the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Nov 2024
A listening experiment exploring the relationship between noise annoyance and sound quality metrics for airborne energy systems
Abstract. The present study investigates the relationship between sound quality metrics (SQMs) and noise annoyance for airborne wind energy systems (AWESs) reported in a listening experiment. A convenience sample of 75 participants rated their annoyance on the International Commission on Biological Effects of Noise (ICBEN) scale in response to nine recordings from in-field measurements of two different fixed-wing and one soft-wing ground-generation AWES. All recordings were normalized to have the same A-weighted equivalent sound pressure level. The acoustical analyses showed that the fixed-wing kites presented a more tonal and narrowband sound signature than the soft-wing kite. Linear-mixed effects models indicated that sharpness was the only SQM predicting participants’ annoyance ratings and that the fixed-wing kites were rated as more annoying than the soft-wing kite. In addition, the effect of some SQMs on annoyance depended on participant characteristics, with loudness having a weaker impact on annoyance for participants familiar with AWESs and tonality having a weaker effect on annoyance for older participants. However, these moderation effects could be random due to the non-probability sampling used.
- Preprint
(1315 KB) - Metadata XML
-
Supplement
(657 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on wes-2024-125', Anonymous Referee #1, 12 Dec 2024
General comments
The present study is interesting and innovative as it gives insight into the properties of a new and developing technology. It applies knowledge about a more established form of wind energy, namely that sound emissions are a central issue for residents, to a context where conflictual cases can yet be prevented. Analyzing the psychoacoustic metrics of different form factors is important in order to understand how design choices affect the experience.
The article is very well written, methodology and results are presented clearly. Where it falls short of its potential yet is in the discussion of what the study means for AWE. Both in the introduction as well as when discussing the results there is little reference to other studies with similar approaches. As there is not a lot of research on AWE specifically, a comparison to the psychoacoustic metrics of other, to some degree similar technologies would help a reader gauge the significance of the results. As it is, a reader without specific psychoacoustic background knowledge would have a difficult time judging whether the results are surprising or expected. Furthermore, it could be clarified what the results imply for the design of AWESs. Do they suggest improvements to be made, for example?
Specific comments
- Section 2.1: How comparable are the three recordings per AWES to each other? Are the recordings selected to cover varying situations or are they supposed to be consistent? The spectrograms in section 3.1 reveal remarkable differences for kite C, whereas the recordings of the other two kites seem fairly homogenous.
- 84: Please add some explanation why only recordings of the reel-out phase were chosen. Are there differing acoustic properties of the phases? Is the reel-out phase more relevant? Are they similar enough that the results of one should translate to the other?
- 256-257: Please elaborate on why the circular flight pattern results in a bump above 1 kHz. How does this explanation relate to the one from l.274-275, stating the bump is a result of rigid materials?
- Section 3.2: Is there a criterium applied to judge which differences are meaningful? Regarding the loudness of kite C, for example, the text says that C3 exhibits higher values than C1. Looking at Figure 6 this is true regarding the maximum values, though not with respect to the distribution or median values. If the statement is based on the 5th percentile what is the threshold of a meaningful difference?
- 329: Please add some words of guidance to Figure 8, help the reader interpret how strong the effect is.
- It was not the aim of this study to compare annoyance levels of AWESs to other emission sources. But can you provide some point of comparison to understand whether the annoyance levels are particularly high or low? At least, it might be worth giving a verbal orientation on how to interpret the annoyance values within the scale (l.314-315).
- You state that the moderation effects might be tied to the sample. There are more specific explanations possible. Familiarity with AWE at a technical university with ties to AWE development might imply a more positive attitude, hence affecting annoyance ratings. On the other hand, the sample is particularly young, raising questions about the effect of age on tonality. Given the age distribution, what age differences result in the effect of age on tonality?
Technical corrections
- 14 & 75: Mostly, in this paper the word “acoustic” is used. In these two instances it is instead “acoustical”. I do not think this is to indicate any semantic difference. Perhaps you might want to keep it consistent.
- 25-27: “during natural disasters“ sounds like the kites are in the air while there is a heavy storm. I assume, you mean something like “following / in the aftermath of natural disasters”?
- 27: I suggest “lower mass” instead of “mass savings”. Also, it should be stated what AWESs are compared to.
- 33: This is the first time the abbreviation AWES is used. It is introduced a few lines later on l.41.
- 55: I suggest “enhance” instead of “enhances” as it refers to multiple aspects.
- 155-156: I suggest “in the analyses this variable was not considered as a confounding factor”. Please clarify that this refers only to experiences with AWE sounds as familiarity is considered in later analyses.
- Figure 3: There is no indication what the color grading represents.
- 230-231 “Examining the spectrograms...”: It is suggested that this refers to spectrograms of other recordings than the ones displayed here. If so this should be clarified.
- 251-252: Is the order of SPL below 2 kHz correct? Judging from Figure 4 the levels of C1 look to be higher than those of C2.
- Description of Figure 6: Although 5th percentile seems to be a standard denotation, next to the explanation of the 25th and 75th percentiles I find it quite irritating as both naming conventions count from different directions. Indicating that the highest 5% are highlighted might avoid confusion.
- 305: There seems to be a mismatch between text and table regarding the %HA.
