Quantifying the Rate and Extent of Urbanization Effects on Vegetation Phenology in Mainland China

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The study investigates the impact of urbanization in China on the sensitivity and variability of vegetation phenology, which is a topic of significant importance. By analyzing the response of vegetation phenology to urbanization, this study can help to understand how urbanization changes the seasonal dynamics of vegetation. Moreover, the study considers the combined effects of multiple environmental factors (such as temperature, artificial light, CO2 emissions, and PM2.5 concentrations) on vegetation phenology, which helps to reveal the complex mechanisms underlying phenological changes in the context of urbanization.

 

Main Strengths: The study uses road-network density (RND) as a proxy for urbanization intensity and combines linear regression and partial least-squares structural equation modeling (PLS-SEM) to provide a new method for assessing the impact of urbanization on vegetation phenology. Covering 31 major cities in China over the period of 2014-2022, the study has a rich dataset that can effectively reflect the long-term impact of urbanization on vegetation phenology. The study comprehensively considers the effects of multiple environmental factors on vegetation phenology, including temperature, artificial light, CO2 emissions, and PM2.5 concentrations, making the results more comprehensive and in-depth.

 

Limitations:

1. The study innovatively uses road-network density to quantify urbanization intensity, but there is a lack of argumentation on the reliability of this method in the introduction and methods sections, and no relevant literature is cited. It is suggested to conduct a detailed discussion on the applicability of this method.
2. The study selects multiple cities as research objects, but how are the boundaries of each city defined? Are they the administrative boundaries at the city level? Are urban areas, suburbs, and mountainous regions distinguished? It is suggested to include the remote sensing images of the 31 cities as supplementary figures.
3. The saturation point is a key aspect of this study and should not be referenced only to literature [34]. A brief introduction to the method is recommended.
4. The expression of the model method is incorrect. In SEM, latent variables refer to unmeasurable variables. The indicators in this paper are all measured values, not latent variables.
5. In the analysis based on land-use types, the first three types of vegetation have certain climatic zonality, while the latter three do not and may be widely distributed throughout the country. It seems unreasonable to combine them in the analysis.
6. The content of Fig. S3 is not fully explained. What does R2 on the figure mean? How is the urban-rural gradient determined? These are not introduced. Moreover, there are extra (a) and (b) on the figure, indicating that the graphing was not done carefully.
7. There are extra periods in the tables (e.g., Table S2), and some CO2 is not changed to subscript. These minor issues need to be checked carefully.
8. I suggest providing the absolute values of SOS, EOS, and LOS for each city as supplementary figures (in the form of boxplots or histograms).

Author Response

All point-by-point responses and corresponding revisions are included in the Word file.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

Comments can be found in the attached PDF

Comments for author File: Comments.pdf

Author Response

All point-by-point responses and corresponding revisions are included in the Word file.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

This study focuses on analyzing the sensitivity and variability of vegetation phenology responses to urbanization in mainland China, holding significant scientific value and practical relevance. The topic addresses hot issues such as urban sustainable development and climate change. Using road network density (RND) as a proxy for urbanization, it quantifies the response patterns of vegetation phenology (SOS, EOS, LOS) to urbanization in 31 major Chinese cities (2014-2022) by employing the slope (sensitivity) and range (variability) derived from linear regression between phenology and RND. Partial Least Squares Structural Equation Modeling (PLS-SEM) was used to assess the direct and indirect effects of RND on phenology through all five key environmental factors. However, the study has some limitations. For instance, the specific mechanisms by which urbanization affects each environmental factor, ultimately leading to changes in vegetation phenology, are not clearly elucidated. Additionally, while the PLS-SEM model can handle complex causal relationships, its construction may involve some subjectivity: the rationale and justification for combining the environmental factors ALAN (artificial light at night) and CO₂ emissions into a single latent variable (ALCO) require further explanation and validation. Furthermore, the explanation of the impact mechanisms in the Results and Discussion section is somewhat insufficient. Future studies could enhance the depth of discussion by incorporating field observation data. Addressing these limitations would significantly improve the paper's scientific rigor and persuasiveness.

Specific Reviewer Comments:

1.Introduction: Is the justification for using RND as an indicator of urbanization intensity sufficiently argued? Although the authors mention that RND is superior to the traditional urban-rural gradient (URG) method, the text lacks comparative validation with other urbanization indicators (e.g., population density, impervious surface ratio). It would be beneficial to supplement this with explanations or cite literature demonstrating the general applicability of RND in similar studies.

