Review Reports

Comments

Answers

In my opinion, this study is more like a case study. The knowledge advancing the field of forestry is inadequate.

We agree that our study focuses on a specific case, and we acknowledge the simplicity of the experimental design. However, we would like to emphasize that the aim of the work is not to provide broad generalizations across species or sites, but rather to demonstrate the feasibility of employing a low-cost MEMS accelerometer for tree dynamic stability monitoring. This proof-of-concept is a crucial step toward enabling scalable and continuous monitoring of urban trees, which has been identified as a pressing need in both urban forestry practice and research.

Although case-specific, our results provide significant insights into (i) the operational requirements under which low-cost sensors can deliver reliable data, (ii) their comparative performance against high-precision seismic sensors, and (iii) their potential for integration into IoT-enabled systems for large-scale deployment. These findings represent an important advancement for the community, as they lay the foundation for future studies across more species, sites, and conditions, and they open the way for the development of affordable and practical monitoring systems in urban forestry.

Thus, while the study is intentionally targeted and preliminary in scope, we believe it provides meaningful contributions to the field by delineating operational boundaries, validating a promising low-cost technology, and charting a path for broader application and research.

Please see Line:85-99, 448-461

Comment 2

From lines 193 to 197, the authors used bandpass filters with different frequency ranges for laboratory tests and field tests. Please explain why.

Some big trees may have very low natural frequencies, in the order of 0.1 Hz (see e.g., Jackson et al. (2019) and Chau et al. (2022)). Setting the bandpass filter with lower bounds of 0.1 Hz (line 196) or 0.5 Hz (line 197) will incorrectly identify the natural frequencies. The bandwidth should avoid the possible range of the oscillation frequencies of the trees.

Do you apply the same filtering techniques to the accelerometers for seismic-level accuracy? Please clarify.   

Comment 14

The length and materials of the beam and the installation locations of the sensors should be reported.

We have revised the manuscript to clarify the rationale for the choice of bandpass filter ranges in the laboratory and field tests. The use of different ranges was intentional and reflected the distinct characteristics of the tested systems. Laboratory tests were conducted on a steel cantilever beam (140 cm long, cross-section 50 × 5 mm), rigidly fixed at one end and free at the other. The sensors were installed at the free end of the beam, where displacement is maximal, and the filter settings were tailored to isolate the relevant higher natural frequencies while minimizing noise.

For the field tests, the bandpass filter was set between 0.5 Hz and 9 Hz. We acknowledge that very large trees may exhibit natural frequencies as low as 0.1 Hz (Jackson et al., 2019; Chau et al., 2022). However, the trees investigated in this study were of moderate size (DBH = 17.5 cm, height = 11 m), resulting in higher dominant natural frequencies. Accordingly, the chosen lower bound of 0.5 Hz ensured that the primary oscillation modes were captured.

Finally, we confirm that the same filtering methodology was consistently applied to all sensors, including the high-precision seismic accelerometers, thereby ensuring a fair and reproducible comparison.

Please see Line: 156-161, 195-198, 222-240

Comment 3

Please report the heights and diameters at breast height of the trees in the field. The authors may provide photos of some examined trees. It is beneficial to know the leaf distribution and the branching architecture of the examined trees. This information is important to estimate the natural frequencies of trees to check if the results mentioned in the study make sense or not.

We agree that this information is important to report. In the revised manuscript, we have added the diameter at breast height (DBH = 17.5 cm), the total height of the tested tree (11 m), and the estimated crown area (5 m²). We have also included photographs of the examined tree to document its crown structure and leaf distribution.

Please see LINE: 196–198.

Comment 4

I would suggest that the authors install the MEMS sensors at higher locations. In the current tests, they installed the sensors at about 220 to 270 cm (a very low position). The vibrations at these locations are very low (given the ages of these trees). It is expected that the measurements of MEMS sensors are not reliable, especially when the trees are under ambient vibrations.