- Figure 7: In some boxes the central horizontal line is hard to see. I suggest the lines and frames of the boxes be black rather than match the color.
- 369-370: There is a contradiction to the results chapter. Here it says that the fixed-wing kites show higher loudness, whereas in the results the soft-wing kite had the highest loudness values.
Citation: https://doi.org/10.5194/wes-2024-125-RC1 -
RC2: 'Reply on RC1', Anonymous Referee #1, 12 Dec 2024
To specifiy what I was getting at in specific comment 7: The question is if you see any merit to / evidence for these hypotheses.
Citation: https://doi.org/10.5194/wes-2024-125-RC2 -
AC1: 'Reply on RC1', Helena Schmidt, 14 Jan 2025
Reviewer 1
Response to reviewer
1) General comment
The present study is interesting and innovative as it gives insight into the properties of a new and developing technology. It applies knowledge about a more established form of wind energy, namely that sound emissions are a central issue for residents, to a context where conflictual cases can yet be prevented. Analyzing the psychoacoustic metrics of different form factors is important in order to understand how design choices affect the experience.
The article is very well written, methodology and results are presented clearly.
We thank you for your kind words.
2) General comment continued.
Where it falls short of its potential yet is in the discussion of what the study means for AWE. Both in the introduction as well as when discussing the results there is little reference to other studies with similar approaches. As there is not a lot of research on AWE specifically, a comparison to the psychoacoustic metrics of other, to some degree similar technologies would help a reader gauge the significance of the results. As it is, a reader without specific psychoacoustic background knowledge would have a difficult time judging whether the results are surprising or expected. Furthermore, it could be clarified what the results imply for the design of AWESs. Do they suggest improvements to be made, for example?Thank you for highlighting this important point. We have expanded the introduction and discussion to include comparisons with (psycho)acoustic studies on similar technologies, such as wind turbines and drones. These comparisons help contextualize the findings and demonstrate whether the observed patterns in sound quality metrics (SQMs) and annoyance ratings are consistent or unexpected. Additionally, we now address the design implications of our results.
SPECIFIC COMMENTS:
3) Section 2.1: How comparable are the three recordings per AWES to each other? Are the recordings selected to cover varying situations or are they supposed to be consistent? The spectrograms in section 3.1 reveal remarkable differences for kite C, whereas the recordings of the other two kites seem fairly homogenous.
The three recordings per AWES were selected to represent typical sound emissions during the reel-out phase of the respective system. Given that kite C was towed by a truck, there is larger variability within the 3 recordings in comparison to kites A and B. We have added this explanation in Section 2.1, “Sound recordings”.
4) 84: Please add some explanation why only recordings of the reel-out phase were chosen. Are there differing acoustic properties of the phases? Is the reel-out phase more relevant? Are they similar enough that the results of one should translate to the other?
The reel-out phase was chosen because it is acoustically the most dominant and operationally consistent phase across all AWESs. During this phase, the kite operates in crosswind maneuvers at high flight speeds while the reel-out speed is kept relatively low to maximize the energy production period. This operational setup implies that sound emissions from the kite, including contributions from the onboard ram-air turbine, wing flutter, and tether vibrations, are the most significant. In contrast, sound emissions from the ground station (e.g., the generator) are comparatively minor due to the low reeling speed. Although other phases, such as reel-in and hovering, also produce sound emissions, these are typically less prominent and exhibit greater variability due to differences in operational settings. By focusing on the reel-out phase, the study ensured that the results reflected a consistent and acoustically relevant scenario, providing a reliable basis for assessing noise annoyance. We have clarified this in Section 2.1.
5) 256-257: Please elaborate on why the circular flight pattern results in a bump above 1 kHz. How does this explanation relate to the one from l.274-275, stating the bump is a result of rigid materials?
We appreciate the opportunity to clarify this point. We now specified in the results that the circular flight patterns of AWESs B and C could cause the bump sound pattern, possibly due to induced specific turbulent flow characteristics around the kite's surfaces and structures for frequencies higher than 1 kHz. Additionally, at the end of the paragraph of Figure 5, we mentioned that for B3, the noise bump in the 1-2 kHz range might be due to vortex shedding frequencies around the kite's body or edges and that for C3, the additional components introduced by the turbine could shift these frequencies upwards to the 2-3 kHz range.
6) Section 3.2: Is there a criterium applied to judge which differences are meaningful? Regarding the loudness of kite C, for example, the text says that C3 exhibits higher values than C1. Looking at Figure 6 this is true regarding the maximum values, though not with respect to the distribution or median values. If the statement is based on the 5th percentile what is the threshold of a meaningful difference?
Thank you for raising this point. In psychoacoustic studies, differences in individual SQMs are considered meaningful if they are larger than a threshold value called Just Noticeable Difference (JND). These threshold values can be found in the literature, e.g., Osses et al., 2023. We computed the JND for A1, B1, and C1 as references and compared them to the values from the literature. The bar plots below display the JND per metric to show what deltas are significant (i.e. perceived as different). If desired, we can include the plots in the supplements.
[1] Osses, A., Greco, G.F., and Merino-Martinez, R., “Considerations for the perceptual evaluation of steady-state and time-varying sounds using psychoacoustic metrics,” 10th Convention of the European Acoustics Association (Forum Acusticum), 11-15 September 2023, Torino, Italy.
