Decoding the Role of Urban Green Space Morphology in Shaping Visual Perception: A Park-Based Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. Introduction

Existing literature has explored the relationship between green space morphology and health benefits to some extent. However, this paper fails to clearly distinguish its theoretical breakthrough points. At the beginning, too much emphasis is placed on the long - term symptoms of COVID - 19, which has a weak connection with the core topic of "green space morphology - visual perception". It is recommended to streamline the epidemiological background. The transition from health benefits to public perception is not natural. It is advisable to more clearly identify the research gap in the introduction and highlight its policy background.

 

2. Methods

The potential interference of the surrounding environment of green spaces on visual perception is not mentioned. It is recommended to include "spatial context factors" (such as land use mix) as covariates in the model.

 

The demographic characteristics (such as age, professional background) of the 150 volunteers are not described in detail, which may affect the objectivity of the perceptual evaluation. It is necessary to supplement the rationality explanation of the sampling method.

 

For landscape indicators, the definitions and calculation formulas of key parameters need to be supplemented.

 

The segmentation accuracy of Mask2Former on the ADE20K dataset is not clear. It is possible to supplement the IoU value to prove its reliability.

 

The visual entropy formula is questionable: in formula (4), the weight coefficient of 0.3 for the color feature parameter lacks a theoretical basis. It is recommended to cite relevant literature or conduct a sensitivity analysis.

 

3. Results

Figures 3 and 4 only show the regression coefficients, lacking confidence intervals or significance markings. It is recommended to supplement error bars or p - values to enhance the credibility of the results.

 

4. Discussion

The connection between the results and the discussion needs to be further strengthened. For example, for the finding that "there is no significant association between shape complexity and pleasure", the discussion does not deeply explain this phenomenon. It is suggested to add interpretations and supports from an interdisciplinary perspective (such as environmental psychology).

 

The conclusion mentions planning suggestions such as "connecting scattered green spaces through green corridors", but it does not combine specific cases or policy contexts. It is recommended to supplement practical design strategies.

Comments on the Quality of English Language

The overall English level of this article is good, meeting the requirements of academic papers. However, there are still some details that can be improved. For example, there are quite a number of long and complex sentences. 

Author Response

We sincerely appreciate reviewer 1 for their valuable feedback that we have used to improve the quality of our manuscript. We value the expert advice you provided and have made revisions throughout the text accordingly. We appreciate your enthusiastic efforts and hope that the revised manuscript meets your expectations.

Comments 1: Existing literature has explored the relationship between green space morphology and health benefits to some extent. However, this paper fails to clearly distinguish its theoretical breakthrough points. At the beginning, too much emphasis is placed on the long - term symptoms of COVID - 19, which has a weak connection with the core topic of "green space morphology - visual perception". It is recommended to streamline the epidemiological background. The transition from health benefits to public perception is not natural. It is advisable to more clearly identify the research gap in the introduction and highlight its policy background.

Response: Thank you for the reviewer’s suggestions. We have simplified the background discussion of COVID-19 to make the research topic more focused. Additionally, in the introduction, we have added a more logical explanation of how "green space morphology influences visual perception, and in turn, affects health," which makes the transition smoother. Furthermore, we have clarified the research gap in the introduction and included a discussion on the policy implications of the research findings, such as in urban planning and green infrastructure development, to enhance the practical application value of the study.

Comments 2: The potential interference of the surrounding environment of green spaces on visual perception is not mentioned. It is recommended to include "spatial context factors" (such as land use mix) as covariates in the model.

Response: Thank you for the reviewer’s suggestion. We agree that spatial context factors may influence visual perception, and we have added a discussion on their potential role in the discussion section. However, since the data in this study primarily focuses on the internal morphological characteristics of green spaces and lacks high-precision surrounding environment data, we did not directly incorporate relevant variables into the model. Our study emphasizes the impact of the internal morphology of green spaces on visual perception, and we will consider exploring the role of spatial context factors in future research.

Comments 3: The demographic characteristics (such as age, professional background) of the 150 volunteers are not described in detail, which may affect the objectivity of the perceptual evaluation. It is necessary to supplement the rationality explanation of the sampling method.

Response: Thank you for the reviewer’s suggestion. We have added detailed demographic information of the volunteers, including age distribution, professional background, etc., in the revised manuscript. This information has been included in the Methods section to improve the transparency and objectivity of the study.

Comments 4: For landscape indicators, the definitions and calculation formulas of key parameters need to be supplemented.

Response: The definitions of all key landscape indicators and their corresponding calculation formulas have been presented in the figures. We have also clarified the software tools and data sources used to calculate these indicators. These revisions help ensure that readers can accurately understand our analytical methods.

Comments 5: The segmentation accuracy of Mask2Former on the ADE20K dataset is not clear. It is possible to supplement the IoU value to prove its reliability.

Response:Thank you for the reviewer’s suggestion. Mask2Former was trained on 20,210 images from the ADE20K dataset, which includes complex daily scene images, with 2,000 images used for validation and 3,000 images for testing. The dataset contains various objects from natural spatial environments, with each image having an average of 19.5 instances and 10.5 object classes. The trained model achieves an excellent pixel-level accuracy of 84.59, which effectively meets the requirements of this study.

Comments 6: The visual entropy formula is questionable: in formula (4), the weight coefficient of 0.3 for the color feature parameter lacks a theoretical basis. It is recommended to cite relevant literature or conduct a sensitivity analysis.