We agree that mounting the sensors at 220–270 cm is not optimal for capturing the full extent of ambient vibrations. However, we emphasize that for the objectives of this feasibility study, co-locating the low-cost MEMS unit and the seismic reference at 2.70 m provided reliable data for comparative analysis. This height was deliberately chosen to reflect realistic urban deployment and scalability constraints: (i) ensuring operator access for battery replacement and inspection without lifting equipment; (ii) reducing exposure to casual tampering; and (iii) providing a sufficient lever arm for forced-excitation tests such that sway-induced accelerations exceeded the device noise floor. We have clarified this rationale in the revised Methods section. While we acknowledge the limitation for ambient-wind monitoring, we stress that the measurements remain valid for the study’s aims. As noted in the Discussion and Conclusions, future work will extend validation to multiple mounting heights along the trunk and within the crown to better capture ambient responses.

Please see LINE: 145-150, 203-210, 455-457

Comment 5

Please report the frequencies associated with the highest peaks in Figure 6. The text “around 2 Hz” in line 230 is a bit vague. Please modify. As seen from the blue line of Figure 6, it seems that there are also two peaks in the range of 0 to 1 Hz. There is also a peak at around 4 Hz. Please explain those peaks.

Also, in line 226, the authors used the term “poorly damped”. It is unclear what it means. Do you mean “lightly damped”? Please clarify. Moreover, for both Figures 6 and 7, please report on the damping ratios of free decay.

We agree that reporting the exact frequencies is important for replicability. In the revised manuscript, we now provide the precise frequencies associated with the main peaks in Figure 6, avoiding vague expressions such as “around 2 Hz.” The dominant peak appears at approximately 1.7 Hz in both sensors, corresponding to the natural frequency of the cantilever beam. In the spectrum of the low-cost sensor, additional peaks are visible in the lower frequency range (0–1 Hz) and at approximately 4 Hz. Since these peaks are absent in the high-cost reference sensor, we attribute them to sensor noise rather than to actual vibration modes of the system; this clarification has been added to the Results section.

We have also corrected the terminology by replacing “poorly damped” with “lightly damped” to avoid ambiguity. Finally, as suggested, we now report the damping ratios estimated from the free decay analysis for the laboratory tests in both Figures 6 and 7.

Please see LINE:  272-277, 283-284, 329-347

Comment 6

Figure 9 clearly indicates that the authors did not install their sensors at sufficiently high locations. They should have a small estimation of the vibration levels of trees before installing these sensors on trees; otherwise, you cannot measure anything. Also, the results in Figure 9 show that the measurements between the low-cost MEMS sensors and the PCB Piezotronics Model 356B18 are quite noticeable. Do you use the same filtering techniques for both sensors? Are they a fair comparison? Please clarify.

 Please improve Figure 9 as well. Its font size is too small.

As noted in our response to Comment 4, we agree that the installation height (220–270 cm) is not optimal for detecting ambient crown vibrations. However, we deliberately co-located the low-cost MEMS unit and the seismic reference at this height to reflect realistic urban deployment and scalability constraints: (i) facilitating operator access for battery replacement and inspection without lifting equipment; (ii) reducing exposure to casual tampering; and (iii) providing a sufficient lever arm for forced-excitation tests, ensuring that sway-induced accelerations exceeded the device noise floor. We have clarified this rationale in the revised Methods section, while also acknowledging the limitation and noting in the Discussion and Conclusions that future tests will involve installations at higher points along the trunk and into the crown to better capture ambient vibrations.

Regarding the comparison between instruments, we confirm that the same filtering methodology was consistently applied to both the low-cost MEMS sensors and the high-cost seismic accelerometers, thereby ensuring a fair and reproducible comparison. Finally, we have improved the readability of Figure 9 by increasing the font size and enhancing the overall clarity of the plots, as suggested.

Please see LINE: 227-229, 314-318

Comment 7

In Eq. (1), the authors state that the measured crown area is 5 m2. How do you measure it? Please clarify. The authors used Eq. (1) to identify the “threshold” of wind speed in line 346 as 13.9 m/s. This is a very high wind speed! I am not sure if it makes sense. Note that Eq. (1) is valid when the force is acting on the crown. However, in the tests, the pulling force acted on the stems (I guess). The estimations of the required wind speed are a bit questionable.

