This study presents a semiautomatic method for landslide identification in Germany. I find the topic relevant and promising; however, improvements are necessary, particularly in the methodology section, which requires a more detailed description. Additional comments and suggestions are outlined below:

1. The authors use the term “landslides” in the introduction. In English, this is a general term encompassing all types of mass movements (e.g., shallow landslides, debris flows, rockfalls, etc.). Did your analysis identify all these types? If not, I recommend using a more precise term to reflect the specific process addressed in the study.
2. Study Area section: Please provide information on recorded damage and economic losses in the region, if available. What is the primary triggering factor for landslides in Thuringia? Is it related to tectonic activity, climatic conditions, or other factors?
3. Line 98: The manuscript states that “the area has experienced periods of landslide activity.” Please specify which periods are being referred to.
4. Data section: What criteria were used for visual landslide mapping? Which types of landslides were identified and mapped? This information is essential and should be included.
5. GEOBIA-based landslide inventory mapping section: Please specify the versions of the GIS software used (e.g., eCognition, ArcGIS).
6. Figure 1: I suggest incorporating the symbol for landslide features (currently shown in white) into the map legend itself rather than only in the figure caption. This will enhance immediate understanding, as the current legend indicates landslides in green, which is confusing.
7. Line 126 – Step 1: What were the specific criteria applied for visual landslide mapping? This information is crucial. Please also state the total number of landslides identified and the total mapped area.
8. Line 136: Indicate the software versions used for ArcGIS and R.
9. Lines 145–150: What were the proportions of the samples used for landslide scarps, landslide bodies, and non-landslide areas? Please include this breakdown.
10. STAGE II – Segmentation and Classification: Include the segmentation parameters such as shape and compactness.
    
    Suggestion: A figure illustrating the mapped landslide scarps and bodies would enhance clarity.
11. Line 162: Please elaborate on the “refinement process.” What specific criteria were used to determine when the result was satisfactory?
12. Lines 162–166: Provide the threshold values used in the ruleset applied during the classification (e.g. shape and compactness).
13. Line 194: (Dias et al., 2023). Ensure proper in-text citation formatting and consistency.
14. Figures: Improve the resolution and overall size for better readability and visual interpretation.
15. Section 4.2 – GEOBIA-based landslide modeling results: Include the total area identified for landslide features by both Method I and Method II for comparison.
16. The Results section needs to be more comprehensive. Please include more descriptive analysis and interpretation of the findings.
17. There are two references listed for Dias et al. (2023), labeled “a” and “b” in the references. However, only Dias et al. (2023) is cited in the main text. Please ensure that the correct designation (a or b) is used consistently in both the text and the reference list.

The manuscript presents a very interesting and potentially valuable contribution to the NHESS. The incorporation of GEOBIA into landslide mapping represents a notable advancement in this domain. However, several parts of the manuscript require further modifications and improvement before it can be considered for publication in the journal. 

Authors should outline the specific landslide features that their method is able to identify. In the current case, it seems that we are discussing structured landslide failures such as translational slides.

The authors mention that the area of interest has witnessed several landslide events, but without any clarification of the type of movement or the temporal resolution of the events. Are there event-based failures or is there a temporal scale of their occurrence?

Regarding the manual mapping of landslides, authors should provide clear information on the procedure they followed to map the landslide features. This is crucial for the reader to understand the process of the accuracy assessment in the later stage.

The highlighted advantage of this work is the application of the GEOBIA. The process of identifying objects instead of pixels is crucial and it gives the power for semantic labeling and contextual information incorporation. In this case authors should talk and discuss further the parameters for the segmentation phase, such as scale, shape/color, and compactness. More information is needed on the ruleset development and an explanation of the chosen parameters.

The section on Refinement and accuracy assessment (AA) needs more clarification. I propose to improve it by incorporating more information on how and why the procedure is critical for assessing the performance of the method.

The Results section would benefit from a more detailed and thorough presentation. Please provide a deeper interpretation of the findings to enhance clarity and understanding for the reader. There are several figures that look blurred on the manuscript. Please take a look at them and provide better quality as outputs to enhance the quality of the work.