Assessing Huanglongbing Severity and Canopy Parameters of the Huanglongbing-Affected Citrus in Texas Using Unmanned Aerial System-Based Remote Sensing and Machine Learning

1. From your article, I learned that huanglongbing is a great harm to citrus, but why should we distinguish its severity? And do you have any basis for the division you mentioned in Table1? I suggest you explain it more clearly.

Thank you for pointing that out, and we agree with your comments. Therefore, more detail on the importance of HLB severity assessment was clearly explained in line 62 – 68. In addition, we followed the criteria reported by Archer et al. (2022) to perform visual rating of HLB for the division of Table 1. (line 113 – 116)

2.Please specify the data collected for each period in Table 2. I think you are not clear description because 14 months mentioned in Line 234. This kind of messy data is2 a little difficult to realize

Thank you for pointing that out and here is the clarification. After the treatments were injected into HLB-affected citrus, we observed visual ratings and flew the UAS over the trial targeting by monthly. For various reasons, we were not able to collect the data monthly as planned (line 113-114). The 14 months (mentioned in line 240) were counted from the treatments injection (February 2023) to the last month of UAS data collection (April 2024). line 240.

3.This article achieve the good results, the highest accuracy reached about 0.92. But you also mentioned in line 352 that "and accuracy ranging from 0.72 to 1.00 was reported by Lan et al." So what's the innovation in this article?

Our work, for the first time, reported using UAS-based remote sensing approach to assess the severity of HLB. As has not been reported before, we could not find the same work to make such a comparison of accuracy. Therefore, we compared our results with the most similar one instead. Research, done by Lan et al., was very similar to ours where they used UAS-based remote sensing to detect HLB positive, while we were using UAS-based remote sensing to assess HLB severity (line 357 – 361). An innovation in their work was that their approach (AdaBoost) achieved up to 100% accuracy for HLB detection.

1. The paper proposes to obtain visible images and multispectral images by UAS, and then extract VI and TF indexes from these images to evaluate HLB and its impact on citrus. Although the method used in the paper is not particularly novel, the paper describes the entire research process in detail and can provide a useful reference for relevant researchers. The author mentioned in the paper "Our results highlight the potential of using UAS-based features in assessing the severity of HLB-affected citrus trees", and I agree with the author's evaluation of this achievement.

Thank you for your comments

2. Since the author used the TF index, the spatial resolution of the images obtained by the UAS and the visualization results of the TF index extracted from the images should be given in the paper, but no information about the TF index is given in Figures 6 and 7.

We agree with your comments, so a spatial resolution of 2 cm has been already added in the revised manuscript (line 280 and 294)

3. Since the focus of this study is the remote sensing method, and the VI and TF indicators are closely related to the flight altitude of the UAS, it is recommended that the author supplement the results of data processing based on different flight altitudes, so as to provide the best data collection plan.

It is a very interesting topic if a higher resolution could improve accuracy and efficacy of this approach. However, that was not the objective of this manuscript. Therefore, we have included this recommendation as future research in the revised manuscript (line 381 – 382)

Nevertheless, the logical explanation of why we have chosen of flying at 40 m above ground levels was given in line 131 – 137.