Response: Thank you for the reviewer’s suggestion. In the revised manuscript, we have cited the relevant literature (Qin, Yang, and Wangari, 2024) to provide a theoretical basis for the weight coefficient of 0.3 for the color feature.

Comments 7: Figures 3 and 4 only show the regression coefficients, lacking confidence intervals or significance markings. It is recommended to supplement error bars or p - values to enhance the credibility of the results.

Response: Figure 3 shows the regression coefficients along with error bars. Figure 4 uses the asterisk "*" to mark the significance levels, clearly indicating the statistical significance of the results.

Comments 8: The connection between the results and the discussion needs to be further strengthened. For example, for the finding that "there is no significant association between shape complexity and pleasure", the discussion does not deeply explain this phenomenon. It is suggested to add interpretations and supports from an interdisciplinary perspective (such as environmental psychology).

Response:Thank you for your suggestion. Based on your feedback, we have further explained the finding that "there is no significant association between shape complexity and pleasure" in the discussion section. In this part, we explored this result from an environmental psychology perspective. Specifically, shape complexity is not the sole determinant of pleasure; other landscape elements, such as color and spatial layout, may have a more direct impact on emotional responses. Additionally, the theory of visual fatigue could also influence our findings, as excessive shape complexity in highly complex environments may lead to visual fatigue, thus reducing pleasure.

Comments 9: The conclusion mentions planning suggestions such as "connecting scattered green spaces through green corridors", but it does not combine specific cases or policy contexts. It is recommended to supplement practical design strategies.

Response:Thank you for your suggestion. To enhance the practicality of the planning recommendations, we have added a discussion of the Spanish case of green corridors (Iungman et al., 2025) in the discussion section. This highlights the potential of nature-based green corridor plans in creating more sustainable and health-promoting urban environments.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper titled Decoding the Role of Urban Green Space Morphology in Shaping Visual Perception investigates how the morphology of urban green spaces influences visual perception preferences. Using data from ten parks in Chengdu, China, the study used high-resolution satellite imagery to generate green space maps and calculate metrics such as quantity, fragmentation, connectivity, and shape complexity. The relationships between visual perception factors (comfort, pleasure, quietness, etc.) and landscape elements are analysed using a multi-objective regression model combined with structural equation modelling. The findings suggest that green space connectivity, aggregation, and shape complexity have a stronger influence on visual perception than green space quantity alone, emphasising the importance of spatial morphology in health-oriented urban design.

The paper addresses a relevant and timely topic, especially in the context of post-pandemic urban planning. The use of high-resolution satellite imagery and advanced analytical methods adds robustness to the findings. The arguments are well-supported by evidence, and the analysis is thorough. However, while the analysis is thorough, the paper could benefit from a deeper exploration of the underlying mechanisms behind the observed relationships. In addition, the paper could provide more detailed analysis on how different socio-demographic factors influence visual perception preferences.

My main concern about this paper is its originality, as similar studies have been conducted in Chengdu. This should be better highlighted in the manuscript to distinguish the unique contributions of this research.

Below I have included my specific comment:

Abstract:

Please include a brief mention of the international relevance of the study to highlight its broader applicability.  You should emphasise the innovative aspects of the methodology used.

The text includes typos and grammar errors: We analyzed the links between perception ns of comfort, 14 pleasure, quietness, and other factors with landscape element proportions, visual rich-15 ness, and entropy using a multi-objective regression model combined with structural 16 equation modeling, controlling for geographical and demographic variables.

Introduction

-Please expand the literature review to include more international references, particularly studies from Europe, North America, and other regions with diverse urban green space characteristics.

In addition, you should state the research gap and how this study addresses it in a global context.

Page 2. Urban green spaces provide opportunities for residents to connect with nature, promoting physical activity [6-8]. Cite more studies. Also include how green spaces improve healthy living.

Page 2: Perceptions and preferences regarding green spaces are closely linked to their restorative qualities. Cite several studies

Page 3: In this report change to in this paper…

Methods

Please provide a more detailed explanation of the selection criteria for the parks and justify the focus on Chengdu.

Discussion

You should discuss the findings with a comparative perspective, discussing how the results align or contrast with international studies.

In addition, highlight any unique findings that could be of interest to an international audience.

Furthermore, you should discuss underlying mechanisms behind the observed relationships. Include references to international studies that have explored similar mechanisms.

Finally discuss the implications of the findings for urban green space planning in different cultural and geographical contexts.

Limitations: Include all the limitations

Conclusion:

The conclusion section needs major rewriting. You should summarise the key findings with an emphasis on their international relevance.

Finally, you should provide specific recommendations for urban planners and policymakers, considering different international contexts.

 

References

Land has an international readership so you should provide a balanced mix of local and international references. Please include more studies from diverse regions to improve the paper's global relevance.

Author Response

We sincerely appreciate reviewer for their valuable feedback that we have used to improve the quality of our manuscript. We value the expert advice you provided and have made revisions throughout the text accordingly. We appreciate your enthusiastic efforts and hope that the revised manuscript meets your expectations.

Comments 1: Please include a brief mention of the international relevance of the study to highlight its broader applicability.  You should emphasise the innovative aspects of the methodology used.

