This study proposed and established a technical framework based on the lidar and photogrammetry carried by drones, utilizing image recognition technologies to identify plant species to obtain accurate BVOCs emissions. It is expected that the combination of the Lidar characterization technology, the identification technology of tree species, and the tree-species emission factor database could create a new way to accurately quantify the biogenic emissions over a large region. However, in current form, details of technique and the uncertainty discussion are somewhat less satisfactory for AMT journal. In addition, the language of this manuscript need refinement. Overall, I suggest providing more information on method descriptions and uncertainty sources before the manuscript can be accepted. Specific suggestions are listed below.

1. Line 127: What does the ‘forest gap’ mean? Why is it needed to avoid the gap?
2. Method 2.4: Since the drone would obtain large number of photos on different terrain and elevations, it is important to reconstruct the whole targeted area from each photo and avoid the replicated identification of every single tree. Please provide more details on how to combine the information of individual tree location from lidar data with photo taken.
3. Methods: Although this study provide an innovative method to recognize BVOC emission from drone-based lidar and camera, details of specific techniques used here still require further clarification, for example, how to design the flight route, the lidar data processing and the algorithm of identifying tree species.
4. Discussion 4.1: The authors presented several types of uncertainty sources. Then it is curious for readers to know the dominant uncertainty and the exact uncertainty level. In addition, authors could propose some possible solution and research directions to mitigate these uncertainties on emission estimation.
5. Another uncertainty of this method could come from the emission of vegetation below the tree canopy which cannot be detected by lidar or photo. I suggest providing some algorithm to approximate those emissions or at least carry out some sensitivity test.

The study proposes a workflow utilizing drone-mounted lidar and photogrammetry technologies to identify plant species and estimate BVOCs emissions in biodiverse forest ecosystems, and underscores the importance of advancing image recognition technology and sharing data within research communities to enhance BVOC emissions estimation accuracy. However, the description of the methodology in this study is rather crude, lacking detailed explanations, and the uncertainty analysis lacks quantification, resembling a simple compilation of different methods, which fails to demonstrate the superiority of this method. Moreover, the readability of the article is poor, suggesting a need for language revision throughout the manuscript. Therefore, I believe the current version of the article is not suitable for publication in the AMT journal. Some specific comments are as follows:

1. This study only considers the influence of tree species and uses emission factors from MEGAN and literature reports for different tree species for calculation. However, by only considering tree species and not factors such as land use type, leaf biomass, emission factors, and meteorological effects, which are considered in MEGAN, does the accuracy of the calculated results surpass that of MEGAN?
2. The study mentions significant limitations in the research methodology, constrained by the reported tree species, emission factors, and photochemical conditions, leading to potential variations in BVOCs emission results for the same species in different regions and ecosystems. Thus, it restricts the further application and transferability of this method to other forests, necessitating targeted studies specific to local conditions. In such a highly uncertain scenario, what is the practical application value of this method?
3. The analytical methods employed in the article, such as LiDAR-Based Tree Segmentation and Canopy Structure Calculation, lack detailed descriptions of relevant aspects and improvements made to achieve fine-grained segmentation and canopy structure calculation. It is recommended to supplement the description and discussion in this regard.
4. The article mentions that results obtained from different tree species identification software may exhibit certain discrepancies. How should identification and selection be carried out in such cases? How should the resulting uncertainty be considered? It is suggested to supplement relevant descriptions and discussions.
5. The article also acknowledges numerous uncertainties inherent in the method itself, including drone flight altitude, image resolution, selection of image recognition tools, and the inability to identify emissions from vegetation below the canopy. However, the uncertainty analysis lacks quantitative representation. How can targeted improvements and enhancements be made to address the sources of these uncertainties? As AMT is a journal focused on measurement technology, the article should emphasize detailed quantitative descriptions of measurement technology upgrades and modifications, rather than simply combining different methods. It is recommended to supplement descriptions and discussions related to these aspects.
6. There are many grammatical errors in the article, such as lines 56-57, 74-75, and 252-253. Additionally, many sentences are incomplete, such as line 63-64. Furthermore, there is a lot of repetitive expression, such as lines 64-65, 79-80, and 242-243. The overall impression of the article is rushed and lacks careful scrutiny. It is recommended to thoroughly revise the manuscript.