We sincerely thank the reviewer for this important comment. The crown area of approximately 5 m² was estimated by delineating the crown boundary from a top-down orthomosaic. As the reviewer correctly points out, Eq. (1) is formulated for the case of wind force acting directly on the crown. In our pulling tests, however, the pulling device and load cell were attached to the stem at the pulling point, about 3 m above ground level. This setup follows the standard practice for rapid-release pulling tests in arboricultural risk assessment (Wessolly & Erb, 1998; Brudi & van Wassenaer, 2002; Giachetti & Ferrini, 2020). For this reason, the wind speed value of 13.9 m/s derived from Eq. (1) should be regarded only as an indicative reference, not as an exact physical equivalence. This limitation has been clarified in the revised manuscript. Please see LINE: 243-257, 403-408, 196-198

Comment 8

In lines 244 to 246, the authors mention that “However, when setting up the tests in the field, we observed that the damping ratio of the tree system was significantly higher.”. This is as expected, because trees use multi-modes to vibrate and dissipate energy, especially under free vibrations (Rodriguez et al. 2008, Xuan et al. 2021, Loong et al. 2023). Certain discussions connecting the observations from the tests and the knowledge of tree dynamics are suggested. 

Thanks, helpful suggestion. We have revised the Discussion to explicitly connect our field observations with the established knowledge that trees dissipate energy through multi-modal vibrations under free oscillations (Rodriguez et al., 2008; Xuan et al., 2021; Loong et al., 2023). This explains the higher damping ratios we recorded in the field compared to the laboratory beam tests. In addition, we added a short statement in the Introduction to highlight that multi-modal vibration is a characteristic feature of tree dynamics, which provides context for the interpretation of our results.

Please see LINE: 67-68, 392-395

Comment 9

Some studies apply other technologies to measure tree vibrations (see Chau et al. (2022, 2023)). Any advantages of using low-cost MEMS sensors over the state-of-the-art technology (Zanotto et al. 2024)? The discussions along this direction could be improved.

We have expanded the Discussion to compare low-cost MEMS sensors with state-of-the-art technologies (Chau et al. 2022, 2023; Zanotto et al. 2024). While advanced systems provide higher sensitivity and accuracy, their use in urban contexts is limited by cost and technical complexity. By contrast, MEMS sensors, although less accurate for low-amplitude signals, are more affordable, easier to install, and scalable. We now highlight that MEMS can complement high-end systems, serving as practical first-level monitoring tools within large-scale networks

 Please see LINE:  422-437

Comment 10

I am curious about the power consumption of MEMS sensors in the field. Are these sensors transmitting data wirelessly? How much is this wireless transmission system? Note that throughout the manuscript, the specifications of MEMS sensors (the manufacturing company, the specification names, and the weight) are unclear.

The MEMS sensors used in this study are part of a new prototype, and some technical specifications cannot be disclosed in detail due to ongoing internal work aimed at securing a patent for the instrument. However, we have clarified in Section 2.1 (Sensors) that the instrument is based on an STMicroelectronics triaxial MEMS chip, with a configurable full-scale range of ±19.6 m/s². The device operates with dual power supply (battery or continuous), and in this study it was used in battery mode. Wireless transmission enables field measurements lasting several hours. 

 Please see LINE: 121-126, 428- 437

Comment 11

What do gpk and spk mean in lines 95-96 and 103? Do you mean “g” (earth gravitational acceleration) and second? Please clarify.

We thank the reviewer for pointing this out. We have revised the text to avoid ambiguity, and the previous abbreviations (gpk and spk) have been replaced with the correct SI notation for acceleration, expressed in meters per second squared (m/s²).

Please see LINE: 101-121

Comment 12

How do the authors estimate the damping ratios? Are you using a half-power bandwidth or other methods? Please clarify. Normally, the damping ratios have high uncertainty. The authors may report the standard deviations of measured damping ratios in Table 2.