Response: Thank you for your feedback. We have added a brief mention of the international relevance of the study to emphasize its broader applicability. This study’s findings can offer valuable insights for urban green space planning and design in cities worldwide, especially those undergoing rapid urbanization or seeking to improve the quality of public spaces.

Comments 2: The text includes typos and grammar errors: We analyzed the links between perception ns of comfort, 14 pleasure, quietness, and other factors with landscape element proportions, visual rich-15 ness, and entropy using a multi-objective regression model combined with structural 16 equation modeling, controlling for geographical and demographic variables.

Response: Thank you for your valuable feedback. We appreciate your attention to detail regarding the typos and grammar issues. However, we believe that the text you encountered may have been affected during translation or formatting, rather than being an issue with the original manuscript.

Comments 3: Please expand the literature review to include more international references, particularly studies from Europe, North America, and other regions with diverse urban green space characteristics.

Response: Thank you for your valuable feedback. We have expanded the literature review to include additional international references. These new references provide a broader perspective on how green space morphology influences visual perception in different cultural and environmental contexts. We believe this expansion enhances the comprehensiveness of the review and strengthens the overall foundation of the study. 

Comments 4: In addition, you should state the research gap and how this study addresses it in a global context.

Response: Thank you for your insightful comment. We have revised the manuscript to clearly highlight the research gap and explain how this study addresses it in a global context.

The existing literature on urban green spaces predominantly focuses on general relationships between green space and well-being, but there is limited research specifically examining how the morphological characteristics of green spaces, such as connectivity, aggregation, and shape complexity, influence visual perception preferences. Additionally, there is a lack of comprehensive studies that explore the generalizability of these relationships across different urban contexts, especially in cities with diverse green space characteristics.

This study addresses these gaps by integrating data from 10 parks in different urban settings, using high-resolution satellite imagery and advanced statistical methods, including a multi-objective regression model and structural equation modeling. By doing so, we provide evidence on how various morphological characteristics of green spaces independently influence visual perception, and we explore the applicability of these findings in diverse urban environments. This contributes to the global understanding of how urban green space design can be optimized for enhancing human well-being, especially in the context of rapidly urbanizing cities worldwide. 

We hope this clarification helps in emphasizing the global relevance and innovation of the study.

Comments 5: Urban green spaces provide opportunities for residents to connect with nature, promoting physical activity [6-8]. Cite more studies. Also include how green spaces improve healthy living.

Response: Thank you for your insightful comment.We have expanded the literature review to include additional references on environmental psychology and its role in promoting health. These studies emphasize how the physical environment, particularly green spaces, influences psychological well-being, reduces stress, and enhances overall health outcomes. We believe that incorporating these references enriches the context of our study and underscores the importance of green space design in fostering both physical and mental health.

Comments 6: Page 2: Perceptions and preferences regarding green spaces are closely linked to their restorative qualities. Cite several studies

Response: Thank you for your valuable suggestion. We have included studies on environmental psychology promoting health in the literature review, which highlight the connection between perceptions and preferences regarding green spaces and their restorative qualities. Research shows that exposure to nature, especially green spaces, helps reduce stress, improve mood, and promote overall psychological restoration. These studies support the idea that green spaces not only offer aesthetic value but also provide significant restorative benefits that enhance well-being.

Comments 7: Page 3: In this report change to in this paper…

Response: Updated.

Comments 8: Please provide a more detailed explanation of the selection criteria for the parks and justify the focus on Chengdu.

Response: Thank you for your valuable feedback. This study indeed focuses primarily on Chengdu, a representative city in southwest China, which has unique climatic, cultural, and urban planning characteristics. These factors may limit the generalizability of the findings to other regions. The results from Chengdu may not be fully applicable to areas with different cultural contexts, such as northern cities. We acknowledge this limitation and suggest that future research could conduct comparative studies in regions with distinct cultural differences to explore how culture influences green space morphology and visual preferences.

Although this study is centered on Chengdu, we believe the findings can still provide valuable insights for other rapidly urbanizing areas, particularly cities in central and western China. To enhance the broader applicability of the research, we suggest future studies expand their sample to include more cities (such as coastal cities or cities of different types). Moreover, cross-cultural comparative studies would help validate the generalizability of our findings and provide more targeted policy recommendations for urban green space planning and design in different regions.

Comments 9: You should discuss the findings with a comparative perspective, discussing how the results align or contrast with international studies.

Response: Thank you for your insightful suggestion. As noted in the manuscript, we have made an effort to incorporate a comparative perspective by discussing how our findings align with or contrast to studies conducted internationally. We believe this approach helps to contextualize our results within the broader research landscape, though we are happy to elaborate further if necessary.

Comments 10: In addition, highlight any unique findings that could be of interest to an international audience.

Response: Thank you for your valuable suggestion. We have highlighted several unique findings in the manuscript that we believe could be of interest to an international audience. Specifically, our research offers new insights into how specific green space morphologies, such as connectivity and shape complexity, influence visual perception at the individual level. These findings contribute to the growing body of knowledge on the relationship between urban green space design and well-being, particularly in regions where such relationships have not been extensively studied. We hope that these novel results will add value to international discussions on the role of urban green spaces in promoting human health and well-being.

Comments 11: Furthermore, you should discuss underlying mechanisms behind the observed relationships. Include references to international studies that have explored similar mechanisms.