7) 329: Please add some words of guidance to Figure 8, help the reader interpret how strong the effect is.
We appreciate the feedback. We have added a description of the relationship in the text to guide readers through Figure 8. Specifically, we wrote that “Fig. 8 displays a significant and strong relationship between sharpness and annoyance (r = 0.863, p = 0.002)
8) It was not the aim of this study to compare annoyance levels of AWESs to other emission sources. But can you provide some point of comparison to understand whether the annoyance levels are particularly high or low? At least, it might be worth giving a verbal orientation on how to interpret the annoyance values within the scale (l.314-315).
Thank you for this suggestion. While direct comparisons to other noise sources were beyond the scope of this study, we agree that providing context is helpful. We have added comparisons with Merino-Martínez et al. (2021) and Schäffer et al. (2016), who assessed annoyance for wind turbine sound in laboratory settings under similar conditions. Specifically, we note: ”The mean annoyance ratings for the different AWES types ranged from approximately 34 for AWES A to 54 for AWES C (Fig. 7). In comparison, Merino-Martínez et al. (2021) reported average annoyance ratings of about 61-72 (converted from the ICBEN 11-point scale to a 0-100 scale) for wind turbine sound in a laboratory experiment, although the LAeq values in their study were lower, at around 38 dBA, than those used here. The percentage of highly annoyed participants (%HA) per AWES type varied between approximately 7% and 19% (Table 4), with AWES C showing the highest %HA, followed by AWES B and then A. This trend aligns with the previously reported higher tonality and sharpness values for AWES C compared to B and A. The observed %HA range is slightly narrower than the 2 to 34% of HA predicted by Schäffer et al (2016) for wind turbine sound exposure in laboratory settings with an LAeq range of 35 to 45 dBA.”
Unfortunately, we couldn’t find more relevant studies because existing research on wind turbine noise annoyance varies significantly in methodology, including differences in annoyance scales, thresholds for defining "highly annoyed," sound pressure levels, and study settings (laboratory vs. field).
9) You state that the moderation effects might be tied to the sample. There are more specific explanations possible. Familiarity with AWE at a technical university with ties to AWE development might imply a more positive attitude, hence affecting annoyance ratings. On the other hand, the sample is particularly young, raising questions about the effect of age on tonality. Given the age distribution, what age differences result in the effect of age on tonality?
We agree with you that participants’ familiarity with AWESs may reflect a more positive attitude, leading to lower annoyance ratings. We have added the following: “This familiarity may be intertwined with more positive attitudes toward AWE, potentially explaining the lower levels of noise annoyance observed—a pattern consistently reported in studies on wind turbines.”
Regarding age, we acknowledge that the sample is relatively young, which may limit our ability to generalize the age-related effects on tonality. That is why we mention in the Discussion that the moderation effects should be viewed cautiously due to the non-probability sampling used and the sample’s lack of representativeness. Future studies with a more balanced age distribution would help clarify this relationship.
TECHNICAL CORRECTIONS:
10) 14 & 75: Mostly, in this paper the word “acoustic” is used. In these two instances it is instead “acoustical”. I do not think this is to indicate any semantic difference. Perhaps you might want to keep it consistent.
We have corrected this inconsistency, using “acoustic” throughout the manuscript.
11) 25-27: “during natural disasters“ sounds like the kites are in the air while there is a heavy storm. I assume, you mean something like “following I will schedule some time for us to connect. in the aftermath of natural disasters”?
We have clarified this to “in the aftermath of natural disasters” to avoid ambiguity.
12) 27: I suggest “lower mass” instead of “mass savings”. Also, it should be stated what AWESs are compared to.
Thank you for the suggestion. We have revised the sentence to: “With its substantially lower mass compared to conventional wind turbines, AWE also has a smaller environmental footprint (Hagen et al., 2023).”
13) 33: This is the first time the abbreviation AWES is used. It is introduced a few lines later on l.41.
We appreciate that you noticed the mistake. The abbreviation AWES is now introduced upon its first mention in the Introduction.
14) 55: I suggest “enhance” instead of “enhances” as it refers to multiple aspects.
Corrected as suggested.
15) 155-156: I suggest “in the analyses this variable was not considered as a confounding factor”. Please clarify that this refers only to experiences with AWE sounds as familiarity is considered in later analyses.
We have revised the sentence to: “However, this familiarity was largely theoretical, as most participants had not heard AWES sounds prior to the experiment. Therefore, experience with AWES sounds was not considered as a confounding factor in the analyses.”
16) Figure 3: There is no indication what the color grading represents.
We apologize for the confusion. The original label got covered up during formatting. We have changed the layout so that it is visible again.
17) 230-231 “Examining the spectrograms...”: It is suggested that this refers to spectrograms of other recordings than the ones displayed here. If so this should be clarified.
Thank you for the feedback. We didn’t mean to imply spectrograms other than the ones displayed and discussed. We have changed the sentence to avoid misunderstandings to “The spectrograms confirmed that the recordings were representative.”
18) 251-252: Is the order of SPL below 2 kHz correct? Judging from Figure 4 the levels of C1 look to be higher than those of C2.
