The study presented by Chang et al. is a valuable contribution to the analysis of landslide failure using existing seismic stations, expanding the application limits of previous research. However, several aspects require further review and justification to ensure the accuracy and rigor of the results.

Although the authors have selected existing numerical models (SF) and explained them clearly, a more detailed evaluation of their suitability for the study area and previous validation under similar conditions is essential. This would involve assessing the potential biases and limitations of the models and providing a detailed justification for their selection. Additionally, incorporating a more comprehensive validation of landslides would help avoid biases and ensure result accuracy. Considering terms like "probable landslide" could also enhance clarity and precision. I tried to review the video in S1, but was not available.

The detection of individual landslide events is a critical aspect of the study. To ensure the accuracy of the results, it is crucial to verify that detected events are individual and not the result of superimposed seismic signals. The authors could provide evidence to support that only one landslide is being observed at each time step, thereby strengthening the reliability of the findings.

In Section 4.2, correlating landslide events and explicitly locating them on the area's geology and soils is vital. This would enable the assessment of the accuracy of the used density and avoid biases in mass/volume estimation. Furthermore, correlating events could facilitate the identification of patterns and trends in landslide occurrence, ultimately contributing to more effective prediction and prevention strategies. I would suggest increase the discussion using https://doi.org/10.29382/eqs-2020-0034

Although the usage of digital elevation models (DEMs) is a strength of the study, the comparison between LiDAR and Drone-derived DEMs requires careful evaluation. Considering control points can clarify potential biases in reported heights or the need for vertical bias correction. This would ensure that the DEMs accurately represent the study area's topography. I would suggest consider https://doi.org/10.1111/tgis.12819 in your assessment.

While the study provides valuable insights into landslide failure in the selected study area, it would be beneficial to discuss the potential applications of the methodology in different environmental contexts. For instance, how would the approach perform in regions with varying geological formations, climate conditions, or vegetation cover? Exploring these questions could enhance the study's relevance and generalizability, ultimately contributing to a more comprehensive understanding of landslide dynamics. The authors may consider incorporating case studies or comparative analyses to demonstrate the versatility of their approach. Please, consider discuss https://doi.org/10.5194/nhess-21-3015-2021, https://doi.org/10.1016/j.jhydrol.2021.126294, https://doi.org/10.1007/s10346-023-02179-4