Response: Thank you for your suggestion. We have addressed potential mechanisms linking green space morphology and visual perception preferences, emphasizing the role of visual richness and entropy. Our study found that landscape element proportions mediate the relationship between green space morphology and visual perception. Existing research supports our findings, indicating that diverse green space elements can enhance cognitive function (Ricciardi et al., 2022) and emotional responses (Yin et al., 2023b; Chen et al., 2024a). Additionally, higher visual entropy and richness reflect increased complexity in visual perception, which is influenced by element density and spatial organization. We agree that further longitudinal studies would provide additional insights into these mechanisms.

Comments 12: Finally discuss the implications of the findings for urban green space planning in different cultural and geographical contexts.

Response: Thank you for the suggestion. We have included a discussion on the implications of our findings for urban green space planning in different cultural and geographical contexts. Our study highlights the importance of green space morphology in influencing visual perception and well-being. Given that cultural and geographical contexts shape people’s interactions with and preferences for green spaces, urban planners should consider local values, aesthetic preferences, and environmental conditions when designing green spaces.

Comments 13: Limitations: Include all the limitations.

Response: Thank you for your valuable suggestions. We have addressed in the text that this study lacks an investigation into the time and frequency of people's contact with green spaces, which is crucial for understanding the impact of green space morphology on natural contact at the individual level. Additionally, this study focuses only on the visual perception preferences of individuals aged 18 to 40 and does not address the changes in perception in individuals over 40, which may influence the relationship between green space morphology and visual perception preferences. The study emphasizes green space coverage but does not consider green space accessibility, so it cannot accurately explain how green spaces contribute to social cohesion. Future research could further explore and address these aspects.

Comments 14: The conclusion section needs major rewriting. You should summarise the key findings with an emphasis on their international relevance.

Response: Thank you for your valuable feedback. We have addressed the comment by emphasizing the significance of green space morphology in influencing visual perception, particularly for larger, connected, aggregated, cohesive, and complexly shaped green spaces. Our findings suggest that urban green space investments that increase average area, connectivity, aggregation, cohesiveness, and shape complexity—while enhancing visual richness and diversity of landscape elements—can improve visual perception and produce better therapeutic effects. Additionally, we acknowledge that green space morphology may impact residents' accessibility to, frequency of, and duration of green space exposure, thereby affecting visual perception preferences and therapeutic outcomes. These insights have significant implications for urban green space planning, offering valuable evidence for city and landscape planners to consider in creating alternative green space designs.

Comments 15: Finally, you should provide specific recommendations for urban planners and policymakers, considering different international contexts.

Response: Thank you for your valuable comment. We have already included specific recommendations for urban planners and policymakers in the text, emphasizing the importance of considering green space morphology when planning urban areas. These recommendations suggest that investments should focus on increasing connectivity, aggregation, cohesiveness, and shape complexity of green spaces, as well as enhancing visual richness and diversity of landscape elements. Additionally, we highlight the need for a greater focus on green space accessibility and how it influences residents' engagement and well-being. These recommendations are presented with a consideration of international contexts, acknowledging the diverse needs and characteristics of different urban environments.

Comments 16: Land has an international readership so you should provide a balanced mix of local and international references. Please include more studies from diverse regions to improve the paper's global relevance. 

Response: Thank you for your suggestion. We have already included studies from diverse regions in the text to ensure the global relevance of our paper. These references reflect the international scope of research on urban green spaces and their impact on visual perception and well-being. We believe that the inclusion of these studies strengthens the paper's applicability to various geographical contexts, providing a comprehensive understanding of the topic.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

(1)The actual study is somewhat limited as it focuses solely on parks, which does not fully align with the broader topic of "Urban Green Space." It is recommended to modify the title to better reflect the actual research scope.

 

(2)There are too many paragraphs in the introduction section, and some background discussions, such as "COVID-19," are overly lengthy, which weakens the coherence and logical flow, failing to effectively highlight the innovation and significance of the study. Given that the research subject is the relationship between Urban Green Space Morphology and Visual Perception, is it reasonable to include "COVID-19" as a keyword?

 

(3)The author lacks supplementary descriptions for some experimental procedures, such as details about the volunteers participating in the study, the specific names of the parks selected, the detailed process of AHP weight allocation, and the criteria for selecting Google Street View photos.

 

(4)The author does not provide a detailed explanation for the basis of the indicators selected in the study, such as the categories of visual preference and urban green space morphology, making it difficult to justify that their choices are reasonable and comprehensive.

 

(5)The presentation of the SEM model research results is overly simplistic, lacking sufficient data interpretation. Additionally, in the discussion section that follows, the research data is not effectively utilized, resulting in a weak argument that lacks persuasiveness.

 

Comments on the Quality of English Language

In terms of writing ability, there are no obvious typos or grammatical errors.

Author Response

We sincerely appreciate reviewer 2 for their valuable feedback that we have used to improve the quality of our manuscript. We value the expert advice you provided and have made revisions throughout the text accordingly. We appreciate your enthusiastic efforts and hope that the revised manuscript meets your expectations.

 

Comments 1: The actual study is somewhat limited as it focuses solely on parks, which does not fully align with the broader topic of "Urban Green Space." It is recommended to modify the title to better reflect the actual research scope.

Response: Thank you for the reviewer’s suggestion. To better reflect the specific focus of the research, we have revised the title to "Decoding the Role of Urban Green Space Morphology in Shaping Visual Perception: A Park-Based Study," clarifying that the core focus of the study is the morphology of parks and its impact on visual perception.

