Trend Analysis of Urban Heat Island Intensity According to Urban Area Change in Asian Mega Cities

Round 1

Reviewer 1 Report

The paper presents an attempt to evaluate the relationship between urban development and UHI trends for eight Asian megacities. The use of gridded data (CHELSA) for estimating UHI trends is a new approach, so I think the paper deserves to be published, although the efficiency of the methodology may be open to further discussion. However, it is desired to add some explanation concerning the validity of the data and method.

[Main comments]
The validity of the study is crucially dependent on whether the CHELSA temperature can represent the actual UHI, namely, the temperature distribution for each city area. In order to show the validity of the methodology, it will be needed to explain the outline of the way of constructing the CHELSA data, namely, whether and how the temperature data in each city area have been assimilated.

The authors stress the importance of temperature data constructed on unified conditions among cities (Lines 77-78), but it is possible that gridded data for different cities were based on data having different spatial resolution and observational practice. It is desired to add some explanation and/or discussion on this matter.

It will be better (although not mandatory) to present an example of temperature distribution that indicates that a UHI is well captured by the CHELSA data. Although the temperature field for Seoul is shown in Fig.2, no indication of UHI is seen therein.

Another major point is that the UHI intensity in Dhaka (Fig.5) is close to zero for a large part except in summer. This seems unnatural, so I would recommend checking whether there was any problem in the analysis. For example, was the urban area (urban boundary) properly defined?

[Other comments]
@ I understand that the analysis was based on urban areas that changed year by year, and so the UHI intensity was defined for different areas according to years. It is better to write this situation explicitly.

@ Please explain the definition of "surrounding area" used for defining UHI intensity. For example, was "surrounding area" defined by the area within a specified distance of the urban boundary?

@ In relation to the previous comments, isn't there any possibility that the UHI trend is affected by the change in the definition area that has uneven distribution of natural temperature (e.g., uneven elevation)? It may be better to make a brief discussion of this possibility.

@ Line 9 "possibility of causing global scale climate change" --- It is believed that the contribution of urban warming on the global climate change is negligible (e.g., IPCC AR4, AR5), although there have been a number studies on separating urban and global temperature change (as cited in Lines 47-48). The expression in Line 9 should be changed in order to avoid misunderstanding.

@ Line 109 "LC" --- Please write the full spelling (land cover?) at the first usage.

@ Why is T1+T2+T3 divided by 30 in Eq.(1)?

@ Please avoid overwriting a city name on the city area in Fig.1.

@ Please make the color scale in Fig.2b easier to see.

@ It is better to fix the number of digits in each table. For example, "0.00" in Table 1 should be written as "0.000"

Author Response

Thank you for your comment. Based on the comments, the manuscript has been revised.

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

Please see attached.

Comments for author File: Comments.pdf

Author Response

Thank you for your comment. Based on your comment, The manuscript has been revised. 

Please see the attachment 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

I appreciate the authors' effort of revision. However, I would like to confirm the following points.

One of my previous comments was intended to ask whether CHELSA has successfully captured spatial temperature variations due to urban effects. It will be better to make more specified explanation on this point, although the authors have mentioned the use of "remote-sensing data" (Line 120) that implies continuous spatial coverage of land use and/or surface temperature.

I appreciate that the authors have added Figure 2, but it is better to make clearer the legend of color tones in Fig.2b. In addition, I wonder why the area outside the southeastern border of the city is uniformly in an orange color that corresponds to the highest temperature level, in contrast to the (seemingly) lower temperature over the "surrounding areas" (hatched area) within the city. The discontinuous color changes across the city boundary in Fig.2a will also need explanation.

[Other comments]

@ I think it will be better to write explicitly that "correlation" in this study indicates the correlation between yearly values of urban area and UHI intensity.

@ The meaning of the expression "the dependency of UHI intensity on the economic situation did not show any specific trend" (Line 14) in the Abstract is not clear. Do you want to say the same thing as "the UHI intensity varies depending on --- economic situation" (Line 301) in the Conclusion?

@ Please make clear the meaning of "the degree of constructed data was not taken into consideration" (Line 53).

Author Response

"Please see the attachment."

Author Response File: Author Response.docx

Reviewer 2 Report

Thanks for addressing the comments. 

For the time series analysis is it necessary to try to do a linear fit, specially since obviously they are not a linear fit.  A discussion on the trends could suffice instead of forcing a fit.  Or may be an alternative method which can justify the trends could be used.

There are still some grammatical and editorial errors that need to be addressed.

Author Response

 "Please see the attachment."

Author Response File: Author Response.docx