Land-Use and Land-Cover Changes in Dong Trieu District, Vietnam, during Past Two Decades and Their Driving Forces
Round 1
Reviewer 1 Report
The manuscript is based on a solid research, it applies relevant data and methods. It is concisely and clearly written.
Still, I believe, authors should elaborate few minor issues:
2.1 I suggest including a map showing the location of the study area within Vietnam. I also suggest elaborating further on socioeconomic situation in the study area with respect to industrial activities as coal mining and raw materials quarries - they are not mentioned in the results. Does it mean they are not detected in the satellite images?
2.3 I suggest to explain in more detail class "Barren land", in particular when we see from the results that it was changed into productive land in 2010 and 2019 - what exactly is hidden within this class and why it was "barren" in 2000?
2.5 Driving forces are natural or socioeconomic phenomena causing land use change. Authors analysed merely spatial factors as indicators of those forces - I suggest distinguishing this difference in the text.
In addition, what was the reason behind dividing slope into not uniform classes (less than 2, 2-6, 6-15, 15-26°)? Similarly with population classes. Please explain for the reader.
Author Response
The manuscript is based on a solid research, it applies relevant data and methods. It is concisely and clearly written.
Still, I believe, authors should elaborate few minor issues:
2.1 I suggest including a map showing the location of the study area within Vietnam. I also suggest elaborating further on socioeconomic situation in the study area with respect to industrial activities as coal mining and raw materials quarries - they are not mentioned in the results. Does it mean they are not detected in the satellite images?
Reply:
A map showing the location of the study area within Vietnam is included as suggested.
The area for coal mining and raw material quarries were included in the area of barren land under the Level I LULC classification used.
2.3 I suggest to explain in more detail class "Barren land", in particular when we see from the results that it was changed into productive land in 2010 and 2019 - what exactly is hidden within this class and why it was "barren" in 2000?
Reply:
The definition for barren land is revised in Table 2.
Most of the barren lands in 2000 were converted to forest and orchards due to the orchards orientation policy and forest policy implemented during the study period.
2.5 Driving forces are natural or socioeconomic phenomena causing land use change. Authors analysed merely spatial factors as indicators of those forces - I suggest distinguishing this difference in the text.
Reply:
Thanks for the suggestion. A distinguish is made in the text.
In addition, what was the reason behind dividing slope into not uniform classes (less than 2, 2-6, 6-15, 15-26°)? Similarly with population classes. Please explain for the reader.
Reply:
The classes for slope were divided in accordance with the general guidelines in delineating geographical categories at the district level in Vietnam [28;29]. Typos on two slope classes were also corrected. The population classes were divided based on the Vietnam population density classification scheme [30].
Reviewer 2 Report
This article, seems to me, be correctly arranged and clearly shows the applied method of collecting data on land use changes. In the chapter conclusion, can also be mentioned something about the quality of the applied method.
Author Response
This article, seems to me, be correctly arranged and clearly shows the applied method of collecting data on land use changes. In the chapter conclusion, can also be mentioned something about the quality of the applied method.
Reply:
Thanks for the strong support.
Quality and implications of the applied method are added in the conclusion.
Reviewer 3 Report
The authors present a classic work exploiting Land use and land cover change analysis to modelling the change patterns. Getting to an understanding of actual structural changes of a given area can, of course, be a powerful planning tool with many positive insights on the policy and governance of a specific territory. Let alone when one can shed some light, making inferences on solid findings on the active driving factors. Sadly the presented article is none of that.
Here is a list of the two major limitations the authors should consider:
- Despite the experiment design and methods are adequately described and the results presented, the overall work in its whole outfit looks more like a dry list of mere observations. Most of the supposed key findings can be said just by looking at the maps.
- On line 50, the authors assert as "awareness of the driving components of the LUC is pivotal for land use planning". Which is also the most intriguing and modern aspect of groundbreaking works in land survey and policy. However, no mentions of such potential positive outcomes for the reported Dong Trieu district are given in the article, not even a speculative attempt in the conclusive chapter. All the game connecting changes causes and effects the main driving forces is reduced to a list of correlation indexes.
- What are the limits inherent a 30m pixel map to describe structural complex changes? What implications for a multiscale approach using different data granularity? Why maximum likelihood and not other classification algorithms? What consideration led the authors to use the CVC and not others?
Do you really need to set up a Landsat data classification and correlation statistics to say that "the proximity from urban area influence the reduction of the forest and cropland .. the expansion of built-up areas because the new buildings are often likely to appear near existing buildings." or suggest that "the water area ... often located far from the forest." and "The cropland .. often occurred near the water ... due to easy water accessibility.". really?