Comments 2: There are too many paragraphs in the introduction section, and some background discussions, such as "COVID-19," are overly lengthy, which weakens the coherence and logical flow, failing to effectively highlight the innovation and significance of the study. Given that the research subject is the relationship between Urban Green Space Morphology and Visual Perception, is it reasonable to include "COVID-19" as a keyword?

Response: Thank you for the reviewer’s suggestion. We have streamlined the epidemiological background and retained only the content directly related to urban green space morphology and visual perception, in order to enhance the coherence and logical flow of the introduction.

 

Regarding the rationale for including "COVID-19" as a keyword, our initial consideration was that the event highlighted the need for healthy environments and increased attention to the role of urban green spaces in mental health. However, since the core of this study focuses on the relationship between green space morphology and visual perception, we agree that removing the "COVID-19" keyword would help make the research topic more focused. Thank you again for your valuable suggestion.

Comments 3: The author lacks supplementary descriptions for some experimental procedures, such as details about the volunteers participating in the study, the specific names of the parks selected, the detailed process of AHP weight allocation, and the criteria for selecting Google Street View photos.

Response: Thank you for the detailed review. We recognize that some aspects of the experimental procedures still need further elaboration to enhance the transparency and reproducibility of the study. In section 2.1, we have added details on the social demographic characteristics of the volunteers and the criteria for selecting parks, ensuring the rationality and applicability of the sample selection. However, due to potential impacts on park management, usage, and the fairness of subsequent studies, we are unable to disclose the specific names of the parks. We appreciate your understanding.

Additionally, the park photographs used in this study were taken by the research team on-site, rather than sourced from public databases such as Google Street View. The research photographs have not been publicly released and are solely intended for academic purposes, with no commercial use or further dissemination planned.

Comments 4: The author does not provide a detailed explanation for the basis of the indicators selected in the study, such as the categories of visual preference and urban green space morphology, making it difficult to justify that their choices are reasonable and comprehensive.

Response: Thank you for your valuable feedback. Regarding the basis for the selection of visual preference categories and urban green space morphology indicators, the indicators we selected are directly aligned with the research objectives, which aim to explore the relationship between urban green space morphology and visual perception. By focusing on these specific aspects, we ensure that the study addresses the impact of urban green spaces on users' perception and experience. Furthermore, these indicators are consistent with findings from previous studies, and their rationale is supported by existing literature (as mentioned in the introduction). Therefore, we believe these indicators are well-grounded theoretically.

Comments 5: The presentation of the SEM model research results is overly simplistic, lacking sufficient data interpretation. Additionally, in the discussion section that follows, the research data is not effectively utilized, resulting in a weak argument that lacks persuasiveness.

Response: Thank you for your suggestion. We have provided a more detailed presentation of the SEM model analysis results in the results section. Additionally, the relevant data and statistical test results have been included to further enhance the transparency and interpretability of the model outcomes. In the discussion section, we have conducted a deeper analysis of the SEM model results and compared the data with existing studies. We have also explored the implications of these results for urban green space planning and design, considering the research background and practical significance.

Furthermore, we have strengthened the theoretical and empirical support for the results through references to prior research and conclusions, which has enhanced the depth and persuasiveness of our arguments. We hope that these revisions help to integrate the discussion more closely with the research data and increase the credibility of the conclusions.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript, titled "Decoding the Role of Urban Green Space Morphology in Shaping Visual Perception" investigates how the morphology of urban green spaces (e.g., connectivity, aggregation, shape, etc.) affects visual perception preferences, using high-resolution satellite imagery, deep learning models, and modeling. Findings reveal that spatial characteristics, rather than just the quantity of green space, significantly enhance visual comfort, pleasure, and naturalness, offering practical insights for health-oriented urban design and planning. The article is well-structured and well-written. 

 

Main positive aspects:

- The study uniquely focuses on how the morphology of urban green spaces affects visual perception, moving beyond traditional metrics like green space quantity.

- The combination of high-resolution remote sensing data with deep learning models adds rigor, innovation, and depth to the analysis.

- The study integrates several data such as satellite imagery and eye-level photographs from volunteers, ensuring a holistic view of green space morphology and its visual impact.

- The findings offer practical insights for urban designers.

- By linking green space morphology to potential therapeutic benefits for individuals affected by long-term COVID-19 symptoms, the study addresses a timely and socially relevant issue.

 

 

Main negative aspects:

- The study is conducted exclusively in Chengdu, China, which may limit the applicability of findings to other geographic or cultural contexts.

- It is not clear how the visual perception of individuals was assessed. Later on the limitations, the reader gets more info, but still insufficient. The focus on individuals aged 18 to 40 excludes older populations, who might have different visual perception preferences and health outcomes related to green space exposure.

- The study does not account for how often people interact with green spaces, which could influence the relationship between green space morphology and visual perception.

- The methods used, while robust and innovative, may be difficult for non-specialist readers to interpret, potentially limiting the accessibility of the findings. Including further explanations could improve readability.

- The study implies a universal preference for certain green space morphologies without accounting for cultural or individual differences in aesthetic tastes.

 

 

 

Specific comments:

 

Line 102: Report? Maybe change to paper, article, or work…

 

Line 118: Only parks? Was this a criterion at first? There are a lot of different types of green spaces, maybe specify if you were only looking for parks in the first place, and also explain why. This is the type of green space that is designed to people, so it makes sense. But other types can be highly used (e.g., green squares, tree-lined streets, etc.). Were these parks public? Expand on these aspects to provide more context.

