
5. The manuscript states that "isokinetic sampling is not possible outside the vehicle in a traffic-congested situation." However, representative sampling probes capable of operating over vehicle speeds ranging from 0 to 300 km/h have already been developed (https://doi.org/10.1175/JTECH-D-17-0158.1). The authors should review the relevant literature (including the paper suggested by the reviewer) and revise this statement accordingly.
6. The CFD analysis is described only briefly. The numerical methodology, model validation, and interpretation of the CFD results should be presented in greater detail to demonstrate their reliability. Alternatively, experimental airflow visualization could provide more convincing evidence for the airflow characteristics investigated in this study.
7. Particle transport losses in the sampling line were estimated using PLC software. However, the accuracy of these predictions has not been experimentally validated, making it difficult to assess the reliability of the reported filter efficiency. A more thorough validation of the PLC-based predictions is necessary. Alternatively, reproducing the sampling line in a laboratory environment and experimentally quantifying particle transport losses would provide a more reliable approach.
8. The proposed on-road cabin air filter efficiency evaluation method appears to be highly vehicle-specific. Since vehicle geometry and ventilation characteristics differ among vehicle models, the proposed methodology may require substantial modification for each individual vehicle, limiting its general applicability. The authors should discuss this limitation in greater depth and propose possible strategies to improve the universality of the method.
9. Page 13, Line 299: "Figure 7f" should be corrected to "Figure 8f."
10. The cabin air filter efficiency obtained from the proposed on-road evaluation method should be compared with the efficiency of the same filter measured using a well-established standard indoor test method. Such a comparison is essential to demonstrate the necessity and advantages of the proposed on-road evaluation approach and to strengthen the significance of the present study.