Examining Land Use/Land Cover Change and the Summertime Surface Urban Heat Island Effect in Fast-Growing Greater Hefei, China: Implications for Sustainable Land Development
Round 1
Reviewer 1 Report
This manuscript investigates the relationship between LULC dynamics and SUHI based on surface skin temperatures. Taking Hefei as an example, the authors studied in detail the LULC changes and thermal conditions in the city and its periphery.
English needs to be polished. The authors should generally avoid overly long sentences, nested subordinate clauses, and convoluted wording. The major drawbacks of this manuscript are as follows:
- Research design:
- There are obvious major limitations of this study, such as the inclusion of only 4 morning overpass times in summertime in 4 different years to study SUHI. However, Landsat satellites passed the study area only in the morning, thus, this method cannot generalize (a) seasonal characteristics of the SUHI (summer, winter, spring, fall). Also, since the SUHI normally reaches its greatest magnitude around solar noon, your LST values do not coincide with the time of greatest heating nor the time of heightened heat-health risks (later afternoon, evening). Theoretical and methodological analyses are missing. Therefore, the applicability of this study to inform the decision-making process is challenging and not convincing.
- Introduction:
- This explanation of UHI types is also ignored, and not supported by appropriate references. Specify how each of the UHI types differs from one another, relative to the instrumentation used, media sensed, relevant spatial scales, causal processes, etc. The explanation of UHI (surface) is conceptually weak.
- The 3rd paragraph is incoherent without a theme. Also, I suggest the author review the existing literature on urban expansion and SUHHI and identify research gaps. The contribution is unclear without the research gap and research aims/questions.
- Study area:
- Might be helpful to suggest "why" such studies are needed in Hefei (e.g., for heat mitigation, land use planning, other?).
- Materials and Methods:
- A workflow can make the methods on 6 steps clearer.
- Line 115-117: The sentence is overly long. Please consider dividing it into short sentences.
- Line 118: How many LULC classes and what is the temporal resolution are there in the commercial LULC maps?
- Line 120: How does the ‘supervised signature extraction with maximum likelihood’ method differ from the maximum likelihood classification? If the authors do not intend to elaborate on the method here, please give references.
- Discussion:
- The discussion could be more in-depth, in particular with regard to the finding of the urban-rural corridors, as this is the innovative aspect of the article.
- The authors mentioned that PLS regression is an explicit approach capable of removing the multi-collinearity among the variables. However, the related theory of PLS was not explained. Population density and fraction of developed land are highly correlated, which is a piece of common knowledge. Both variables highly contribute to the increase of SUHI but it could not be explained the collinearity between the two variables.
- The implication of urban planning policies is discussed. However, the policy implications discussed are superficial. For example, the impacts of population density are not discussed.
- A comparison between results based on different measures would significantly enhance the quality of this study.
- Conclusions:
- Line 406-409: The sentence is overly long. Please check and try to separate the sentence into several shorter ones. Also, the use of “as evidence with/by” in this place and throughout the manuscript is super vague.
Author Response
This manuscript investigates the relationship between LULC dynamics and SUHI based on surface skin temperatures. Taking Hefei as an example, the authors studied in detail the LULC changes and thermal conditions in the city and its periphery.
English needs to be polished. The authors should generally avoid overly long sentences, nested subordinate clauses, and convoluted wording. The major drawbacks of this manuscript are as follows:
Response: Many thanks for your valuable comments and suggestion. We carefully revised this manuscript accordingly. Then, we have this manuscript be polished by the native English speaker via MDPI’s language editing service.
- Research design:
- There are obvious major limitations of this study, such as the inclusion of only 4 morning overpass times in summertime in 4 different years to study SUHI. However, Landsat satellites passed the study area only in the morning, thus, this method cannot generalize (a) seasonal characteristics of the SUHI (summer, winter, spring, fall). Also, since the SUHI normally reaches its greatest magnitude around solar noon, your LST values do not coincide with the time of greatest heating nor the time of heightened heat-health risks (later afternoon, evening). Theoretical and methodological analyses are missing. Therefore, the applicability of this study to inform the decision-making process is challenging and not convincing.
