U.S. Farmland under Threat of Urbanization: Future Development Scenarios to 2040

Round 1

Reviewer 1 Report

The authors investigated "U.S. Farmland Under Threat of Urbanization: Future development scenarios to 2040". 

Please check the English of the manuscript thoroughly.

Please check the structure of the work and modify the design.

An obvious conclusion is needed for this work.

What is the novelty of this work? Only the time of your investigation.

Author Response

The authors investigated "U.S. Farmland Under Threat of Urbanization: Future development scenarios to 2040".

Please check the English of the manuscript thoroughly.

** Response: Thank you for your comments and time invested in this paper. Our writing in the revised manuscript has been thoroughly proofread by a professional editor.

Please check the structure of the work and modify the design.

** Response: Our co-authors and professional editor have thoroughly checked the structure of the manuscript, and we all agreed that our current version is clear in terms of method design and result presentation. It might be our writing that made you feel less straightforward or confused, so we have thoroughly checked and revised the language in the revised manuscript.

An obvious conclusion is needed for this work.

** Response: Thank you again for the suggestion. Our conclusion section highlighted the amount (18.2 million acres) and spatial patterns (southern states and metropolitan areas) of possible cropland losses across the continental United States due to urbanization by 2040, which we think are the most important information we want to deliver. Based on these findings, we further highlighted the necessity of cooperative efforts from local to federal policymakers to protect croplands from urbanization. Because urban-cropland interactions are complicated and vary across regions, users could access our online application (http://development2040.farmland.org) and conduct an analysis and make conclusions specific to the areas they are interested in.

What is the novelty of this work? Only the time of your investigation.

** Response: Unlike others, our model can distinguish between highly developed land uses which result in the permanent loss of farmland and low-density residential land use that can limit the options for agricultural production. In addition, we use scenarios of development to show the impacts of doing nothing or doing something at the CONUS level by a more nuanced analysis of ag land quality that includes soils, type of production, production limitations, and land cover. Just as important, we further created an online application that allows users to zoom into their counties and download statistics to help transform the research into actionable policy actions at the county, state, and federal levels.

Reviewer 2 Report

The growth of urbanization has dramatically changed land use. It is of great scientific value to analyze how to analyze the process of land use change in the process of urbanization development.

This paper provides an in-depth study of the relationship between urbanization and land use in the United States using a machine learning approach. The research method is advanced and the results are reliable.

By analyzing the urbanization process in the U.S. during the historical period, we can provide basic data and boundaries for future farmland conservation, etc.

I think this paper is fully worthy to be published in LAND journal, which can promote the scientific community's in-depth understanding of land use and arable land conservation.

There are some additional questions below.

1) Some overviews of land use research in other parts of the world could be added to the research background section as appropriate.

2) Some research on machine learning in land use prediction has been carried out, and the authors can appropriately present the prospects for the application of different machine learning models or artificial intelligence methods for land use, especially for conversion between multiple land forms.

3) Relevant maps, at least in the first one, need to add basic elements such as the compass and scale. It is convenient for international readers to understand the context of the study area.

Author Response

Reviewer 2

The growth of urbanization has dramatically changed land use. It is of great scientific value to analyze how to analyze the process of land use change in the process of urbanization development.

This paper provides an in-depth study of the relationship between urbanization and land use in the United States using a machine learning approach. The research method is advanced and the results are reliable.

By analyzing the urbanization process in the U.S. during the historical period, we can provide basic data and boundaries for future farmland conservation, etc.

I think this paper is fully worthy to be published in LAND journal, which can promote the scientific community's in-depth understanding of land use and arable land conservation.

** Response: Thank you for your comments and time invested on this paper. We have revised the manuscript based on your comments accordingly. Please refer to the revised manuscript for each modification and thanks again for your further efforts.

There are some additional questions below.

1) Some overviews of land use research in other parts of the world could be added to the research background section as appropriate.

** Response: Thank you for the suggestion. In the revised manuscript, we have cited more research focusing on urbanization-induced cropland losses across different countries such as China, India, Canada, Indonesia, Nigeria, Egypt, and Indonesia, as well as studies on globe analysis (Lines 48-53 and Lines 88-92 of the revised manuscript).

2) Some research on machine learning in land use prediction has been carried out, and the authors can appropriately present the prospects for the application of different machine learning models or artificial intelligence methods for land use, especially for conversion between multiple land forms.