2.Section 2.1 (Study Area Selection): Whether the selected cities comprehensively represent the urbanization characteristics and vegetation types across all regions of inland China remains questionable. For example, some medium/small cities or cities in special geographical regions may have unique urbanization patterns and vegetation phenology responses not captured in this study.

3.PLS-SEM Model Construction: ALAN and CO₂ were combined into the latent variable ALCO (due to their strong correlation), but the statistical basis for this combination was not provided, nor was the sensitivity to other latent variable groupings tested. It is recommended to provide validation metrics for the combined latent variable or discuss the impact of different grouping approaches on the model results.

4.Figure 5: An exponential function was used to fit the non-linear relationship between graded RND and phenology metrics, but the rationale for choosing an exponential model over alternatives (e.g., logarithmic or piecewise linear) was not explained. It is suggested to compare multiple models or justify the suitability of the exponential function.

5.Caption for Figure 4: The caption beneath the figure fails to indicate what panels (a), (b), and (c) represent. It is recommended to supplement the explanatory text below the figure with: (a) Intercept distribution, (b) Slope distribution, (c) Range distribution.

6.Results Section: Discussion of model uncertainty could be added to enhance the scientific credibility of the results.

7.Data Time Range: The temporal scope is limited by road network data availability (2014-2022), potentially missing long-term urbanization effects. It is recommended to explicitly discuss the impact of this time range on the conclusions in the Discussion section.

Author Response

All point-by-point responses and corresponding revisions are included in the Word file.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

Dear Authors,

Thank you for the opportunity to review your manuscript. It is well-written and clearly presented, with a sound methodology and well-explained findings.

My primary comment relates to the scientific contribution of the study. While factors such as particulate matter, temperature, precipitation, and urbanization have been widely explored in the literature, along with their interactions, I believe your work has the potential to stand out by emphasizing aspects that are less frequently addressed. In particular, your inclusion of land-cover types is a valuable addition. I encourage you to further highlight these differences and delve deeper into the changes observed across various land-cover categories. This could significantly enhance the originality and impact of your work.

Further (minor) comments are included in the attachment.

Best regards,

Comments for author File: Comments.pdf

Author Response

All point-by-point responses and corresponding revisions are included in the Word file.