The damping ratios were estimated using the logarithmic decrement method, applied to the free decay of oscillations in the time domain. This approach was selected because the rapid decay of tree oscillations, combined with the noise in the MEMS signals, rendered frequency-domain methods such as the half-power bandwidth unsuitable.

We agree that damping ratios are subject to uncertainty. In the laboratory tests, where oscillations were longer and cleaner, we were able to calculate both mean values and standard deviations. For transparency, these results are reported in the table below, although they were not included in the revised manuscript since this level of detail is not central to the study’s primary objectives.

In the field tests, however, the limited number of observable cycles and the higher noise floor did not allow for a robust statistical treatment; therefore, only mean values are reported in Table 2 of the manuscript.

Please see LINE: 250- 256

Comment 13

I am very curious about the results for Preliminary Tests 7 to 10, with different pulling directions. Do you obtain the same damping ratios under different pulling directions (see e.g., Loong et al. (2023))? Please clarify the directions of pulling with respect to the geometry of the trees. In addition, please report the results as well.

Preliminary Tests 7–10 were designed with different pulling directions mainly to explore the feasibility of the setup and to verify sensor synchronization, rather than to extract quantitative parameters. For this reason, no damping ratios were calculated from these tests, and the data were not included in the results. The directions of pulling with respect to tree geometry have now been clarified in Section 2.3 (Preliminary field test), and we have specified in the text that these preliminary trials served to refine the methodology but did not provide directly comparable outcomes

Please see Lines: 175-191, 305-313

References

Chau WY et al. 2022. Understanding the dynamic properties of trees using the motions constructed from multi-beam flash light detection and ranging measurements. Journal of Royal Society Interface, 19(193): 20220319.

Chau WY et al. 2023. Monitoring of tree tilt motion using lorawan-based wireless tree sensing system (IoTT) during super typhoon Mangkhut. Agricultural and Forest Meteorology, 329: 109282.

Jackson et al. 2019. Finite element analysis of trees in the wind based on terrestrial laser scanning data. Agricultural and Forest Meteorology, 265: 137-144.

Loong CN et al. 2023. Reconstruction methods for the mechanical energy of a tree under free vibration. Agricultural and Forest Meteorology, 339: 109541.

Rodriguez M et al. 2008. A scaling law for the effects of architecture and allometry on tree vibration modes suggests a biological tuning to modal compartmentalization. American Journal of Botany, 95(12): 1523-1537.

Xuan Y et al. 2021. The potential influence of tree crown structure on the Ginkgo harvest. Forests, 12: 366.

Zanotto F et al. 2024. Wind-tree interaction: Technologies, measurement systems for tree motion studies and future trends. Biosystems Engineering, 237: 128-141.

The above references only substantiate my statements/comments on the authors' work. I did not recommend the authors add these references.

We thank you for providing these additional references, which help to substantiate the discussion. Although they were not explicitly requested to be added, we have decided to include them in the Introduction and Discussion to strengthen the context and ensure consistency with the new parts we added when comparing low-cost MEMS sensors with state-of-the-art technologies (Chau et al. 2022; Jackson et al. 2019; Loong et al. 2023; Rodriguez et al. 2008; Xuan et al. 2021; Zanotto et al. 2024).

1. While the study demonstrates the technical potential of MEMS accelerometers, the practical aspects of large-scale deployment are not clearly addressed. Specifically, what are the anticipated labour costs for installation across numerous trees, and is such deployment realistically achievable in urban contexts? This question is critical to evaluating the broader applicmability of your approach.

We agree that the practical aspects of large-scale deployment, including labour costs, are an important consideration. As this work is presented as a feasibility study, our primary objective was to assess the technical reliability of MEMS accelerometers in comparison with seismic reference sensors, rather than to conduct a full economic analysis. To provide an order of magnitude, its cost is between one and two orders of magnitude lower than that of professional seismic accelerometers or equipment for conventional pulling tests. 

Relevant details have been added and clarified in the manuscript.