** Response: Thank you for this suggestion. In the revised manuscript, we have thus reviewed and added more citations on land use projections (Lines 94-96 of the revised manuscript).

Almeida, C.d., Gleriani, J., Castejon, E.F., Soares‐Filho, B.S., 2008. Using neural networks and cellular automata for modelling intra‐urban land‐use dynamics. International Journal of Geographical Information Science, 22, 943-963.

Li, X., Yeh, A.G.-O., 2002. Neural-network-based cellular automata for simulating multiple land use changes using GIS. International Journal of Geographical Information Science, 16, 323-343.

White, R., Engelen, G., Uljee, I., 1997. The use of constrained cellular automata for high-resolution modelling of urban land-use dynamics. Environment and Planning B: Planning and Design, 24, 323-343.

Verburg, P.H., Soepboer, W., Veldkamp, A., Limpiada, R., Espaldon, V., Mastura, S.S., 2002. Modeling the spatial dynamics of regional land use: the CLUE-S model. Environmental management, 30, 391-405.

Sohl, T.L., Sayler, K.L., Drummond, M.A., Loveland, T.R., 2007. The FORE-SCE model: a practical approach for projecting land cover change using scenario-based modeling. Journal of Land Use Science, 2, 103-126.

Van Asselen, S., Verburg, P.H., 2013. Land cover change or land‐use intensification: simulating land system change with a global‐scale land change model. Global change biology, 19, 3648-3667.

Liu, X., Liang, X., Li, X., Xu, X., Ou, J., Chen, Y., Li, S., Wang, S., Pei, F., 2017. A future land use simulation model (FLUS) for simulating multiple land use scenarios by coupling human and natural effects. Landscape and Urban Planning, 168, 94-116.

3) Relevant maps, at least in the first one, need to add basic elements such as the compass and scale. It is convenient for international readers to understand the context of the study area.

** Response: We have added North Arrow and Scale Bar for first map (Figure 3 in the revised manuscript).

Reviewer 3 Report

This manuscript developed a method to predict development by 2040 under a business-as-usual scenario. The manuscript was organized well, and the results were presented clearly. Some minor comments are as follows:

1. There are many simulation models used for land simulation, such as some models using Celluar Automata, PLUS, CLUMondo etc…These research progress on land simulation should be reviewed in the introduction.

2.  Please provide more details to describe the business-as-usual scenario.

3.  Some figures need to be improved with higher resolution, such as Figure 1, Figure 2, Figure 3 and Figure 5.

4.  In the section 2.1, there are many equations, and the variables with abbreviations are explained one by one. Section 2.1 can be improved more compactly. 

5.  What is the impact from the different scales of different layer? I think the error and accuracy of the UHD growth from 2001 to 2016 can be discussed. Some paragraphs in Section 3.1 can be removed to the section of discussion.

6.  Figure 4 can be improved to help reader comprehensive the results easily.

Author Response

Reviewer 3

This manuscript developed a method to predict development by 2040 under a business-as-usual scenario. The manuscript was organized well, and the results were presented clearly. Some minor comments are as follows:

1. There are many simulation models used for land simulation, such as some models using Celluar Automata, PLUS, CLUMondo etc. These research progress on land simulation should be reviewed in the introduction.

** Response: We have reviewed and cited more land use simulation models in the introduction section of the revised manuscript (Lines 94-96 of the revised manuscript). Thank you for the suggestion.

Almeida, C.d., Gleriani, J., Castejon, E.F., Soares‐Filho, B.S., 2008. Using neural networks and cellular automata for modelling intra‐urban land‐use dynamics. International Journal of Geographical Information Science, 22, 943-963.

Li, X., Yeh, A.G.-O., 2002. Neural-network-based cellular automata for simulating multiple land use changes using GIS. International Journal of Geographical Information Science, 16, 323-343.

White, R., Engelen, G., Uljee, I., 1997. The use of constrained cellular automata for high-resolution modelling of urban land-use dynamics. Environment and Planning B: Planning and Design, 24, 323-343.

Verburg, P.H., Soepboer, W., Veldkamp, A., Limpiada, R., Espaldon, V., Mastura, S.S., 2002. Modeling the spatial dynamics of regional land use: the CLUE-S model. Environmental management, 30, 391-405.

