Review Reports

Round 1

Reviewer 1 Report

This study used UAV to obtain VIS of paddy fields. Authors investigated two ways of assessment for drought damage. One was the linear regression based on visually assessed Drought Level, the other was k-means clustering without assessed Drought Level. EVI2, as the best prediction parameter of five VIs, could present the damage field level. Estimated damage level by both methods mostly coincided with assessed fields by pest observers. The estimation of method 1 was more accurate, and method 2 was subject to the observation results of POs. This study utilized the feasibility of the UAV-based rapid and objective assessment methods to resolve the subjectivity and time-consuming of artificial observation in damage grading. This achievement provides possibility for large-scale application of crop insurance system, so as to better protect profits of family farmers. But authors still need to give specific explanations and explanations in the following details.

1. In line 19, Whether the word thorough is through.

2. K-means clustering is one of the key methods in this paper. Please elaborate on classified paddy fields based on similarities of 1 to 99 percent of EVI2, which is best illustrated with figures.

3. The data used for calculation in Table 7 comes from the first survey, the second survey or the third survey, please label.

4. Among the five Vis, Whether EVI2 is the most stable and optimal performance in three surveys, and if so, it makes sense to use EVI2 for further analysis. If not, you need to explain the optimal VIs in each survey.

5. In table 6, authors should explain why the determination coefficient in the vegetative stage is higher than those in the heading to harvesting stage, so as to reveal the influence of paddy fields at different growth stages on DL prediction.

Author Response

please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper shows a theme remote sensing/vegetation index for damage assessment for crop insurance in Indonesia. The contribution is significant to the advancement of knowledge. However, some points need to be better detailed for a complete understanding.

Title: ok.

Abstract: I suggest including specific sensor in UAV and analyzed year.

Introduction: is comprehensive whit a good overview of problem in context. However, I suggest detailing with other authors/studies with the use of UAV in crop monitoring.

Materials and Methods: the method description is good. Include the geographic coordinates of the area in figure 3.

Results: is correctly interpreted. Adjust table 7. Include the geographic coordinates of the area in figure 6 and 7.

Discussions: Discussions: I suggest detailing the discussion/authors, for example, comment more on possible limitations of the UAV, other vegetation indices, and compare classification results with other methods/regions/crops.

Conclusions: Detail more and include possible perspectives for future studies.

Author Response

please see the attachment

Author Response File: Author Response.pdf