Cycling Greenway Planning towards Sustainable Leisure and Recreation: Assessing Network Potential in the Built Environment of Chengdu

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors,

Thank you for your study. Please find my feedback as follows.

1.      The research gap and the need for this study are not clear. This study would benefit from a comprehensive literature review to reveal the current studies and their approaches to justify the need for the study.

2.      The last paragraph of section 1.4 is abrupt. If the purpose of the paragraph is to transition to section 1.5., please revisit your flow between sections 1.4 and 1.5.

3.      The use of regression analysis is not clear. Section 1.5 provides some examples from the literature; however, a separate methodology section would benefit the study for a better flow and understanding.

4.      After reading sections 2 and 3, I strongly suggest presenting the methodology before section 2.  Identified data sets and various analysis methods are hard to follow in section 2 before presenting methods.

5.      What is NDVI? Please provide the definition of the abbreviations in their first use in the paper.

6.      What is spatial syntax theory? why is it utilized? There are several numbers and analyses in section 2, why they are separated from the methodology and results?

7.      Why OLS and then GWR and then MGWR? It’s not clear these methods are utilized.

8.      “A review of existing research helped identify preliminary indicators from three key 306 aspects of the built environment—land use, transportation system, and Architectural Form—that could influence DBS leisure cycling. “  Which existing research? The section that presents these variables does not present a review of existing research.

Overall, it is really hard to follow the study in its current form. A clear structure of the paper with an introduction, research questions, literature review, methodology, results, discussion, and conclusion will benefit for the clarity. The research gap and the need for the study are not clear. Several methods and techniques are utilized, but the logic behind selecting these methods is not clear. The value of the work completed is not presented in an understandable flow in the paper, and I believe the necessary updates to improve the quality of work are beyond the scope of revisions.

 

 

Comments on the Quality of English Language

N/A

Author Response

Dear Reviewer,

 

Thank you for your insightful feedback on our manuscript. We appreciate the time and effort you have taken to provide detailed comments and suggestions. Below is our response to your concerns:

 

Comment #1: The research gap and the need for this study are not clear. This study would benefit from a comprehensive literature review to reveal the current studies and their approaches to justify the need for the study.

 

Response:

Thank you for pointing this out. We acknowledge that the research gap and the need for this study may not have been sufficiently clear in the original submission. To address this, we have revised section 1.6 to explicitly highlight the gaps identified in our comprehensive literature review (sections 1.1 to 1.5) and to articulate the unique contributions of our study. The revised section 1.6, now titled "My Contributions," outlines the following key points:

 

1. Comprehensive Approach: Unlike previous studies that focus on singular aspects of the built environment, our study employs MGWR to analyze multiple dimensions simultaneously. This approach provides a more holistic understanding of the factors influencing leisure cycling.

 

2. Spatial Heterogeneity: Our use of MGWR with a 95% confidence interval allows us to capture spatial variations in the significance of environmental factors. This nuanced analysis provides deeper insights into cycling behaviors and helps avoid the inclusion of non-significant influences.

 

3. Data-Driven Greenway Planning: Our method bridges cycling routes with greenways using empirical data from dockless bike-sharing systems. This approach, based on the linear nature of both cycling routes and greenways, ensures that greenway planning is grounded in actual cycling behavior, offering a scalable model for other cities.

 

These revisions clearly articulate the necessity and uniqueness of our study, addressing the research gaps identified in the existing literature. We believe these enhancements significantly strengthen our manuscript and align with your suggestions.

 

Comment #2: The last paragraph of section 1.4 is abrupt. If the purpose of the paragraph is to transition to section 1.5., please revisit your flow between sections 1.4 and 1.5.

 

Response:

Agree. To address your concern about the abrupt transition, we have revised the structure of sections 1.4 and 1.5 to ensure a smoother transition and clearer distinction between the two sections. In the revised manuscript:

 

1. Section 1.4 now focuses on the significance of bike lane and cycling greenway planning, emphasizing the integration of dockless bike-sharing (DBS) data and traditional planning methods. This section discusses how advances in information technology and modern data analysis techniques enhance non-motorized transport planning by connecting urban design with parks, amenities, and services.