 

Line 138: How many photographs? And how many by park? Were there any rules on how the photographs should be taken? Information extracted from a good and bad photo can have very different outputs.

 

Line 152: Explain further what is the ADE20K dataset.

 

Lines 182-183: Can the authors explain a little bit more why those metrics were selected? Why those and not others? What were you looking for exactly? I understand that these are highly used metrics, but a small rationale could improve the text a lot.

 

Lines 200-201: “Landscape preferences are associated with factors such as the observer’s education level, gender, age, and expertise.” This sentence seems a repetition.

 

Line 202: “Studies have shown that perceptions of green spaces differ by gender…” Can you support this statement with more references?

 

Section 2.4: It is not clear for me how you evaluated landscape preference. Did you use data from questionnaires on people? How did gender, education, etc. entered the model? Was it based on real people data? Or was only based on the images (aerial and eye-level)? Were the volunteers the subjects (and therefore their preferences assessed through their photos)? This is important to understand, and from the data results I could not accurately find out. Please explain further.

 

 

Figure 2: This figure needs further discussion. This seems a very interesting result but I could not understand it entirely. The y axis represents proportion of what (provide a axis legend)? What are the orange dots representing (the people)? Please detail this further.

Author Response

We sincerely appreciate reviewer 3 for their valuable feedback that we have used to improve the quality of our manuscript. We value the expert advice you provided and have made revisions throughout the text accordingly. We appreciate your enthusiastic efforts and hope that the revised manuscript meets your expectations.

Comments 1: The study is conducted exclusively in Chengdu, China, which may limit the applicability of findings to other geographic or cultural contexts.

Response: Thank you for your valuable feedback. This study indeed focuses primarily on Chengdu, a representative city in southwest China, which has unique climatic, cultural, and urban planning characteristics. These factors may limit the generalizability of the findings to other regions. The results from Chengdu may not be fully applicable to areas with different cultural contexts, such as northern cities. We acknowledge this limitation and suggest that future research could conduct comparative studies in regions with distinct cultural differences to explore how culture influences green space morphology and visual preferences.

Although this study is centered on Chengdu, we believe the findings can still provide valuable insights for other rapidly urbanizing areas, particularly cities in central and western China. To enhance the broader applicability of the research, we suggest future studies expand their sample to include more cities (such as coastal cities or cities of different types). Moreover, cross-cultural comparative studies would help validate the generalizability of our findings and provide more targeted policy recommendations for urban green space planning and design in different regions.

Comments 2: It is not clear how the visual perception of individuals was assessed. Later on the limitations, the reader gets more info, but still insufficient. The focus on individuals aged 18 to 40 excludes older populations, who might have different visual perception preferences and health outcomes related to green space exposure.

Response: Thank you for your valuable feedback. In this study, visual perception was assessed through both a questionnaire survey and on-site photography. We employed standardized visual perception assessment tools that have been widely used in urban green space research. These tools aim to quantify participants' perceptual preferences for different green space characteristics, such as morphology, color, and complexity. Participants were asked to evaluate a series of urban green space images based on criteria such as visual attractiveness, comfort, and openness. To ensure the validity of the results, we performed a statistical power analysis using G Power.

The study focused on adults aged 18 to 40, exploring how urban green space morphology affects the visual perception of everyday users (such as university students and young professionals). However, we recognize that this age group selection may overlook the visual perception and health benefits for other age groups, particularly older adults. Health conditions in older populations may differ in relation to their interaction with and perception of green spaces. Future research could expand the sample to include older adults, particularly those aged 60 and above, to assess their visual perception and health outcomes. The specific characteristics of the elderly, in terms of perception, cognition, and physical health, may lead to different visual preferences and health responses to green spaces. We plan to explore the differences in visual perception and health effects across different age groups in future studies.

Comments 3: The study does not account for how often people interact with green spaces, which could influence the relationship between green space morphology and visual perception.

Response: Thank you for your valuable feedback. We recognize that the frequency of interaction with green spaces may influence individuals' visual perception of these spaces. People who frequently interact with green spaces may develop different perceptions and preferences for green space morphology compared to those with less frequent interaction, particularly over prolonged periods of use, during which individuals may develop a more profound perception or preference for certain features of the green space. 

However, this study primarily focuses on the impact of different green space morphologies on visual perception, and the frequency of participants' interactions with green spaces was not a primary variable in our research. We chose to concentrate on the visual morphological features of green spaces to explore their direct impact on perception. To address this limitation, future research could incorporate data on participants' frequency of interaction with green spaces in order to examine how this factor might influence their perceptions and preferences.

Comments 4: The methods used, while robust and innovative, may be difficult for non-specialist readers to interpret, potentially limiting the accessibility of the findings. Including further explanations could improve readability.

Response: Thank you for your valuable suggestion. To improve the clarity of the methods section, we have added more background information in the manuscript.

Comments 5: The study implies a universal preference for certain green space morphologies without accounting for cultural or individual differences in aesthetic tastes.