Please see LINE: 362-370, 451-461, 429- 437, 146-150

2. There is an inconsistency between the stated focus on urban trees and the study site shown in Figure 3, which appears more representative of a forest environment. Moreover, Figure 4 suggests a different location, which you stated was on campus on an urban tree, and indicates that the sensors were connected by cables. Such a setup seems impractical for a large-scale urban tree situation.

Further clarification is needed on how the field experiment was conducted in practice:

o How were the sensors powered?

o How was data transmitted or stored?

o To what extent can the presented setup be realistically scaled or adapted for urban tree monitoring?

The study was conducted in three phases: (i) laboratory tests, (ii) preliminary field tests on Alnus glutinosa in a plantation stand (Figure 3), and (iii) operational field tests on Ailanthus altissima located on the University of Florence campus (Figure 4), selected as representative of an urban tree. We have revised Section 2.2 (Experimental design) to make this distinction clearer. The overarching objective of this preliminary work is the development of an instrument for urban tree risk assessment, which represents the broader context of the study.

Regarding the setup, the MEMS sensor is a wireless prototype based on an STMicroelectronics triaxial chip, operated in this study with battery power and Wi-Fi transmission to a server. By contrast, the seismic reference accelerometers are cabled instruments that required direct connection to the acquisition system. We have clarified these aspects in Section 2.1 (Sensors) and in the Discussion. Finally, we emphasize that, while the use of cabled seismic sensors is not scalable for large-scale networks, the MEMS prototype is specifically designed for wireless operation and can realistically be adapted for urban monitoring applications.

Please see LINE: 165-202, 121-127, 128-150

3. The manuscript repeatedly contrasts “low-cost” MEMS sensors with “high-cost” seismic sensors, but does not provide actual price ranges. Including approximate cost figures would strengthen the discussion and better illustrate the economic advantage. And have you considered employing any Remote Sensing approaches? E.g., a Laser or optical sensor for multiple tree observations?

As noted in our response to Comment 1, the prototype MEMS devices tested in this study cost between one and two orders of magnitude less than professional seismic accelerometers and the equipment used for conventional pulling tests, making them substantially more affordable. 

 We have also added a discussion emphasizing that the MEMS sensor is between one and two orders of magnitude less expensive, making it a practical, economical alternative for specific applications. With regard to remote sensing approaches, we agree that they hold great potential for multi-tree observations. Accordingly, we have added a note in the Future Work section of the Conclusions to highlight the importance of integrating MEMS monitoring with complementary technologies such as LiDAR and IoT-based systems.

Please see LINE: 429-431, 456-462

Presentation and Structural Concerns

• The acronym MEMS is used in the title, abstract, and introduction, but the full term is not introduced until line 83. For clarity, the full form should be defined upon first use.

• The abstract and introduction emphasize the role of urban trees in socio-ecological systems, yet the study was carried out on forest trees. The implications of this discrepancy should be explicitly addressed, particularly since cost and accessibility considerations differ substantially between urban and forest contexts. 

• The manuscript would benefit from improved logical flow and clarity of presentation to better guide the reader through the methodology and findings.

• In lines 55–59, reference [24] is cited twice in succession for two different points. Please ensure correct citation usage, and include the publication year within the text for clarity.

We sincerely thank the reviewer for their careful review and for providing such detailed and helpful comments. We have made the following revisions to address the points raised:

• Acronym MEMS: The acronym MEMS is now defined at its first use in the abstract and introduction, ensuring clarity and adherence to standard scientific writing practices.

• Urban vs. Forest Discrepancy: We have explicitly addressed this discrepancy in the manuscript. We clarified that the study site was chosen to provide a controlled environment for a rigorous methodological comparison, while the characteristics of the test tree itself (e.g., its pruning history and physical location) are highly representative of urban trees. This approach ensured that the findings on sensor performance remain applicable to urban contexts, despite the semi-rural setting of the preliminary test site.

• Improved Logical Flow: We have carefully restructured the manuscript to improve its logical flow and clarity of presentation, making it easier for the reader to follow the methodology and findings.