 

2. Section 1.5 is dedicated to various regression models used in quantitative research for urban planning. It begins by highlighting the role of regression analysis in deriving factor weights for urban planning, then delves into different regression models, including Ordinary Least Squares (OLS), negative binomial, logistic regression, and more advanced techniques like Geographically Weighted Regression (GWR) and Multiscale Geographically Weighted Regression (MGWR). This section explains how these models address spatial heterogeneity and improve the interpretability of studies on non-motorized traffic patterns.

 

These revisions ensure a logical flow and coherence between the two sections, enhancing the reader's understanding of the methodological framework and its application in our study.

 

Comment #3: The use of regression analysis is not clear. Section 1.5 provides some examples from the literature; however, a separate methodology section would benefit the study for a better flow and understanding.

 

Response:

Thank you for your valuable feedback. To address your concern about the clarity of our use of regression analysis, we have significantly revised the structure of the manuscript to include a separate and comprehensive methodology section. This new section (Section 2) provides a detailed explanation of our research design, regression models, data preparation, and evaluation model. By clearly distinguishing the methodological approach, we aim to improve the flow and understanding of our study.

 

Additionally, we have expanded Section 1.5 to further elaborate on the examples from the literature and to provide a more in-depth discussion of various regression models.

 

Comment #4: After reading sections 2 and 3, I strongly suggest presenting the methodology before section 2.  Identified data sets and various analysis methods are hard to follow in section 2 before presenting methods.

 

Response:

Thank you for your insightful suggestion. We have undertaken a comprehensive restructuring of the manuscript to present the methodology before discussing the detailed data sets and various analysis methods. This reorganization aims to improve the logical flow and readability of the paper, making it easier for readers to follow our research approach.

 

The revised manuscript now features a comprehensive methodology section (Section 2), which precedes the data and analysis sections. The updated structure is as follows:

• Introduction
• Methodology
  • Research Design and Objectives
  • Regression Models
    • Ordinary Least Squares (OLS) Regression Model
    • Geographically Weighted Regression Model (GWR)
    • Multi-bandwidth Geographically Weighted Regression Model (MGWR)
  • Study Area
  • DBS Data
  • Built Environment Data
  • Data Preparation
  • Evaluation Model of Cycling Greenway Potential
• Results
• Discussion
• Conclusion

 

In Section 2 (Methodology), we provide a detailed explanation of our research design and objectives, the rationale for selecting specific regression models, a description of the study area, and the data collection and preprocessing steps. This section also includes the development of an evaluation model for assessing the potential of cycling greenways. By presenting the methodology first, we ensure that readers have a clear understanding of the research framework before delving into the specifics of the data and analysis methods.

 

We believe these changes significantly enhance the clarity and structure of the manuscript, aligning with your suggestions and improving the overall readability of the study.

 

Comment #5: What is NDVI? Please provide the definition of the abbreviations in their first use in the paper.

 

Response:

Thank you for pointing out the need for clarity regarding the abbreviation NDVI. We have revised the manuscript to include the full term "Normalized Difference Vegetation Index (NDVI)" upon its first use. By explaining that NDVI is derived from multispectral remote sensing imagery, we provide readers with a quick understanding of how this index is used to analyze the spatial patterns of vegetation and built environments within the study area.

 

Comment #6: What is spatial syntax theory? why is it utilized? There are several numbers and analyses in section 2, why they are separated from the methodology and results?

 

Response:

Thank you for your valuable feedback. In response to your comment regarding spatial syntax theory and its utilization, we have revised the manuscript to provide a clearer definition and explanation of this theory. Spatial syntax theory is a method for analyzing spatial configurations, commonly used to assess the accessibility and integration of urban street networks. In our study, we utilize space syntax metrics such as integration and choice as part of the transportation system indicators. Integration measures how well-connected a street segment is within the network, while choice measures the likelihood of a street segment being used as part of a route. These metrics are calculated beforehand and incorporated into the regression analysis to evaluate their influence on DBS usage.