Response: Thank you for your valuable comment. The study may imply a universal preference for certain green space morphologies (such as open spaces, complexity, etc.), which does not fully account for the impact of factors such as cultural background, personal experiences, or social environment on visual perception preferences. In reality, there are indeed differences in aesthetic preferences across cultures and individuals, which could influence people’s preferences and evaluations of green space morphologies. However, in the study, we have discussed how different cultural backgrounds, age groups, and genders may affect visual perception and green space preferences.

Comments 6:  Line 102: Report? Maybe change to paper, article, or work…

Response: Thank you for your careful reading. We have revised "report" to "paper" to ensure the phrasing is more accurate and in line with academic writing conventions.

Comments 7: Only parks? Was this a criterion at first? There are a lot of different types of green spaces, maybe specify if you were only looking for parks in the first place, and also explain why. This is the type of green space that is designed to people, so it makes sense. But other types can be highly used (e.g., green squares, tree-lined streets, etc.). Were these parks public? Expand on these aspects to provide more context.

Response: Thank you for your valuable feedback. The primary reason for initially selecting parks as the focus of this study is that parks are typically designed as public green spaces, where the layout and design often consider multiple factors such as public health, visual perception, and social interaction. As a result, the morphology and layout of parks tend to have a more significant impact on visual perception. Compared to other types of green spaces (such as green belts, tree-lined streets, etc.), parks are one of the most commonly used and encountered spaces in urban areas, making them ideal for studying visual characteristics.

Indeed, apart from parks, many other types of urban green spaces (such as green belts, street greenery, and tree-lined avenues) also affect people's visual perception and well-being. Future research could explore these types of green spaces and examine how their visual effects and people's perception preferences differ. However, in order to maintain the focus of our study, we have not included these other types of green spaces in the current work.

All the parks selected for this study are public parks, meaning they are open to all citizens, not limited to specific groups. Public parks, as common green spaces, usually have a higher frequency of use and are more representative, better reflecting the visual preferences and perceptions of the general population. We will clarify this point in the methods section and provide additional details about the criteria used for park selection to give readers more comprehensive context and background.

Comments 8: Line 138: How many photographs? And how many by park? Were there any rules on how the photographs should be taken? Information extracted from a good and bad photo can have very different outputs. 

Response: Thank you for raising the question. A total of 772 photos were taken from 10 different parks. To avoid potential confounding factors, we selected green spaces of similar size and infrastructure as the units of analysis. After evaluating the fractal dimension and visual complexity of the photos, 10 images were selected from each park, covering different landscape features and elements. This ensures a comprehensive reflection of how green space morphology influences visual perception. To maintain consistency in the photos, the standards for photo-taking and selection have been added to the text. 

Thank you for your valuable comment regarding the high foot traffic during the park photography sessions. While we acknowledge that the presence of more visitors may have impacted the visual aesthetics of the images and the experience of participants, we believe that this factor does not significantly affect the primary objectives of the study, which focused on landscape morphology and visual perception. Moreover, we carefully selected images where the crowd presence did not dominate the scene, ensuring the core features of the park's landscape were the focus.

Comments 9: Line 152: Explain further what is the ADE20K dataset.

Response: Updated.

Comments 10: Lines 182-183: Can the authors explain a little bit more why those metrics were selected? Why those and not others? What were you looking for exactly? I understand that these are highly used metrics, but a small rationale could improve the text a lot.

Response: Thank you for your valuable suggestion. The indicators selected for this study are based on the need to capture the core characteristics of green space morphology that influence visual perception and preferences. In addition to the empirical support for these indicators in the literature, which are closely related to the goals of this study—namely, to understand how green space morphology shapes visual perception and preferences—these indicators can be quantified through high-resolution remote sensing data and machine learning techniques, which facilitates large-scale data processing and comparison. This, in turn, enhances the operability and reliability of the research results. The indicators cover different dimensions (such as landscape complexity and diversity), providing a comprehensive reflection of the morphological features of green spaces. This approach allows us to better understand the multifaceted factors that influence visual perception, rather than focusing on a single visual characteristic.

Comments 11: Lines 200-201: “Landscape preferences are associated with factors such as the observer’s education level, gender, age, and expertise.” This sentence seems a repetition.

Response: Thank you for your suggestion. The sentence has been revised to: "Landscape preferences are influenced by various factors, including the observer's education, gender, age, and expertise."

Comments 12: Line 202: “Studies have shown that perceptions of green spaces differ by gender…” Can you support this statement with more references?

Response: Updated.

Comments 13: Section 2.4: It is not clear for me how you evaluated landscape preference. Did you use data from questionnaires on people? How did gender, education, etc. entered the model? Was it based on real people data? Or was only based on the images (aerial and eye-level)? Were the volunteers the subjects (and therefore their preferences assessed through their photos)? This is important to understand, and from the data results I could not accurately find out. Please explain further.

Response: Thank you for your valuable feedback. The issue you raised regarding how to assess landscape preferences is very important, and we will further clarify the evaluation process and methodology in the paper. We did indeed use a questionnaire survey to assess participants' landscape preferences. Specifically, participants rated images that we provided. Our study is based on real participant data. The assessment of landscape preferences does not solely rely on the images themselves, but on the evaluations made by the volunteers for these images. Therefore, the results of this study reflect the preferences of real people, rather than being based purely on the visual characteristics of the images.

Additionally, the volunteers are the subjects of this study. We collected data through the volunteers' landscape preference ratings to explore how different types of urban green space morphology influence the visual perception of different groups.