To be more precise, we changed 2 kHz to 1.6 kHz. Indeed, C1 presents higher levels than C2. The text has been corrected.
19) Description of Figure 6: Although 5th percentile seems to be a standard denotation, next to the explanation of the 25th and 75th percentiles I find it quite irritating as both naming conventions count from different directions. Indicating that the highest 5% are highlighted might avoid confusion.
We have clarified in the figure caption that the 5th percentile values are the values exceeded 5% of the signal time, as explained in section 2.4.1.
20) 305: There seems to be a mismatch between text and table regarding the %HA.
Thank you for pointing this out. There were indeed some mistakes. The text and table values have been updated.
21) Figure 7: In some boxes the central horizontal line is hard to see. I suggest the lines and frames of the boxes be black rather than match the color.
We have adjusted the figure to use black lines for improved visibility.
22) 369-370: There is a contradiction to the results chapter. Here it says that the fixed-wing kites show higher loudness, whereas in the results the soft-wing kite had the highest loudness values.
We have corrected the text to reflect that the soft-wing kite had the highest loudness values.
Citation: https://doi.org/10.5194/wes-2024-125-AC1
-
RC3: 'Comment on wes-2024-125', Anonymous Referee #2, 08 Jan 2025
The manuscript presents an original study that makes use of a listening experiment to explore noise annoyance of different airborne wind energy systems as emerging renewable energy devices. In doing so, it sheds new light on the common issue of noise emitted from wind energy facilities by analysing psychoacoustic metrics in relation to technological, operational and design-related aspects of airborne wind systems. The objective of the study is two-fold. The paper seeks to investigate the extent to which perception-based sound metrics can predict noise annoyance caused by airborne wind energy systems, while juxtaposing and comparing the noise annoyance of sound emissions from three different airborne wind energy systems. The manuscript addresses an original topic, is well written and clearly organised, and presents novel findings.
However, there are a couple of issues that should be clarified in the manuscript. Moreover, the presentation of the research can be improved overall by embedding the study more thoroughly in the literature as well as by adding a more critical and broader discussion of the findings.
While the literature review in the introduction is short and concise, it would benefit from a greater engagement with more critical research on the assessment of wind turbine noise embedded in the social acceptance literature: For example Haggett (2012) ’Social experience of noise from wind farms’; Dällenbach & Wüstenhagen (2015) ‘How far do noise concerns travel’; Kirkegaard et al. (2025) this journal. The present study seems to divert from the bulk of previous research dealing with noise annoyance and seems to (some extent) speak to this literature not only by focussing on the initial design of an emerging technology, but also by innovatively unpacking the sound quality beyond the loudness of wind turbine sound. It would therefore also be worthwhile to read more about the extent to which the study design, esp. sound quality metrics, have been previously applied in researching noise annoyance from conventional wind turbines and if so, how the findings differ.
Although the description of the methodology provides a lot of detail, it is not entirely clear how sound quality metrics were employed in the study. In the text following line 160, it says that the five metrics were calculated for each sound wave of every recording and it is described how each metric has been calculated. However, it remains unclear how these metrics relate to the participants’ perception of the sound emissions, i.e. if or to what extent participants were knowledgeable about these metrics or had the possibility to engage with or refer to such characteristics of sound quality when articulating their annoyance on the verbal scale. Lines 297-299 hint at some interesting terms that describe different degrees of these characteristics, but there is no indication of how these terms have been applied in the participants’ assessment and rating of AWES noise. This should be clarified. Similarly, I wonder whether participants’ evaluation of individual recordings (line 121) allowed for the possibility of linking their expressions of particular levels of annoyance with certain events in the sound wave.
The conclusion is very short and compact, merely summarising key results and briefly addressing the objective related to the metrics predicting annoyance. A discussion would therefore benefit from a more thorough engagement with the existing literature and much more detail on the significance and wider implications of the findings for the (future) role and utilisation of different airborne wind systems, but also people’s acceptance of novel wind energy technologies (or designs) and annoyance of wind turbine sound. In addition, a few more words about the relevance and potential replicability of the lab experiment in a field study would be highly interesting. A greater contextualisation of the study and its findings would be necessary to strengthen its purpose and value. Otherwise, the interesting insights and novel findings would remain rather abstract.
Further comments:
- It would be helpful for the reader to better understand and interpret the magnitude of and annoyance of sound emissions from AWES by having a comparison to those of other sources, such as conventional wind turbines or drones. How comparable are the insights gathered by drawing on sound quality metrics of AWES?
- Line 126: The definition of annoyance is not entirely clear or at least a bit infelicitously formulated. First, it is tautological, as annoyance is defined by referring to the term annoyance in its explanation (line 127). Second, it seems as if the defined term should be ‘noise annoyance’ rather than annoyance. Thus, I recommend to rephrase the definition.
- Line 142: it would be useful for the reader to see a list of ‘items covering noise sensitivity in different situations’
- Lines 153-155: Having a sample where roughly 50% are familiar of airborne wind seems unusually high. In turn, the fact that only few participants have heard the sound of an AWES before would presumably be even more pronounced in a field experiment, given the novelty of the technology and the fact that only very few people have seen an AWES and have been consciously exposed to its sound. This raises questions about the familiarity of participants with AWES. How does the recruitment of younger participants from a technical university affect the familiarity with AWES and attitude to AWES? Would the attitude be different in a random sample that includes less familiarity? It would be useful to critically consider these issues here or in the discussion/conclusion.