 

Regarding the separation of methodology and results, we would like to clarify that the methodology section details the calculation of space syntax metrics and their role in the overall regression analysis. The results section presents the outcomes of the regression analysis, focusing on how these pre-calculated indicators, along with other built environment factors, influence DBS usage. This approach maintains a clear and logical structure, where the preparation of data, including space syntax calculations, is detailed in the methodology section, and the analytical results are presented in the results section.

 

Comment #7: Why OLS and then GWR and then MGWR? It’s not clear these methods are utilized.

 

Response:

Thank you for your valuable feedback. To address your concern regarding the utilization of OLS, GWR, and MGWR models, we have revised Section 2.2 to provide a more detailed explanation of the rationale behind using these models in a sequential manner. The hierarchical approach is guided by specific objectives and the inherent characteristics of urban data. OLS serves as the foundational regression model to establish baseline relationships between built environment factors and DBS usage. GWR is then employed to account for spatial non-stationarity, providing local variations in regression coefficients. Finally, MGWR is used to further refine the analysis by considering varying spatial scales for different explanatory variables. MGWR's ability to model spatially varying relationships at multiple scales offers a more detailed understanding of spatial dynamics, leading to more precise and reliable results.

 

By progressing from OLS to GWR and then to MGWR, we can incrementally address the complexities and spatial heterogeneity of urban data, ensuring that our analysis captures the nuanced relationships between built environment factors and DBS usage. This hierarchical approach not only enhances the robustness of our findings but also provides a comprehensive framework for urban cycling infrastructure planning.

 

Comment #8: “A review of existing research helped identify preliminary indicators from three key 306 aspects of the built environment—land use, transportation system, and Architectural Form—that could influence DBS leisure cycling.” Which existing research? The section that presents these variables does not present a review of existing research.

 

Response:

Thank you for your valuable suggestion. In response to your comment regarding the existing research that helped identify preliminary indicators from three key dimensions of the built environment—land use, transportation system, and architectural form—we have revised Section 1.2 (Association between DBS and Built Environment) to provide a more detailed literature review. This expanded review offers a comprehensive overview of existing studies on built environment elements and their methods, emphasizing their influence on DBS leisure cycling. Specifically, we have discussed how land use, transportation systems, urban design, and architectural form impact shared bicycle usage. This detailed review now supports the indicators mentioned in Section 2.6.2, ensuring that our selection of variables is well-grounded in existing research.

 

Thank you once again for your valuable feedback. 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Considering the value of mixed land use, accessibility, and human-scale design the research aimed to promote the cycling infrastructure investment -as a sustainable urban mobility mode- through proposing a master plan for its context of the study.  

The whole process including literature review , research design, method, and conclusion seems well. However, I would suggest to address the following points as well: 

+ Why are the older adults and children excluded from your study?

+ Does this study consider shared bikes stations through the main routes?

+ Have you considered about the location of the supporting facilities to biking (repairing stations, etc.) along the principal routes?

+ Is this study applicable to other similar contexts throughout the country?

+ Does the climate and geographical condition of the context support promoting this mode of transport?

+ Are special events (races, etc.) considered in the proposed master plan?

 

Thank you. 

 

Author Response

Dear Reviewer,

 

Thank you for your insightful feedback on our manuscript. We appreciate the time and effort you have taken to provide detailed comments and suggestions. Below is our response to your concerns:

 

Comment #1: Why are the older adults and children excluded from your study?

 

Response:

Thank you for your valuable feedback. The exclusion of older adults and children from our study was primarily due to the characteristics of the dockless bike-sharing (DBS) data, which predominantly represent the usage patterns of adults. Dockless bike-sharing systems often restrict registration to those over 18 years old, and the use of smartphone apps for accessing these services may not be user-friendly for older adults. Additionally, the physical design of the bikes, such as saddle height adjustments, may not accommodate all users.

 

In the limitations section of our manuscript, we have expanded on this point to acknowledge the potential impact of this exclusion on the generalizability of our findings. We have also outlined future research directions that could address these gaps, including targeted studies on green mobility services for children and the elderly using methods like field surveys or panoramic photo records.

 

Comment #2: Does this study consider shared bikes stations through the main routes?