Author Response File: Author Response.docx

Reviewer 5 Report

Comments and Suggestions for Authors This paper examines the sensitivity and variability of vegetation phenology to urbanization across 31 cities in Mainland China from 2014 to 2022. The authors utilized road-network density (RND) as a proxy for urbanization intensity and employed linear regression and partial least-squares structural equation modeling (PLS-SEM) to assess the impact of RND on phenological metrics, such as the start of season (SOS), end of season (EOS), and length of season (LOS). The study identifies that urbanization, especially in its early stages, leads to significant changes in vegetation phenology, with temperature, artificial light at night (ALAN), CO2 emissions, and particulate matter (PM2.5) being key environmental drivers. The study offers insights into how urbanization's effect on vegetation phenology varies across latitudes, hydrothermal conditions, and land cover types. These findings provide valuable guidance for urban planning aimed at minimizing the disruption of natural ecosystems. The author should be a graduate student, therefore, the following are some of my sincere opinions, hoping to master scientific research methods more standardly. 1. Misalignment with Remote Sensing Focus: The study focuses mainly on analyzing geographical data and does not fully leverage remote sensing technologies, which are central to the journal's scope. While MODIS data is used, the analysis does not delve into remote sensing methodologies such as satellite image analysis, object detection, or classification, which would be more appropriate for the journal. I recommend rejecting this paper, as it does not meet the technical scope of the journal. 2. Lack of Novelty in Conclusions: The study provides conclusions that are somewhat predictable based on existing knowledge of urbanization and phenology. For example, the impact of temperature, ALAN, and CO2 emissions on phenological metrics has been well-established in prior research. The paper would benefit from more innovative insights or a novel methodological approach to make its findings stand out. 3. Data Quality and Limitations: The paper relies on publicly available data with relatively coarse resolutions (500 m for MODIS data). This can introduce significant spatial uncertainty, especially for fine-scale urban studies. The authors should discuss the limitations more clearly, specifically the impact of data resolution on their findings. 4. Methodology is Insufficiently Explained: The use of PLS-SEM is mentioned but not explained in enough detail. The paper should provide more clarity on why this method was chosen, particularly in the context of vegetation phenology and urbanization. A clearer explanation of how PLS-SEM contributes to understanding the indirect and direct effects of environmental factors on phenology is needed. 5. Graphical Quality Needs Improvement: Several figures in the paper, including heatmaps and regression plots, suffer from poor quality in terms of color consistency and legibility. For example, the use of inconsistent color schemes across graphs is visually distracting. The authors should adopt a more professional and uniform graphical style, following high-quality publications in the field. 6. Inconsistent Terminology: Throughout the paper, the terminology related to environmental factors and phenology is not always used consistently. For example, "vegetation phenology" and "phenological metrics" are used interchangeably. The paper should adhere to a single, well-defined set of terms to ensure clarity for the readers. 7. Limited Consideration of Biotic Factors: While the paper includes environmental factors such as temperature and PM2.5, it overlooks biotic factors such as species composition, which could significantly influence vegetation phenology. Future studies should incorporate biotic elements or explicitly mention this limitation. 8. Over-reliance on Road-Network Density (RND): The use of RND as a proxy for urbanization intensity is innovative but limited. While the study argues that RND provides a continuous measure of urbanization, it would benefit from a discussion of other potential proxies or methods that might yield complementary insights, such as population density or satellite-derived impervious surfaces. 9. Superficial Discussion of Mechanisms: The paper touches on the mechanisms driving the observed phenological changes but does not provide enough in-depth analysis of how these factors interact at a finer scale. A more detailed discussion on the biophysical processes behind the sensitivity and variability of phenology to urbanization would be valuable. 10. Lack of Robust Validation: There is no clear validation of the models or results presented. For instance, it would be important to compare the study's predictions with real-world field data or other established models to assess the reliability of the results. Without such validation, the findings may not be convincing. 11. Data Presentation Issues: While the figures present the data, the lack of proper labeling and caption clarity hampers the ease of interpretation. For example, some graphs lack axis labels or units, making it difficult to understand the scale of measurements. The authors should pay more attention to the presentation of their results to improve readability. 12. Recommendation for Improved Statistical Analysis: While the paper uses linear regression and PLS-SEM, it would benefit from more advanced statistical techniques, such as machine learning models or time-series analysis, to predict the impact of urbanization on phenology more accurately. Incorporating these methods could enhance the robustness and predictive power of the research. In conclusion, the paper has potential but falls short in aligning with the journal’s focus on remote sensing and in providing new insights or robust methods. The quality of the visuals and the depth of the analysis must be improved. Given these issues, I recommend rejecting the paper in its current form.

Author Response

All point-by-point responses and corresponding revisions are included in the Word file.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

It can be seen that the author has made very meticulous and comprehensive revisions. Now I only have two minor questions:

First, for Fig. S3, each city has a separate legend with very small and unclear fonts. It is recommended to combine them into one clear legend.

Second, using the distance from the urban core (buffer zones) to measure the urban-rural gradient is a feasible method, but there are still some uncertainties. In most cities, urban development is spatially asymmetric. At the same distance from the urban core in different directions, the degree of urbanization can vary. For example, in Beijing, there are more mountains in the northwest and more plains in the southeast. It is suggested to briefly analyze the above uncertainties in the discussion section.

Author Response

I appreciate your thoughtful feedback and the opportunity to improve the manuscript. All point-by-point responses are in the Word file.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

I have no more questions.

Author Response

Thank you for your time and for providing such constructive feedback.  We sincerely appreciate your thoughtful suggestions.

Reviewer 4 Report

Comments and Suggestions for Authors

Dear Authors,

Thank you for revising the manuscript in line with the suggestions. The improvements are substantial, and the paper is now much closer to being ready for publication. I have included a few additional comments in the PDF. Given the information I requested throughout the text, I would kindly suggest addressing it in the discussion section as well. This could help explain why a notable portion of the samples or cities do not align with the observed patterns.

I haven't noticed this before: "the environmental and phenological indicators were derived from different years". So I would like to see additional information on the data. I don't understand how environmental factors and phenology could be analyzed if the data is not from the exact same period? 

Comments for author File: Comments.pdf

Author Response

Thank you very much for your valuable comments and suggestions. All point-by-point responses are in the Word file.

Author Response File: Author Response.docx

Reviewer 5 Report

Comments and Suggestions for Authors

The revision is very meticulous, and the current version meets the publication requirements.

Author Response

Thank you for your time and for providing such constructive feedback. We sincerely appreciate your thoughtful suggestions.