• Citation Usage: The citation for reference [24] has been corrected to ensure proper usage and clarity throughout the text.

Overall Assessment

This study provides useful documentation of MEMS accelerometer testing for tree monitoring, and the results are promising. However, substantial revisions are required to strengthen the methodological transparency, align the study context with its stated objectives, and improve overall presentation. Addressing the above issues will significantly enhance the clarity, credibility, and impact of the paper.

We thank the reviewer for these valuable comments. We believe the observations have contributed significantly to improving the manuscript, as was also the case with the suggestions from the other two reviewers. We have revised the text to enhance clarity and have added further details across the relevant sections to strengthen methodological transparency and improve the overall presentation of the work.

The Abstract - it is well written and have a proper structure, it presents a small background for the subject of the article, the aim of the study, some methodological aspects and the principal results and conclusions.

In this section, I suggest to use only the word ”mitigate” not both ”mitigate/reduce” – please see line 3. 

We have amended the text and now consistently use the term “mitigate” as recommended.

Please see LINE: 3-4

Keywords – the keywords are chosen well and reflect the information and the topic from the article.

Introduction - it is comprehensive and related with the subject of the paper. It presents the aim of the paper and information related to the importance of urban forests and urban trees, as a key component in cities, which could contribute to a better environment, but in some case also could produce disservices.

The introduction section presents also detailed information about the static and dynamic techniques which could be used to evaluate the health of trees and their response to different stress factors, like wind.

Knowing the reaction of trees in various scenarios and how the external factors affect the dynamic parameters of the trees are crucial points in ovoid their failure and protect people.

As recommendations, I suggest the following:

• Line 26 – please use another word instead of ”previous” from the context; maybe ”other” will be better for my point of view – ”… and other studies …”;
• Line 45 – please note that in the text is mentioned only the source no. 17, so is ok to write “recent studies”? or will be better to putt “a recent study”?  

We have revised the text accordingly, replacing “previous” with “other” and changing “recent studies” to “a recent study” for greater precision.

Please see LINE: 27, 50

Methodology – is very well structured and all the stages of the working protocol al detailed described, which is very important for a scientific article, because in some cases the same protocol is used by other authors in different conditions, to evaluate a similar process.

The authors present the two seismic accelerometers used in the study and also the low-cost sensor’ characteristics. They presented in detail the methodology applied and the protocol specific to the laboratory tests, preliminary tests and field tests.

Questions:

1. Line 142 – please mention in how many trees were made the preliminary field investigations?
2. Lines 157 - which was the height of the tree subjected to the field tests? Could it influence the results? In other cases (other dimensions of the trees), which will be the results?
3. Lines 166-167 – Why do you changed the height where the sensors were placed comparing to that of pulling tests? Could this change the obtaining results?

Firstly, this height was deliberately chosen to reflect realistic urban deployment and scalability constraints: (i) ensuring operator access for battery replacement and inspection without lifting equipment; (ii) reducing exposure to casual tampering; and (iii) providing a sufficient lever arm for forced-excitation tests such that sway-induced accelerations exceeded the device noise floor. We have clarified this rationale in the revised Methods section.  

Secondly, within this standard range, the precise placement was adjusted according to the trunk morphology of each tree, ensuring the most stable and secure mounting point possible. While the absolute mounting height of the sensor can influence the recorded acceleration amplitude, this variation does not affect the primary conclusions of our study. 

The central aim was the methodological comparison between the low-cost and high-cost sensors. For this reason, in every test both sensors were mounted at exactly the same height, thereby ensuring a direct, one-to-one comparison that underpins our findings. We have added a note to the manuscript to clarify this point.