 

Response:

Thank you for your valuable question. This study focuses on dockless bike-sharing (DBS) systems, which have become the predominant mode of bike-sharing in China. DBS platforms leverage modern technology, such as cashless payments and GPS tracking, to enable users to locate, unlock, and pay for bikes using their smartphones. Shared bikes that are not bound by stations can be more reflective of riders' preferences for starting and ending points and routes. This system provides real-time data feedback, allowing for an in-depth analysis of cycling behaviors on an unprecedented scale.

 

A survey conducted in 2017 indicated that over 70% of Chinese users preferred dockless systems over traditional station-based ones (China Bike Sharing 13 Cities User Research Report in Chinese-Future Think Tank -2017.pdf (ambchina.com)). As a result, the shared bike industry in China has evolved through market competition these years to predominantly feature dockless systems, effectively eliminating the need for fixed bike stations. Our study utilizes data from these dockless systems to analyze cycling behavior and infrastructure along main routes.

 

To ensure a clearer understanding of DBS systems for all readers, we have expanded the explanation in Section 1.1 to include more details about the evolution and characteristics of dockless bike-sharing in China.

 

Comment #3: Have you considered about the location of the supporting facilities to biking (repairing stations, etc.) along the principal routes?

 

Response:

Thank you for your valuable question. Our study primarily focuses on evaluating and identifying potential cycling greenways based on the existing urban fabric. However, we have considered the importance of supporting facilities such as repairing stations along the principal routes. In Section 4.4 (Implications), we have discussed how municipalities can use the evaluation model to strategically expand or retrofit cycling greenways, optimizing resource allocation. We have now added a specific mention of supporting facilities like repairing stations to emphasize their importance in creating a comprehensive cycling network.

 

We believe these additions address your concerns and enhance the practical applicability of our study. Thank you once again for your insightful suggestion.

 

Comment #4: Is this study applicable to other similar contexts throughout the country?

 

Response:

Thank you for your insightful question. This study is designed to be applicable to other urban contexts with similar characteristics. The methodology and findings provide a scalable model that other cities with dockless bike-sharing (DBS) services can adopt. By leveraging local DBS data and built environment characteristics, cities can adapt the evaluation model to suit their unique contexts, enhancing urban mobility in a way that is responsive to local needs and preferences.

 

In Section 1.6 (Our Contributions), we discuss how this method aligns with public preferences and promotes widespread applicability and transferability of the findings. Additionally, in Section 4.4 (Implications), we have highlighted the potential for other municipalities to use this model to strategically expand or retrofit cycling greenways. This adaptability ensures that the evaluation model remains relevant and effective across different urban settings, making it a valuable tool for promoting sustainable urban mobility and infrastructure development.

 

Furthermore, as of the end of August 2019, dockless bike-sharing services in China have covered more than 360 cities nationwide, including every corner of the built-up areas in large cities (Blue Book of Sharing Economy: Annual Report on the Development of Shared Mobility in China, 2019). This extensive coverage means that other cities can readily use DBS data as a key reference for planning and development.

 

Comment #5: Does the climate and geographical condition of the context support promoting this mode of transport?

 

Response:

Thank you for your valuable question. In our study, we have considered the climate and geographical conditions of Chengdu. As mentioned in Section 2.3 (Study Area), Chengdu is primarily a plain area, which is conducive to promoting cycling. Additionally, in Section 2.4.1, we noted that the weather records for the data collection dates indicate favorable conditions, ensuring that the data reflects typical local cycling behavior.

 

While our study focuses on Chengdu, future research should indeed consider the climate and geographical conditions of other cities when applying the evaluation model. Different climatic conditions and geographical features could significantly impact cycling behavior and infrastructure needs. Therefore, incorporating local climate and geography into the planning and analysis process is crucial for accurately promoting this mode of transport in different urban contexts.

 

Comment #6: Are special events (races, etc.) considered in the proposed master plan?

 

Response:

Thank you for your question. The proposed master plan in our study is part of the National Key Research and Development Program of China's 14th Five-Year Plan: Research on technology and method of urban renewal comprehensive evaluation based on urban ‘physical examination’ evaluation. This project is designed to provide policy recommendations to the Chengdu municipal government.