- Lines 283-285. This sentence requires more explanation. It is not clear why C3 is obviously different from C1 and C2 (Fig4). How is the higher loudness of C3 related to the ‘frequency-dependent human ear sensitivity’? Does this mean that the recording of C3 recorded a louder sound of the reel-out phase?
- Lines 297-299. With reference to my comment above, the terminology highlighted in these two sentences seems to be crucial for qualifying the different perception-based sound metrics (e.g. harshest, strongest beating …), but it is not obvious where these terms are coming from and how they have been applied in the study. Were these terms included in the verbal scale, were they articulated or rated by the participants, or were they merely used by the researchers to illustrate the results?
- Are there any further limitations in the study that should be mentioned? Are there any challenges concerning the recordings, considering the distance to the microphone and moving kites (also compared to the recording of the sound from conventional turbines)?
Minor comments:
- Line 27: If mentioned, it is perhaps useful to elaborate a bit more on mass savings and how they relate to capacities and energy production of AWE in comparison to conventional wind energy?
- Line 28: Explain how and why AWE is seen as a potential game-changer in the energy transition.
- Line 61-62: It would be interesting to know what AWES was used at this test site, fixed-wing or soft-wing.
- Line 65: It may be better to write ‘sound emissions of different AWES’ instead of ‘AWES sound emissions’ in order to hint at the scope of your study
- Line 69: What is meant by global in global Psychoacoustic Annoyance (PA)? The term global is not used again in the remainder of the manuscript.
- Line 89: Normalisation and A-weighting should be explained for the unfamiliar reader.
- Line 109: the caption of the figure requires a source, even if it is an own photograph
- Line 138: ‘reasonably valid’ sounds more like guesswork rather than scientific severity
- Line 152: the information about the duration of each experiment could be moved forward to section 2.2.2, as I was wondering about that while reading this section
- Line 160: the verb ‘describe’ annoyance is used here, whereas the term ‘predict’ annoyance is predominantly used later on (e.g. lines 355-362). Is there a difference in the meaning? If not, it might be preferable to use one verb consistently in order to avoid confusion for the unfamiliar reader
- Line 175: Di et al. (2016) does not seem to appear in the bibliography
- Line 335-336: Elaborate a bit on this sentence. What does ‘varying effects on annoyance rating among individuals’ mean?
- Lines 355 and 359: add years to references and ‘et al.’ to Di
Citation: https://doi.org/10.5194/wes-2024-125-RC3 -
AC2: 'Reply on RC3', Helena Schmidt, 14 Jan 2025
Reviewer comment
Response to reviewer
1) While the literature review in the introduction is short and concise, it would benefit from a greater engagement with more critical research on the assessment of wind turbine noise embedded in the social acceptance literature: For example Haggett (2012) ’Social experience of noise from wind farms’; Dällenbach & Wüstenhagen (2015) ‘How far do noise concerns travel’; Kirkegaard et al. (2025) this journal. The present study seems to divert from the bulk of previous research dealing with noise annoyance and seems to (some extent) speak to this literature not only by focussing on the initial design of an emerging technology, but also by innovatively unpacking the sound quality beyond the loudness of wind turbine sound. It would therefore also be worthwhile to read more about the extent to which the study design, esp. sound quality metrics, have been previously applied in researching noise annoyance from conventional wind turbines and if so, how the findings differ.
Thank you for your thoughtful comment. In response, we have expanded the introduction to engage more deeply with research on the social acceptance of wind turbine noise, emphasizing the broader context of noise annoyance and its connection to psychosocial and individual factors. This expanded discussion highlights the complexity of noise perception and its relevance to the acceptance of wind energy technologies.
We also strengthened the rationale for focusing on sound quality metrics (SQMs) in the context of AWESs by underscoring the innovative contribution of this approach. Unlike conventional studies that primarily emphasize overall loudness, our study seeks to "unpack" the perceptual dimensions of sound, such as sharpness and tonality.
Furthermore, in the revised Discussion section, we provide additional context on how our findings diverge from or align with existing research on wind turbine noise.
2) Although the description of the methodology provides a lot of detail, it is not entirely clear how sound quality metrics were employed in the study. In the text following line 160, it says that the five metrics were calculated for each sound wave of every recording and it is described how each metric has been calculated. However, it remains unclear how these metrics relate to the participants’ perception of the sound emissions, i.e. if or to what extent participants were knowledgeable about these metrics or had the possibility to engage with or refer to such characteristics of sound quality when articulating their annoyance on the verbal scale. Lines 297-299 hint at some interesting terms that describe different degrees of these characteristics, but there is no indication of how these terms have been applied in the participants’ assessment and rating of AWES noise. This should be clarified. Similarly, I wonder whether participants’ evaluation of individual recordings (line 121) allowed for the possibility of linking their expressions of particular levels of annoyance with certain events in the sound wave.