Sohl, T.L., Sayler, K.L., Drummond, M.A., Loveland, T.R., 2007. The FORE-SCE model: a practical approach for projecting land cover change using scenario-based modeling. Journal of Land Use Science, 2, 103-126.

Van Asselen, S., Verburg, P.H., 2013. Land cover change or land‐use intensification: simulating land system change with a global‐scale land change model. Global change biology, 19, 3648-3667.

Liu, X., Liang, X., Li, X., Xu, X., Ou, J., Chen, Y., Li, S., Wang, S., Pei, F., 2017. A future land use simulation model (FLUS) for simulating multiple land use scenarios by coupling human and natural effects. Landscape and Urban Planning, 168, 94-116.

2. Please provide more details to describe the business-as-usual scenario.

** Response: The Business-as-Usual scenario assumed that urban development trends from 2016 to 2040 remains the same as that between 2001 and 2016. We added the description of the scenario in the revised manuscript (Lines 151-153 of the revised manuscript).

3. Some figures need to be improved with higher resolution, such as Figure 1, Figure 2, Figure 3 and Figure 5.

** Response: The figures in our original manuscript (word document) had a resolution of 300 dpi. It might be the review system that converted the word document to PDF, which greatly reduced the quality of the figures. For your reference, we added Figures 1-3 and 5 here.

Figure 1. Framework of our modeling methods. (1) projecting urban demands, (2) calculating development suitability, (3) creating urban development probability, and (4) generating binary UHD and LDR projections. UHD: urban and highly developed; LDR: low-density residential. Datasets used across modeling steps are highlighted under dotted frames.

Figure 2. State-level relationship between population growth and urban area increases from 2001-2016 (a: urban and highly developed; b: Low-density residential).

Figure 3. Overview of urban and highly developed UHD suitability.

Figure 5. Distribution of 10 selected cities/metropolitans for accuracy assessment.

4. In the section 2.1, there are many equations, and the variables with abbreviations are explained one by one. Section 2.1 can be improved more compactly.

** Response: Thanks for the suggestion. In Section 2.1, we aimed to explain how we calculate urban land demands. Although it might be long, we think our current Section 2.1 is easy to follow with models and rationales clearly explained. Thus, we decided to keep the original content in Section 2.1. However, we have thoroughly checked and revised the language to make the section easier to follow.

5. What is the impact from the different scales of different layer? I think the error and accuracy of the UHD growth from 2001 to 2016 can be discussed. Some paragraphs in Section 3.1 can be removed to the section of discussion.

** Response: Good suggestion. However, the main focus of our work is to evaluate farmland under threat and implications, so we prefer to keep projection evaluation in Section 3.1 and implication discussion in Section 4. Given the fact that accuracy of land use simulation can be impacted by many factors (data availability, model design, and inherent uncertainty in land use simulation), it is not easy to evaluate the impact of each input variable. Instead, like other studies (Liu et al., 2017; Sohl et al., 2014), we evaluated the overall performance of our model.

Sohl, T.L., Sayler, K.L., Bouchard, M.A., Reker, R.R., Friesz, A.M., Bennett, S.L., Sleeter, B.M., Sleeter, R.R., Wilson, T., Soulard, C., 2014. Spatially explicit modeling of 1992–2100 land cover and forest stand age for the conterminous United States. Ecological Applications, 24, 1015-1036.

Liu, X., Liang, X., Li, X., Xu, X., Ou, J., Chen, Y., Li, S., Wang, S., Pei, F., 2017. A future land use simulation model (FLUS) for simulating multiple land use scenarios by coupling human and natural effects. Landscape and Urban Planning, 168, 94-116.

6. Figure 4 can be improved to help reader comprehensive the results easily.

** Response: Thanks for the suggestion. Figure 4 is a demonstration of residential forestlands that can remain undeveloped. In the revised manuscript, we further revised the figure caption to clarify the implications of this figure. Please check the revised caption of Figure 4 (Lines 263-265).

Reviewer 4 Report

The article is suitable for publication. In particular, the Theoretical Approach, Methodology and Data Sources section was sufficiently supplemented.

Author Response

Reviewer 4

The article is suitable for publication. In particular, the Theoretical Approach, Methodology and Data Sources section was sufficiently supplemented.

** Response: Thank you for the positive feedback.

Round 2

Reviewer 1 Report

The authors investigated my comments.