Response: Thanks for your comments. We are answering your questions via following aspects: (1) This study intends to depict the relationship between LULC change and summertime SUHI effect, given Greater Hefei is subject to the extreme UHI effect in summer but the UHI effect is weak or negligible in the other seasons. We then corrected this manuscript’s title accordingly.
(2) On the application of satellite thermal remote sensing for UHI study, there have been numerous studies using the coarse TIR bands such as MODIS and relatively high-resolution bands of Landsat series satellites. The MODIS can provide four TIR images everyday with relatively higher temporal resolution than the Landsat series images. However, the MODIS TIR images are too coarse to spatially explicate the patterns of SUHI effect for Greater Hefei, given the coarse LULC classification. For instance, some small patches of water bodies and vegetated in the urban and around the urban area were mixed with the impervious surface. We admit that for Landsat series satellites, their overpassing time during 10:30-11:00 at local time makes the retrieved LST values be lower than the extreme high LSTs on noon or the other time. But, due to the limitation of satellite-based thermal remote sensing, Landsat series images undoubtedly the practical datasets that can be integrated into city-level master planning and guide the policy towards mitigation of UHI effect. It is noteworthy that the Landsat series images can provide more detailed LULC information and generate accurate LST than the MODIS images, though Landsat series images have low temporal resolution.
(3) Due to the weather limitation, particularly the cloud contamination, only four high-quality Landsat images acquired in the summer were used for this study. As you concern, the retrieved LSTs are lower than the true values during the solar noon and the other time. But, only a few weather stations sparsely set in the urban and rural sites failed to provid the spatially explicit UHI effect information covering the whole study area. Alternatively, combined with the LULC change, the Landsat images overall depicted the spatial patterns of the LSTs and SUHI intensity and associated extent influenced by the SUHI effect. Moreover, for Greater Hefei, the 30-m maps of LULC, SUHI intensity, and associated extent influenced by SUHI effect can provide sufficient information for city/regional decision-making across sustainable land development and UHI mitigation.
- Introduction:
- This explanation of UHI types is also ignored, and not supported by appropriate references. Specify how each of the UHI types differs from one another, relative to the instrumentation used, media sensed, relevant spatial scales, causal processes, etc. The explanation of UHI (surface) is conceptually weak.
Response: Thanks for your comments. We then strengthened these essential explanations and listed some appropriate references accordingly. Please see the highlighted words and sentences.
- The 3rd paragraph is incoherent without a theme. Also, I suggest the author review the existing literature on urban expansion and SUHHI and identify research gaps. The contribution is unclear without the research gap and research aims/questions.
Response: Thanks for your comments. Based on literature review, we then corrected this paragraph accordingly, focusing on the researching gaps.
- Study area:
- Might be helpful to suggest "why" such studies are needed in Hefei (e.g., for heat mitigation, land use planning, other?).
Response: Thanks for your comments. We then corrected this paragraph by adding the new contents that may interest the readers.
- Materials and Methods:
- A workflow can make the methods on 6 steps clearer.
Response: We added a workflow named Figure 2.
- Line 115-117: The sentence is overly long. Please consider dividing it into short sentences.
Response: Many thanks. We rephrased this paragraph to make it more readable.
- Line 118: How many LULC classes and what is the temporal resolution are there in the commercial LULC maps?
Response: The commercial LULC maps have six level-1 classes (including Built/developed land, Forest/shrub, cropland, water bodies, grassland, and bare land) and further 25 level-2 classes. But in this study, the level-1 classes were adopted because the LULC classes in Greater Hefei were relatively simple. These LULC maps were produced with the multiple Landsat images on demand. In this study the LULC products in 1995,2002,2007, and 2016 were purchased.
- Line 120: How does the ‘supervised signature extraction with maximum likelihood’method differ from the maximum likelihood classification? If the authors do not intend to elaborate on the method here, please give references.
Response: The maximum likelihood classification (MLC) methods include hierarchical MLC, supervised MLC, and fuzzy MLC, and so on. We think the so-called supervised signature extraction with maximum likelihood’ method is the same as the well-known supervised maximum likelihood classification method, since there are different menu names among the Remote sensing and image analysis systems. Herein, we replace ‘supervised signature extraction with maximum likelihood’ with the ‘supervised maximum likelihood classification’ (reference see Chen, D., & Stow, D. The effect of training strategies on supervised classification at different spatial resolutions. Photogrammetric Engineering and Remote Sensing, 2002,68, 1155-1162).