Please see LINE: 195-197, 178-181, 143-150

As recommendations, I suggest the following:

• Line 115 – please ovoid to use the word “we”, because the scientific papers have to be written in an impersonal way, so please change it;
• Line 128 – figure 2 – please move the Figure 2 after her first mention in the text, because in this way the readers could find it rapidly and could understand better the presented information;
• Line 132 – please ovoid to use the word “We”, because the scientific papers have to be written in an impersonal way, so please change it;
• Line 136 – please move the Table 1 (which now is placed at the page 7 – so after 3 pages from it is mentioned for the first time), immediately after Figure 3;
•  
• Line 148 – please ovoid to use the word “our”, because the scientific papers have to be written in an impersonal way, so please change it;

Line 153 – please delete the word “see” when you mention the Table 1;.

We thank the reviewer for these valuable suggestions to improve the manuscript’s readability and compliance with scientific writing standards. The manuscript has been carefully revised to adopt impersonal language, all figures and tables have been repositioned to appear immediately after their first mention in the text, and the word “see” has been removed from table references.

• Line 218 – what means “U”? 

We apologize for the earlier lack of clarity. The symbol “U” denotes the estimated wind speed, calculated using Equation (1). We have now explicitly defined this variable in the text at the point where the equation is introduced.

Please see LINE: 258-262

Results and Discussion – are good sections and illustrate the results obtained by the author after applying the methodology. The authors presented the results obtained at the laboratory tests, preliminary tests and field tests, comparing the results of seismic accelerometer (high-cost sensor) to those of the low-cost sensor, in time and frequency domains.

At the beginning, the results indicate similar data for the two models of used sensors, but the low-cost sensor is more influenced by the Ambiental noise, aspect which highlighted the limitation of the low-cost sensors under real-world conditions.

Question:

1. Line 291 – why you highlighted only the test D and F for the data registered for damping, because in table 2 are presented data for all the tests you performed? 

We thank the reviewer for this question, which allows us to clarify our reasoning. Tests D and F were highlighted because they provide clear and representative examples of the high variability observed in the damping estimates from the low-cost sensor. In particular, these tests show marked deviations from the more stable values recorded by the reference sensor, thereby effectively illustrating the “scattered” performance described in the text. Our intention was to use these concrete examples to make the general trend of inconsistency more accessible to the reader. To improve clarity in light of this feedback, we have added a phrase to the manuscript explicitly noting that these are illustrative examples.

Please see LINE: 330-347, 348-355

As recommendations, I suggest the following:            

• Line 230 – figures 7 and 8 – please move these figures after their first mention in the text<
• Lines 242, 245, 246, 247, 251, 265 – please ovoid to use the word “we”, because the scientific papers have to be written in an impersonal way, so please change it;
• Lines 242 – 256 – usually, between the name and number of a chapter and the first subchapter it is not recommended/allowed to find text; so, I suggest to insert a name and a number for this paragraph and to reorder the next ones;
• Lines 292-293 – table 3 – please move the Table 3 after his first mention in the text;
• Lines 289-292 – And please mention something also about the amplitude values, because those are also presented in table 2;
• Lines 293-300 – maybe will be better to talk firstly about the Root Mean Square (RMS) and after that about Signal-to-Noise Ratio (SNR), as these were presented in the table. 

We thank the reviewer for this question, which provides the opportunity to clarify our reasoning. Tests D and F were selected because they represent clear and illustrative examples of the high variability observed in the damping estimates from the low-cost sensor. These tests, in particular, show substantial deviations from the more stable values recorded by the reference sensor, thereby effectively demonstrating the “scattered” performance described in the text. Our intention was to use these examples to make the general trend of inconsistency more evident to the reader. In response to the reviewer’s feedback, we have added a phrase to the manuscript explicitly noting that these are presented as illustrative examples.

Please see LINE:  339-347, 348-354 

Finally, the authors presented at the end of the article come interesting conclusions, related and supported by the results.

As suggestion, at line 378 maybe will be more relevant to write the simulated wind speed (as was calculated with formula 1), and not the force (above 600 N).  

References – all the references from the list are mentioned in the text, they are recent and in relation with the subject of the article.

To maintain a concise conclusion focused on the direct experimental finding of the 600 N force threshold, we have presented the detailed conversion to the equivalent wind speed in the Discussion section, where it is more appropriately contextualized.

Please see LINE: 444-451, 402-408