 

 

Thank you once again for your valuable feedback.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This thesis provides an in-depth study on urban greenway planning and the enhancement of urban leisure activities, especially in the context of Chengdu, China, which is highly original and of practical application. The study provides a comprehensive analysis of the potential of bicycle greenways in the urban built environment through the use of DBS big data, which is important for promoting urban green mobility and enhancing urban sustainability.Through detailed analysis and simulation of DBS data, the authors demonstrated the bicycle flows on the existing network in Chengdu, thus assessing the potential for future bicycle greenway infrastructure improvements.

However, the paper requires extensive revision before publication.

1. Is the DBS data used in the paper fully representative of the behavior of all cyclists? It is recommended that the authors discuss the representativeness of the data and its potential impact on the results.

2. There are a large number of built environment factors related to cycling, and the authors should provide more detail in the literature review on the arguments of existing studies on built environment elements and related methods, and highlight the innovative nature of this paper. The literature review section needs to be reorganized.

3. For urban design in cycling, image segmentation is currently based on street view, but street view is mostly based on the car perspective, which is somewhat different from the cycling perspective. More information is needed from the authors on how to account for the difference between the two.

4. According to the authors, the routes are simulation-generated rather than actual routes, and the authors do not go into detail on how to filter out non-commuter cycling behavior. The authors say that the distinction is made by time of day, but the commuter group cannot be excluded during this time period, is further screening needed in combination with the attributes of OD?

5. The results of the MGWR analysis are not presented in detail in the Results section. The results presented in the discussion section are still questionable, the authors generated that MUR has the strongest effect on cycling, but in fact the mean and the maximum and minimum values show a low correlation, especially when compared to CND and SVF, the authors need to explain the results further.

6. The discussion section needs to be analyzed based on the data of the existing results. No new findings are presented from the existing results. A more specific presentation of the new findings of this paper based on the independent variables is needed.

Author Response

Dear Reviewer,

 

Thank you for your insightful feedback on our manuscript. We appreciate the time and effort you have taken to provide detailed comments and suggestions. Below is our response to your concerns:

 

Comment #1: Is the DBS data used in the paper fully representative of the behavior of all cyclists? It is recommended that the authors discuss the representativeness of the data and its potential impact on the results.

 

Response:

Thank you for your valuable question. The DBS data used in this study primarily represents the usage patterns of adults, excluding certain age groups like children and the elderly due to registration restrictions, user-friendliness of smartphone apps, and physical design constraints of the bikes. This exclusion limits the generalizability of the findings and might lead to planning and design biases in public spaces.

 

In Section 4.5 (Limitations and Further Study), we have expanded on this point to acknowledge the potential impact of this exclusion on the generalizability of our findings. While DBS data provides extensive insights into the behavior of the majority of cyclists, it may not fully capture the cycling patterns of all demographic groups. Future research should include targeted studies on green mobility services for children and the elderly, using methods like field surveys or image semantic segmentation based on panoramic photo records to gather relevant data and understand their cycling preferences.

 

Comment #2: There are a large number of built environment factors related to cycling, and the authors should provide more detail in the literature review on the arguments of existing studies on built environment elements and related methods, and highlight the innovative nature of this paper. The literature review section needs to be reorganized.

 

Response:

Thank you for your valuable suggestion. We have revised Section 1.2 to provide more detail in the literature review on the arguments of existing studies on built environment elements and related methods. Additionally, in Section 1.6 (Our Contributions), we have detailed the innovative aspects of our study, such as the use of MGWR to analyze multiple dimensions of built environment factors simultaneously. This approach allows coefficients to vary spatially and enables scale proportions to adapt across different explanatory variables while minimizing the impact of multicollinearity. These enhancements provide a robust basis for the selection of built environment factors for this study and highlight the innovative nature of our paper by situating our work within the broader context of built environment and cycling research.

 

Specifically, we have discussed how land use, transportation systems, urban design, and architectural form influence shared bicycle usage. We have also highlighted the transition from subjective assessments to data-driven evaluations using urban big data, emphasizing the varying impacts of built environment factors across different urban contexts and cultures.