We appreciate your comments and the opportunity to clarify the methodology. The study did not assume that participants were knowledgeable about specific SQMs, such as sharpness, loudness, or tonality, nor were they informed of these metrics during the experiment. Instead, these metrics were derived objectively from the acoustic analysis of the recordings. The intention was to identify whether these objectively measured characteristics could predict participants' subjective annoyance ratings (see Section 2.2.3). We have added this explanation at the end of the Introduction, so that readers unfamiliar with SQMs immediately grasp the study rationale.
The descriptive terms mentioned (e.g., “harsh” and “beating”) were used to interpret the results of the acoustic analysis and were not terms provided by participants during the experiment. We acknowledge that this distinction could be clearer in the manuscript and have added a sentence at the end of Section 2.4.1, “Acoustic analyses,” stating that such terminology arose from the researchers’ analysis rather than participant feedback.
Regarding your last point, we want to clarify that participants provided a single annoyance rating for each entire recording, rather than for specific segments within a recording. As a result, our analysis focused on comparing how sound recordings with relatively higher or lower values for certain SQMs (e.g., sharpness, loudness, or tonality) were rated in terms of overall annoyance across recordings. This approach allowed us to investigate whether specific acoustic characteristics of the recordings, as quantified by SQMs, influenced participants’ average annoyance ratings. However, because participants evaluated the recordings as a whole, we did not assess their annoyance in relation to specific events or changes within the sound wave of an individual recording. We hope this clarifies the methodology, but we are happy to provide further details if needed.
3) The conclusion is very short and compact, merely summarising key results and briefly addressing the objective related to the metrics predicting annoyance. A discussion would therefore benefit from a more thorough engagement with the existing literature and much more detail on the significance and wider implications of the findings for the (future) role and utilisation of different airborne wind systems, but also people’s acceptance of novel wind energy technologies (or designs) and annoyance of wind turbine sound. In addition, a few more words about the relevance and potential replicability of the lab experiment in a field study would be highly interesting. A greater contextualisation of the study and its findings would be necessary to strengthen its purpose and value. Otherwise, the interesting insights and novel findings would remain rather abstract.
Thank you for your valuable feedback. In response, we have added a comprehensive discussion section to provide a more thorough engagement with the existing literature and to contextualize the study’s findings. The new section summarizes the key results and compares them to prior research on wind turbines and drones. We also address the limitations of our findings, particularly regarding their applicability to real-world settings. Furthermore, we outline specific recommendations for future research and derive actionable insights for industry, emphasizing design modifications and noise mitigation strategies to enhance the social acceptance of AWESs. These additions aim to strengthen the study’s purpose, highlight its novel contributions, and demonstrate its broader implications for the development and acceptance of AWE.
FURTHER COMMENTS:
4) It would be helpful for the reader to better understand and interpret the magnitude of and annoyance of sound emissions from AWES by having a comparison to those of other sources, such as conventional wind turbines or drones. How comparable are the insights gathered by drawing on sound quality metrics of AWES?
Thank you for this suggestion. While direct comparisons to other noise sources were beyond the scope of this study, we agree that providing context for the observed annoyance levels is helpful. We have added a comparison with two studies that assessed annoyance for wind turbine sound in laboratory settings under similar conditions. Specifically, we note: ”The mean annoyance ratings for the different AWES types ranged from approximately 34 for AWES A to 54 for AWES C (Fig. 7). In comparison, Merino-Martínez et al. (2021) reported average annoyance ratings of about 61-72 (converted from the ICBEN 11-point scale to a 0-100 scale) for wind turbine sound in a laboratory experiment, although the LAeq values in their study were lower, at around 38 dBA, than those used here. The percentage of highly annoyed participants (%HA) per AWES type varied between approximately 1% and 19% (Table 4), with AWES C showing the highest %HA, followed by AWES B and then A. This trend aligns with the previously reported higher tonality and sharpness values for AWES C compared to B and A. The observed %HA range is slightly narrower than the 2 to 34% of HA predicted by Schäffer et al (2016) for wind turbine sound exposure in laboratory settings with an LAeq range of 35 to 45 dBA.”Unfortunately, we couldn’t find more relevant studies because existing research on wind turbine noise annoyance varies significantly in methodology, including differences in annoyance scales (e.g., 5-point vs. 11-point), thresholds for defining "highly annoyed," sound pressure levels, and study settings (laboratory vs. field). Regarding comparisons with other studies on SQMs, we address this point in response to your first comment.
5) Line 126: The definition of annoyance is not entirely clear or at least a bit infelicitously formulated. First, it is tautological, as annoyance is defined by referring to the term annoyance in its explanation (line 127). Second, it seems as if the defined term should be ‘noise annoyance’ rather than annoyance. Thus, I recommend to rephrase the definition.
We appreciate your observation and agree that the definition can be clarified to avoid tautology and better align with the term "noise annoyance." We propose the following revisions to address this concern: "Noise annoyance was defined in accordance with the ISO 15666 standard as an individual’s adverse reaction to noise, which may include feelings of dissatisfaction, bother, or disturbance caused by noise exposure (International Organization for Standardization, 2021)."
