Predicting Land Use Changes in Philadelphia Following Green Infrastructure Policies

Round 1

Reviewer 1 Report

I commend the authors for their work. The research is sound and well researched. The paper is also well written and adds knowledge-base to the topic. The paper considers the challenge of predicting land use changes in Philadelphia following green infrastructure policies. The authors point out that the study use of potential models and Markov Chain analysis to predict high resolution land use changes in Philadelphia, PA for the year 2036 was validated. The accuracy of the validation, using Kappa statistics and disagreement scores is also correct. Overall, the article makes a valueable contribution to our understanding of the issue. However, a few minor revisions should be addressed:

Figure 5 – The map is not clear, increase the map size as this is a very conclusive part of the finding and very important in demonstrating GI changes.

Figure 6 – Same as Figure 5, increase map size.

Line 337 – capitalize Table 5

Line 362 – capitalize Table 6

Line 495 – remove (UHI) as it is not referenced as an acronym in the paper

I therefore recommend accepting the article after some minor revisions.

Author Response

Dear Reviewer,

Thank you for your kind words. We have corrected for all of your comments accordingly. We appreciate the time and effort you have given to review our research.

Sincerely,

Charlotte & Peleg

Author Response File: Author Response.pdf

Reviewer 2 Report

Introduction

Table 2 What do you mean by respective LULC models? How were the boundaries of the city set to make these calculations?

Materials and Methods

I would like to see the aim at the end of the introduction section, I do not see the point of the first paragraph in methods, since it is a summary

Why not using the OBIA method, or the same method for the three analyzed dates?, that should make the LULCC more robust when comparing different dates

Please provide the resolution of the imagery used in text too.

Table 3 the Philadelphia GI were the existing or the projected or both? 

3.2.3 There are not future conservation areas in the 2036 projection that should be included as a restriction?

3.3 I need further explanation on how the GI was modeled, how did you identify the future scenarios of GI, how do you allocate different GI or where to allocate those, based on which rules?

It would have been great to validate the GI predicted with the authority in charge of GI, to see the likelihood far than just statistics

The results from the projected GI look very clustered, would you mind explaining the pattern?, I am impressed over the large loss of tree canopy, It also seems in some areas 2015GI has become tree canopy in 2036? Could you show on the maps of gain and losses which one correspond to GI that you mention on text?

Table 7 you need to disclose what the acronyms mean at the end of the table, since it should be self-explanatory, also revisit the caption

I do not see the usefulness of landscape metrics for the manuscript, I would like a further explanation on how those connect with GI. It seems from your discussion that just connectivity metrics would be enough to evaluate the urban landscape, and that especially depends on the type of GI put on the ground.

I think it would be really important to account for the implications of increasing GI, discussing the distribution of those in terms of equity, and refer to which type of GI is expected to be found where according to your simulations

It would be great to see a discussion towards the robustness of your simulation contrasted with projected growth by the local or regional government and discuss the agreement of the model with planning authorities, so the work is not merely an academic exercise and can be seen as something the would showcase current decision making.

Author Response

Dear Reviewer,

Thank you for edits, comments, and suggestions. We have condensed the paper by a few pages, making it more concise, while also enlarging the maps so they can be understood better. We addressed all of your comments accordingly, removed some of the spatial metrics, and have added some context in the discussion to match your comments. 

We appreciate the time and effort that you have dedicated to reviewing our research.

Sincerely,

Charlotte Shade

Author Response File: Author Response.pdf

Reviewer 3 Report

On line 107 the authors write: The aim of this study is to forecast the LULCCs resulting from continued urbanization and the GI policies in Philadelphia, Pennsylvania, USA for the year 2036 using detailed landcover models at a very high resolution,” and on line 156 the authors note: “The aim of this study is to assess the impact of nature-based solution policies on landcover change and future spatial urban structure” then in the conclusion state (line 507): “This study finds that LULCCs from GI policies can successfully be modelled in a heterogeneous intra-urban environment using fine scale data.” These are somewhat out of agreement—is the point of the study to evaluate the impact of policy or forecast change or determine whether GI policy can be modeled. These are related but different approaches. I mention this because I think that focusing on one could help resolve a concern I have with the paper, which is it is too long. The introduction is focused on modelling methods, which suggests that is the true focus. But I’m not sure all of this is needed as it is more literature review than I think an empirical article should have. (This is obviously a personal preference, I just think the results will be more widely read if the paper is more concise. As presented the paper appeals to remote sensing audiences more than green infrastructure audiences.)

Considering all of the details about previous studies, the authors don’t return to how their results show improvement over what came before. Obviously finer scale matters, but the literature presented isn’t directly informing the original analysis. Again, I think the paper can be shortened and tightened quite a bit.

Also in the literature review, I think Table 1 may not be needed and what information is required can be discussed in the context of Table 2.

Since policy is important as an assumption, or at least related, I’d like to see what policy actually looks like (e.g. what are “Nature-based solution policies?).

Why are 2008-2015 used as points for historical data, and can these been improved with additional years? What are limitations of using these years?

Author Response

Dear Reviewer,

Thank you for edits, comments, and suggestions. We have condensed the paper by a few pages, making it more concise, while also enlarging the maps so they can be understood better. We addressed all of your comments accordingly, added some context on nature-based solution policy, and have added some clarification in the methods to match your comments. 

We appreciate the time and effort that you have dedicated to reviewing our research.

Sincerely,

Charlotte Shade

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

From my point of view the authors have addressed all my concerns, therefore I am OK for the paper to get publish in this revised version