 

Comment #3: For urban design in cycling, image segmentation is currently based on street view, but street view is mostly based on the car perspective, which is somewhat different from the cycling perspective. More information is needed from the authors on how to account for the difference between the two.

 

Response:

Thank you for your valuable question. We acknowledge that street view imagery used for urban design analysis is typically captured from the perspective of cars, which differs from the cycling perspective. However, this imagery provides valuable insights into the built environment that are relevant for non-motorized transport studies. Street view imagery transforms traditional top-down urban studies by allowing comprehensive quantitative research on street quality at an urban scale.

 

Although the car perspective differs from the cycling perspective, street view imagery primarily captures elements such as sky, trees, and built structures, which contribute to the overall environment rather than focusing solely on visual or psychological experiences of cyclists. To address the limitations of this perspective in current cycling studies, we propose including additional cycling-related covariates to enhance the specificity and relevance of the research.

 

We have added a discussion on this topic in Section 2.5.3 (Urban Design) to explain how we account for the differences between the car perspective and the cycling perspective, and how our methodology mitigates these differences.

 

Comment #4: According to the authors, the routes are simulation-generated rather than actual routes, and the authors do not go into detail on how to filter out non-commuter cycling behavior. The authors say that the distinction is made by time of day, but the commuter group cannot be excluded during this time period, is further screening needed in combination with the attributes of OD?

 

Response:

Thank you for your valuable question. To address your concern regarding the distinction between leisure and commuter cycling behavior, we have provided additional details in Section 2.4.1 (DBS Data Collection).

 

Our data collection was conducted using the API of the WeChat 'Mobike' application, capturing over a million real-time parking points between 13:00 and 17:00 on October 20 and 21, 2018, which were a Saturday and Sunday. Previous research (Wilkesmann et al., 2023) indicates that weekend afternoons are peak periods for leisure cycling. Additionally, we performed a preliminary data exploration by conducting a simple linear regression analysis between the collected DBS data and commonly recognized work-related Points of Interest (POI), such as company and enterprise locations. The p-value from this regression was 0.823014, indicating no significant correlation between the two, thus confirming that the data primarily represents leisure cycling rather than commuter cycling.

 

Comment #5: The results of the MGWR analysis are not presented in detail in the Results section. The results presented in the discussion section are still questionable, the authors generated that MUR has the strongest effect on cycling, but in fact the mean and the maximum and minimum values show a low correlation, especially when compared to CND and SVF, the authors need to explain the results further.

 

Response:

Thank you for your insightful comments and suggestions. We have revised the manuscript to address the concerns about the presentation and interpretation of the MGWR results.

 

Presentation of MGWR Results:

We have significantly expanded Section 3.2 to provide a more detailed presentation of the MGWR results. This section now includes a thorough analysis of the MGWR coefficients and their spatial variability. Specifically, Table 9 presents the MGWR coefficient results and the weights for the evaluation model, highlighting the influence of each built environment factor on leisure cycling. The expanded discussion emphasizes the substantial influence of several indicators based on Table 9 and Figure 7.

 

Explanation of Results:

In the revised Section 4.2, we provided a more nuanced discussion of the MGWR results:

 

• Land Use: MUR is identified as having the strongest influence within the land use dimension. While previous research has highlighted the impact of mixed-use areas on cycling, our study finds that the influence of MUR is significantly greater than any single land use POI density, which differs from prior findings. This suggests a strong preference for cycling in mixed-use areas over areas dominated by a single type of land use.
• Transportation System: The results show considerable spatial heterogeneity in SSI, with both positive and negative impacts. High integration in central urban areas may deter cycling due to congestion, while increased integration in peripheral areas promotes cycling. In contrast, SSC consistently showed a positive influence across all regions, indicating a universal preference for routes that offer freedom of movement and diverse path choices.
• Urban Design: SVF exhibited considerable spatial variability, with both positive and negative effects. The impact of SVF varies widely, reflecting different regional preferences for sky visibility and openness in urban design. For example, in some central urban areas, a higher SVF is associated with increased cycling, while in peripheral areas, complete sky obstruction is not preferred by cyclists.
• Architectural Form: Differences between old and new urban areas are notable. In the southern new city, FAR shows insignificant effects, similar to SGF and PSP, indicating that these factors are less influential in newly developed areas. Conversely, AGF in the old city displays widespread insignificance, reflecting the impact of distinct planning and spatial layouts from different eras on leisure cycling.