6) Line 142: it would be useful for the reader to see a list of ‘items covering noise sensitivity in different situations’
We agree that including a few sample items from the NoiSeQ scale would improve the clarity and comprehensiveness of the manuscript. However, we have opted not to list all the items to avoid unnecessarily lengthening the text. To address this, we propose the following revised wording: “Noise sensitivity was assessed using the condensed 12-item version of the NoiSeQ scale. Participants rated their agreement with statements related to noise sensitivity in various contexts on a 4-point scale, ranging from “strongly disagree” (0) to “strongly agree” (3) (Griefahn, 2008). Sample items are “When I am at home, I quickly get used to noise” and “When people around me are noisy, I find it hard to do my work”.
7) Lines 153-155: Having a sample where roughly 50% are familiar of airborne wind seems unusually high. In turn, the fact that only few participants have heard the sound of an AWES before would presumably be even more pronounced in a field experiment, given the novelty of the technology and the fact that only very few people have seen an AWES and have been consciously exposed to its sound. This raises questions about the familiarity of participants with AWES. How does the recruitment of younger participants from a technical university affect the familiarity with AWES and attitude to AWES? Would the attitude be different in a random sample that includes less familiarity? It would be useful to critically consider these issues here or in the discussion/conclusion.
We value your feedback and agree that participant familiarity with AWES warrants further elaboration. The higher level of familiarity with AWES in our sample can be attributed to the presence of a highly active and renowned research group on airborne wind energy at our faculty. Many participants are students or employees and may have been indirectly exposed to the concept of AWES through institutional activities or research dissemination. However, it is important to note that this familiarity was primarily theoretical. Given the limited number of operational test sites worldwide, most participants had never encountered an AWES in real life, seen detailed footage, or heard the sound emissions prior to the experiment. This explains why only a small number of participants had previously heard the sound of an AWES. We added this explanation in Section 2.3 “Participant characteristics”.
We acknowledge that participant familiarity may differ in a random sample. However, although the proportion of people familiar with AWE is likely higher in our sample than in the general population, there was still an even split between those familiar and unfamiliar in the sample. This balance allowed us to compare the effect of familiarity on annoyance ratings within both the familiar and unfamiliar subgroups within a population that may generally be more exposed to such technology. Moreover, the main effect of familiarity on annoyance was statistically insignificant, suggesting that familiarity alone did not significantly influence annoyance ratings. Regarding field studies, they typically focus on individuals who have been exposed to the sound emissions of the technology in question. Familiarity with AWES, in this context, would likely be a prerequisite for obtaining meaningful results regarding real-world noise annoyance.
The first reviewer also noted a potential entanglement of familiarity with attitudes in that participants’ familiarity with AWESs may reflect a more positive attitude, leading to lower annoyance ratings. To address this point, we have added the following explanation in the Results: “This familiarity may be intertwined with more positive attitudes toward AWE, potentially explaining the lower levels of noise annoyance observed—a pattern consistently reported in studies on wind turbines.”
8) Lines 283-285. This sentence requires more explanation. It is not clear why C3 is obviously different from C1 and C2 (Fig4). How is the higher loudness of C3 related to the ‘frequency-dependent human ear sensitivity’? Does this mean that the recording of C3 recorded a louder sound of the reel-out phase?
In Figure 4c, the spectra of C1, C2, and C3 are plotted. A noticeable difference in sound pressure levels (Lp) is observed for frequencies below 1.6 kHz, particularly for C2, which exhibits lower Lp compared to C1 and C3. Additionally, within this frequency range, C3 elucidates peaks that suggest a tonal behavior. Human hearing is most sensitive to frequencies between 2 kHz and 5 kHz, as illustrated by the equal-loudness contours (ISO 226). Observing Figure 4c, it can be noted that C3 displays higher Lp values in the frequency range of 3–5 kHz.
9) Lines 297-299. With reference to my comment above, the terminology highlighted in these two sentences seems to be crucial for qualifying the different perception-based sound metrics (e.g. harshest, strongest beating …), but it is not obvious where these terms are coming from and how they have been applied in the study. Were these terms included in the verbal scale, were they articulated or rated by the participants, or were they merely used by the researchers to illustrate the results?
Thank you for raising this point. As explained in response to your earlier comment, these descriptive terms were used to interpret the results of the acoustic analysis and were not terms provided by participants during the experiment. We acknowledge that this distinction could be clearer in the manuscript and have added a sentence at the end of “Section 2.4.1 Acoustic analyses,” stating that such terminology arose from the researchers’ analysis rather than participant feedback.
10) Are there any further limitations in the study that should be mentioned? Are there any challenges concerning the recordings, considering the distance to the microphone and moving kites (also compared to the recording of the sound from conventional turbines)?
Thank you for your suggestion to add a limitation section. We have done so. The following part pertains to the aspects you mentioned: “[...] the study's methodology faced several challenges related to sound recordings, particularly concerning the varying distances to the microphone (100–700 m) and the moving nature of the kites. While normalization to 45 dBA mitigated some inconsistencies, the dynamic sound signatures created by the kites’ flight patterns introduced additional variability compared to the noise emissions of stationary wind turbines. Additionally, the location of the observer or microphone significantly influences noise perception because the acoustic prominence of different system components varies depending on the vantage point. For example, certain components, such as the kite, may dominate acoustically when the observer is positioned directly below or in line with the kite’s trajectory. In contrast, noise from the generator or tether vibrations may become more prominent at close distances to the ground station. Furthermore, environmental factors such as wind noise and ground reflections may have influenced the recordings despite mitigation efforts using windscreens. These limitations underscore that the results have only restricted applicability to the field. Schäffer et al. (2016) highlighted these challenges, emphasizing that laboratory and field studies should be viewed as complementary rather than directly comparable.”