- Discussion:
- The discussion could be more in-depth, in particular with regard to the finding of the urban-rural corridors, as this is the innovative aspect of the article.
Response: Many thanks for your valuable comments. We then added a figure (Figure 6) to show the proposed ecological network towards mitigating the SUHI effect along the urban-rural corridors (details see section 5.2).
- The authors mentioned that PLS regression is an explicit approach capable of removing the multi-collinearity among the variables. However, the related theory of PLS was not explained. Population density and fraction of developed land are highly correlated, which is a piece of common knowledge. Both variables highly contribute to the increase of SUHI but it could not be explained the collinearity between the two variables.
Response: To make the concise expression and to be coincide with some programming syntax in the R, we replaced the term PLS regression with PLSR across this manuscript. We added some points to strengthen the capability of PLSR models and explain the related theory of PLSR (please see sub-sections 3.2.5 and 5.1). Also, we refined the explanations of Population density and fraction of developed land in contributing to the increasing SUHI effect as with’ Given the SUHII was measured with the LST difference between the LULC categories, the relative importance of Pop_density cannot be directly embodied via its standardized coefficient, though there is the highly significant correlation between Developed land and Pop_density……Nevertheless, the relative importance of Pop_density in contributing to the SUHI effect indicators cannot be directly detected and thus may be underestimated, given the simply measured LST difference between the impervious surface and cooling surface but neglecting the interactions of the LULCs and Pop_density.’Please see the highlighted sentences in sub-sections 4.3 and 5.1.
- The implication of urban planning policies is discussed. However, the policy implications discussed are superficial. For example, the impacts of population density are not discussed.
Response: Thanks for your comments. We then strengthen the policy implications discussed in section 5.2.
- A comparison between results based on different measures would significantly enhance the quality of this study.
Response: Since the in-situ measured air temperature-based Canopy UHI (CHI) effect cannot be directly compared with SUHI effect, we then suggest the combination of thermally sharpened SUHII indicators and in-situ measured micro-climatic factors to generate the simulated CUHI indicators via the computational fluid dynamics (CFD), which can be compared with the real CUHI indicators.
- Conclusions:
- Line 406-409: The sentence is overly long. Please check and try to separate the sentence into several shorter ones. Also, the use of “as evidence with/by” in this place and throughout the manuscript is super vague.
Response: We ask the native English speaker to edit this manuscript, focusing on such vague expression.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The present work is well written:the methodology and the results are very well explained and
the importance of these studies is current subject of
interest in the international scientific community.
China is the biggest urbanization country and the research on
the consequences on the increase of the urban heat island
becomes of fundamental importance to understand the
mechanisms and feedbacks with global climate change.
Only a correction:the captions on Figure 3 and 4 are
not clear.
Figure 3: it is not clear because in the figure are present
(a) and (b) and in the caption are reported other letters.
Figure 4: the same
It seems the caption are esxchanged
Author Response
The present work is well written:
the methodology and the results are very well explained and the importance of these studies is current subject of interest in the international scientific community.
China is the biggest urbanization country and the research on the consequences on the increase of the urban heat island becomes of fundamental importance to understand the
mechanisms and feedbacks with global climate change.
Only a correction:the captions on Figure 3 and 4 are not clear. Figure 3: it is not clear because in the figure are present (a) and (b) and in the caption are reported other letters.
Figure 4: the same It seems the caption are esxchanged
Response: Many thanks for your kind and valuable comments. We then corrected these captions accordingly.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
In the "authors' response", the authors did not imply the lines where the related changes are. It is super hard for me to detect them and provide comments. I strongly suggest the authors do a well-written report as "authors' response". Thanks.
Author Response
Based on your valuable comments and suggestion, we tried to correct the manuscript and highlighted all the changes in red. Also, we highlighted some sentences in red based on the discussion among the authors. In other words, almost all the changes marked in red are in response to your major concerns to enhance quality of this manuscript. Besides, only the changes marked in red (Figures 4 and 5) is in response to Reviewer-2’s concern since he/she only requires us to alter the captions of former Figures 3 and 4 (which are present Figures 4 and 5) .
Author Response File: Author Response.docx