 

Correlation vs. Influence:

The MGWR model coefficients indicate the strength and direction of the relationship between each built environment factor and leisure cycling. It's essential to clarify that these coefficients represent the "influence" rather than the "correlation" in a simple linear sense, such as Pearson correlation coefficients. The MGWR approach allows for spatially varying relationships, meaning the influence of a particular factor can differ across different areas.

 

Key Insights and Visualization:

We have included visualizations (Figure 7) that map the regression results across the study grid. This helps articulate the varying impacts of urban design elements on cycling preferences across Chengdu, providing an intuitive understanding of environmental influences on cycling activity. Additionally, we have further refined and detailed the key findings in Section 4.2 to highlight the novel insights gained from our study. These findings underscore the unique contributions of our research, particularly the significant regional differences and threshold effects identified through the MGWR analysist.

 

Threshold Effect:

We have added a discussion on the threshold effect observed in the MGWR results. Once certain areas achieve adequate accessibility and road density, the influence of SSI and CND stabilizes, as evidenced by the large areas of insignificance in the 95% confidence interval comparison (Figure 7). This suggests that further improvements in these metrics do not significantly enhance cycling potential, guiding future greenway planning and urban development strategies to focus resources on other factors.

 

We believe these revisions address your concerns and provide a clearer, more detailed explanation of the MGWR analysis and its findings. Thank you again for your valuable feedback.

 

Comment #6: The discussion section needs to be analyzed based on the data of the existing results. No new findings are presented from the existing results. A more specific presentation of the new findings of this paper based on the independent variables is needed.

 

Response:

Thank you for your detailed feedback and suggestions. We have carefully revised the discussion section to ensure it is thoroughly analyzed based on the data of the existing results and presents new findings from our study.

 

In the revised Section 4.2, we have highlighted the key findings derived from our analysis, emphasizing the novel insights and contributions of this research. These key findings are directly based on the independent variables and the MGWR results, and they provide specific, data-driven insights into the impact of various built environment factors on leisure cycling. Here are the key points addressed:

 

Land Use Dimension:

MUR's Influence: MUR exerts the strongest influence on leisure cycling within the land use dimension, significantly greater than single land use POI densities. This indicates the importance of mixed-use areas in promoting cycling activities, a finding that diverges from previous studies that did not highlight this strong differential impact.

 

Transportation System:

Spatial Syntax Integration (SSI): SSI shows considerable spatial heterogeneity, with both positive and negative impacts. High integration in central urban areas may deter cycling due to congestion, while increased integration in peripheral areas promotes cycling. This nuanced understanding helps tailor urban design strategies more effectively.

Public Transit POIs: Public transit POIs, particularly subway stations, positively influence leisure cycling more than any single type of leisure and entertainment venues. This significant influence suggests that well-integrated transit options can greatly enhance the attractiveness of cycling routes, a finding that previous studies have not explicitly confirmed.

 

Urban Design:

Street Sky-view Factor (SVF): SVF's impact exhibits considerable spatial variability, with both positive and negative effects. This indicates different regional preferences for sky visibility, light conditions, and open views. For instance, higher SVF in central urban areas is associated with increased cycling, while complete sky obstruction in peripheral areas is not preferred by cyclists.

 

Architectural Form:

Differences Between Old and New Urban Areas: Differences between old and new urban areas are notable. In the southern new city, FAR (similar as SGF and PSP) show insignificant effects, while AGF in the old city displays widespread insignificance. This reflects the impact of distinct planning and spatial layouts from different eras on leisure cycling, emphasizing the need for tailored urban design strategies.