MINOR COMMENTS:
11) Line 27: If mentioned, it is perhaps useful to elaborate a bit more on mass savings and how they relate to capacities and energy production of AWE in comparison to conventional wind energy?
The first reviewer also wondered what AWESs are compared to here, so we revised the sentence: “With its substantially lower mass compared to conventional wind turbines, AWE also has a smaller environmental footprint.”
For the sake of conciseness, we refrain from giving more details in the Introduction but refer you and other interested readers to the following two sources for more information:
Hagen et al. (2023) found that future large-scale hard-wing AWE systems deployed in onshore wind parks could save 70% of the mass of a comparable HAWT park. The global warming potential (GWP) and cumulative energy demand (CED) were 60% and 65% of the respective impacts of the HAWT. Coutinho (2024) investigated mobile soft-wing AWE systems and found that the mass savings were also significant but less pronounced. Similarly, the GWP and CED were reduced less than the HAWT installations.12) Line 28: Explain how and why AWE is seen as a potential game-changer in the energy transition.
We appreciate your suggestion and have expanded on this point to clarify why AWE is considered a game-changer: “While AWE is regarded as a potential game-changer for the energy transition because it can harness higher-altitude winds, requires fewer materials compared to wind turbines, and can be deployed in remote or distant offshore locations (IRENA, 2021), the technology has not yet converged towards a single standard configuration.”
13) Line 61-62: It would be interesting to know what AWES was used at this test site, fixed-wing or soft-wing.
Thank you for pointing this out. We have specified that a soft-wing AWES is used at the test site.
14) Line 65: It may be better to write ‘sound emissions of different AWES’ instead of ‘AWES sound emissions’ in order to hint at the scope of your study
The sentence in question refers to the field study by Schmidt et al. The study only investigated one type of AWES. Therefore, we will retain the phrasing.
15) Line 69: What is meant by global in global Psychoacoustic Annoyance (PA)? The term global is not used again in the remainder of the manuscript.
We appreciate your attention to detail. The term 'global' in Global Psychoacoustic Annoyance (PA) refers to a composite metric that integrates multiple sound quality metrics to estimate annoyance, as explained. We have removed the word global as it is redundant.
16) Line 89: Normalisation and A-weighting should be explained for the unfamiliar reader.
We added a brief explanation of these terms: "Normalization refers to adjusting the sound pressure levels of recordings to a common reference value, ensuring comparability. A-weighting is a standard method to adjust sound measurements to reflect the human ear's sensitivity to different frequencies (approximately between 2 and 5 kHz)."
17) Line 109: the caption of the figure requires a source, even if it is an own photograph
We added a source to the figure caption, specifying that it is our photograph.
18) Line 138: ‘reasonably valid’ sounds more like guesswork rather than scientific severity
We have rephrased this to convey scientific rigor: "Self-evaluations have been shown to provide a valid measure of individual hearing ability in the absence of audiometric testing (Hong et al., 2011)."
19) Line 152: the information about the duration of each experiment could be moved forward to section 2.2.2, as I was wondering about that while reading this section
We agree and have relocated the information to Section 2.2.2.
20) Line 160: the verb ‘describe’ annoyance is used here, whereas the term ‘predict’ annoyance is predominantly used later on (e.g. lines 355-362). Is there a difference in the meaning? If not, it might be preferable to use one verb consistently in order to avoid confusion for the unfamiliar reader
We standardized the terminology to avoid confusion. Since "predict" is statistically more precise, we use it consistently.
21) Line 175: Di et al. (2016) does not seem to appear in the bibliography
Thank you for noticing. We have added the missing reference to the bibliography.
22) Line 335-336: Elaborate a bit on this sentence. What does ‘varying effects on annoyance rating among individuals’ mean?
We have clarified this point: “This suggests that all tested SQMs, except for fluctuation strength, influenced annoyance ratings differently across individuals. These variations may reflect individual differences in factors such as noise sensitivity, age, or familiarity with AWES, which can shape how participants perceive and react to specific sound qualities. To explore whether these individual characteristics could account for the observed differences, interaction effects between the SQMs and participant characteristics (i.e., age, AWE familiarity, and noise sensitivity) were included in the models with the random SQM effects.”
23) Lines 355 and 359: add years to references and ‘et al.’ to Di
We have corrected these citations to include the publication years and the proper "et al." formatting.
Citation: https://doi.org/10.5194/wes-2024-125-AC2
Data sets
Respondents' noise annoyance ratings for airborne wind energy systems measured in a laboratory listening experiment Helena Schmidt, Renatto Yupa Villanueva, Daniele Ragni, Roberto Merino-Martinez, Piet van Gool, and Roland Schmehl https://data.4tu.nl/private_datasets/11wCOgnga-GfjzL4oufzdd1UPRPP2YVFVnjRaKcHJhE
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 207 | 71 | 12 | 290 | 41 | 4 | 3 |
- HTML: 207
- PDF: 71
- XML: 12
- Total: 290
- Supplement: 41
- BibTeX: 4
- EndNote: 3
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1