The personal background of the participants, such as gender, age, and educational background, is indeed an important factor in our model. All volunteers (a total of 150) filled out a background information form before participating in the questionnaire survey, which included details such as gender, age, and education background. This information was used as control variables. During the analysis, we considered the potential impact of these personal characteristics to ensure that the evaluation of landscape preferences reflected the opinions of a broad group, rather than being influenced by individual traits. Specifically, we included variables like the number of participants, education background, and gender as covariates in the statistical model and controlled for their effects during the analysis. This means that when evaluating landscape preferences, we accounted for how the participants' personal backgrounds might influence their ratings.

Comments 14: Figure 2: This figure needs further discussion. This seems a very interesting result but I could not understand it entirely. The y axis represents proportion of what (provide a axis legend)? What are the orange dots representing (the people)? Please detail this further.

Response: Thank you for your valuable suggestion. The figure has been updated in the manuscript. This violin plot illustrates the central tendency and dispersion of the data, combining features of both box plots and density plots. The vertical axis (Y-axis) represents landscape preference. The width of the violin indicates the data density. The wider the violin, the more concentrated the sample data in that region; conversely, the narrower the violin, the fewer data points in that area. The orange dots represent individual data points, which influence the width of each individual violin plot.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

In the discussion section, it is mentioned that there is a correlation between green space form and visual perception preference, and that the proportion of landscape elements plays an intermediary role, but the underlying mechanisms of these relationships are not sufficiently explored, and it is suggested that more in-depth discussions be conducted.

Similar studies in different countries or regions can be compared to analyze the differences in their applicability in different contexts and geographic environments, so as to make the planning recommendations more generalizable.

Author Response

Comments 1: 

In the discussion section, it is mentioned that there is a correlation between green space form and visual perception preference, and that the proportion of landscape elements plays an intermediary role, but the underlying mechanisms of these relationships are not sufficiently explored, and it is suggested that more in-depth discussions be conducted.

Similar studies in different countries or regions can be compared to analyze the differences in their applicability in different contexts and geographic environments, so as to make the planning recommendations more generalizable.

Response 1: Thank you for your valuable comments. We appreciate your suggestion to explore more deeply the underlying mechanisms behind the relationship between green space morphology and visual perception preferences, as well as the mediating role of landscape element proportions. In the discussion section, we have acknowledged these mechanisms and highlighted that "the association between green space morphology and visual perception preferences is mediated by the proportion of visual landscape elements." However, we agree that further clarifying the specific processes of interaction between these factors can strengthen the overall discussion. Future studies, including longitudinal research, could provide additional insights into how different spatial and morphological characteristics directly influence visual perception and well-being.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for your efforts in revising the manuscript and for addressing my  comments. However, after reviewing both the response to my comments and the revised manuscript, it is frustrating to see that the changes made in the response are not fully reflected in the text. While you acknowledge the feedback and state that revisions have been incorporated, the modifications in the manuscript appear to be minor and do not sufficiently address the key concerns raised.

In particular, the literature expansion, methodological justifications, comparative perspectives, conclusion  and the discussion on international relevance remain underdeveloped. Given these inconsistencies, I am afraid that, in its current form, the manuscript cannot be accepted. A more substantial revision is needed to ensure that the response to my comments  aligns with actual improvements in the manuscript.

Author Response

Comments 1: 

Thank you for your efforts in revising the manuscript and for addressing my  comments. However, after reviewing both the response to my comments and the revised manuscript, it is frustrating to see that the changes made in the response are not fully reflected in the text. While you acknowledge the feedback and state that revisions have been incorporated, the modifications in the manuscript appear to be minor and do not sufficiently address the key concerns raised.

In particular, the literature expansion, methodological justifications, comparative perspectives, conclusion  and the discussion on international relevance remain underdeveloped. Given these inconsistencies, I am afraid that, in its current form, the manuscript cannot be accepted. A more substantial revision is needed to ensure that the response to my comments  aligns with actual improvements in the manuscript.

Response 1: 

Thank you for your patience and detailed feedback on our revisions. We understand your concerns and sincerely apologize for any inconsistencies in the modifications. We fully recognize that although we mentioned the changes made in our responses, they may not have fully met your expectations in the manuscript itself.

After reviewing the comments from round 1, we have made the following updates:

1. Literature Expansion: We will further expand the literature review by incorporating more relevant studies from different regions and international research. This will enhance the global relevance and applicability of our study.

2. Methodological Justification: We will provide a more detailed explanation of why we chose our research methods, particularly the multi-objective regression model and structural equation modeling, to ensure that the methodology section is clearer and more well-justified.

3. Comparative Perspectives: We will include a comparative perspective from international studies in the discussion section, analyzing the differences in how green space morphology affects visual perception preferences in different regions and cultural contexts. This will improve the generalizability and depth of the study.

4. Conclusion: We will strengthen the conclusion by clearly summarizing the key findings and further exploring the implications and practical significance of these findings in a global context. We will emphasize the applicability of this study in different cultural and geographical environments and highlight the potential global impact of green space morphology on urban human well-being.

We hope these revisions will address your concerns and enhance the overall quality of the manuscript. Thank you again for your thoughtful and constructive feedback.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The author has solved most of the problems raised

Author Response

Comments 1: The author has solved most of the problems raised.

Response 1: Thank you for your valuable feedback on our research. We are pleased to hear that you believe we have addressed most of the issues raised.