 

Threshold Effect:

Road Network Accessibility and Density: The influence of road network accessibility and density stabilizes once a certain threshold is reached. This suggests that further improvements in these metrics do not significantly enhance cycling potential. This conclusion, based on the MGWR results and the 95% confidence interval significance comparison (Figure 7), guides future greenway planning and urban development strategies to focus resources on other impactful factors.

 

These detailed analyses and findings, presented in the discussion section, underscore the novel contributions of our study. They are grounded in the data and existing results, providing a specific and thorough presentation of how various independent variables impact leisure cycling.

 

 

Thank you once again for your valuable feedback.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for your extensive efforts to improve the study. They address the previous concerns well. I have a couple of minor suggestions: 

- The last paragraph in the introduction would start with: "Specifically, this research focuses on several key items..." or somethings similar since the following expressions are not in the questions form.

-Is there a specific reason using italics on page 3 lines 95-104? If not, please revise.

- Please double-check the resolution and borders of Figure1. 

- "While street view imagery used for urban design analysis is typically captured from 381 the perspective of cars, it provides valuable insights into the built environment that are 382 relevant for non-motorized transport studies." better to provide a citation for this statement. 

- For all colored figures please double check the resolution and readability.

- I suggest including limitations and future research in the conclusion. 

 

Author Response

Dear Reviewer,

 

We appreciate your recognition of the improvements made in the revised manuscript. Thank you very much for your detailed comments and suggestions on our manuscript in the 2^nd round. Below is our response to your concerns:

 

Comment #1: The last paragraph in the introduction would start with: "Specifically, this research focuses on several key items..." or somethings similar since the following expressions are not in the questions form.

 

Response:

Thank you for pointing this out. We have revised the last paragraph in the introduction to start with "Specifically, this research focuses on several key items..." as recommended. This change ensures that the following expressions are clearly stated in a non-question format.

 

Comment #2: Is there a specific reason using italics on page 3 lines 95-104? If not, please revise.

 

Response:

Thank you for pointing out the use of italics on page 3, lines 95-104. The italics were not intentional and occurred due to formatting inconsistencies during the editing process. We have revised the text to correct this issue.

 

Comment #3: Please double-check the resolution and borders of Figure1.

 

Response:

Thank you for your suggestion regarding the resolution and borders of Figure 1. We have updated the figure to a clearer version with improved resolution and borders. Additionally, all figures, graphics, and images in the manuscript have been double-checked for resolution and readability.

 

Comment #4: "While street view imagery used for urban design analysis is typically captured from 381 the perspective of cars, it provides valuable insights into the built environment that are 382 relevant for non-motorized transport studies." better to provide a citation for this statement.

 

Response:

Thank you for your suggestion to provide a citation for the statement regarding street view imagery. We have added the appropriate reference to support this statement in the manuscript.

 

Comment #5: For all colored figures please double check the resolution and readability.

 

Response:

Thank you for your suggestion to double-check the resolution and readability of all colored figures. We have reviewed and ensured that all figures, graphics, and images in the manuscript have clear resolution and high readability. The original image files have been uploaded for verification.

 

Comment #6: I suggest including limitations and future research in the conclusion.

 

Response:

Thank you for your suggestion to include limitations and future research in the conclusion. We have revised the conclusion to reflect the limitations and potential future research directions discussed in Section 4.5.

 

Thank you once again for your valuable feedback.

Reviewer 3 Report

Comments and Suggestions for Authors

The topic is intriguing and aligns well with the journal. The practical approach meets scientific standards. The revised article significantly enhances the previous version. Its structure now fulfills the requirements of a scientific publication. The sentences have been shortened, making them more understandable for readers. The conclusion review now aligns with the research results.

Author Response

Dear Reviewer,

Thank you for your positive feedback and suggestions. We appreciate your recognition of the improvements made in the revised manuscript.

In response to your comments:
- We have reviewed the empirical research results section to ensure that all figures, tables, and textual explanations are as clear and detailed as possible.
- Additionally, we have double-checked the references to ensure that all statements are adequately supported and that recent and relevant literature is included.
- To address any potential concerns about the resolution and readability of our figures, we have separately uploaded the original image files for all figures, graphics, and images used in the manuscript.

Thank you again for your valuable feedback.

Best regards.