Minors,
- check the fonts e.g. line 173 and 174
- at the beginning of section 2.1 a clear statement to the actual territorial extension of the studied district and a map would improve readers understanding of the place.
that's all
Author Response
The authors present a classic work exploiting Land use and land cover change analysis to modelling the change patterns. Getting to an understanding of actual structural changes of a given area can, of course, be a powerful planning tool with many positive insights on the policy and governance of a specific territory. Let alone when one can shed some light, making inferences on solid findings on the active driving factors. Sadly the presented article is none of that.
Reply:
The mentioned applications are parts of another paper to be submitted soon. We purposely omitted this part to avoid violation of ethical issues.
Here is a list of the two major limitations the authors should consider:
- Despite the experiment design and methods are adequately described and the results presented, the overall work in its whole outfit looks more like a dry list of mere observations. Most of the supposed key findings can be said just by looking at the maps. On line 50, the authors assert as "awareness of the driving components of the LUC is pivotal for land use planning". Which is also the most intriguing and modern aspect of groundbreaking works in land survey and policy. However, no mentions of such potential positive outcomes for the reported Dong Trieu district are given in the article, not even a speculative attempt in the conclusive chapter. All the game connecting changes causes and effects the main driving forces is reduced to a list of correlation indexes.
Reply:
Thanks for the advice. Simplified inferences (same reason as above) on the policy and governance for Dong Trieu district based on findings on the active driving factors are given in the conclusion.
- What are the limits inherent a 30m pixel map to describe structural complex changes? What implications for a multiscale approach using different data granularity? Why maximum likelihood and not other classification algorithms? What consideration led the authors to use the CVC and not others?
Reply:
Methodologically, the analyses of 30m pixel optical satellite images are limited by the similarities in spectral reflectance across landscapes. For example, different crops or tree species with similar phenological characteristics may be indistinguishable from the image, which can result in similar land cover features. Ancillary data (such as DEM and population density) that are varying in format, accuracy, and spatial resolution than satellite data provided more insights on the potential driving factors.
A preliminary study by the first author indicated that the maximum likelihood classification scheme serves our purpose well in classifying the land use types in the studied region.
There are several reasons to use CVC but not others.
- Comparing to other measures for measuring association like Phi or Pearson's Contingency Coefficient (C), CVC can compare multiple ê“2 test statistics and is generalizable across contingency tables of varying sizes.
- CVC is not affected by sample size and therefore is very useful in situations where a statistically significant chi-square was the result of a large sample size instead of any substantive relationship between the variables.
- CVC can be interpreted as a measure of the relative (strength) of an association between two variables.
Do you really need to set up a Landsat data classification and correlation statistics to say that "the proximity from urban area influence the reduction of the forest and cropland. the expansion of built-up areas because the new buildings are often likely to appear near existing buildings." or suggest that "the water area ... often located far from the forest." and "The cropland .. often occurred near the water ... due to easy water accessibility.". really?
Reply:
What we are trying to do is to discover the priority and weighting of each driving factor on the effects of changes between different land classes. Although it may seem obvious in some cases, it is a necessary step to determine the logistic regressions to be applied in a land change model used to simulate land use changes in future scenarios.
Minors,
check the fonts e.g. line 173 and 174
Reply:
Corrected as suggested.
at the beginning of section 2.1 a clear statement to the actual territorial extension of the studied district and a map would improve readers understanding of the place.
Reply:
A map is included as suggested.
Reviewer 4 Report
It is recommended to use the same line spacing in all tables. For example, the table 1 has line spacing one while the table 4 has bigger line spacing.
I recommend to arrange the text so that the tables will not divided between separate pages.
The figures in tables 6 (A. B and C) does not have units. The figures are obviously in hectares but however, it should be indicated in the title of the tables. The second column (first column of figures) does not have the column names in those tables.
The figures appear in the tables in different style. The thousands separator (,) is used randomly. Such slip ups must be corrected.
Author Response
It is recommended to use the same line spacing in all tables. For example, the table 1 has line spacing one while the table 4 has bigger line spacing.
Reply:
Corrected as suggested.
I recommend to arrange the text so that the tables will not divided between separate pages.
Reply:
Corrected as suggested.
The figures in tables 6 (A. B and C) does not have units. The figures are obviously in hectares but however, it should be indicated in the title of the tables. The second column (first column of figures) does not have the column names in those tables.
Reply:
Corrected as suggested.
The figures appear in the tables in different style. The thousands separator (,) is used randomly. Such slip ups must be corrected.
Reply:
Corrected as suggested.
Reviewer 5 Report
The authors present an interesting topic, However, there are still some issues that need to be addressed before considering this work for publication. - - The methodology section is poor and not well documented. I suggest to add a flowchart presenting a methodology used for this study , -- The discussion of your results need to be improved. - The quality of the output maps is not good ( see figure 1)Author Response
The authors present an interesting topic, However, there are still some issues that need to be addressed before considering this work for publication.
The methodology section is poor and not well documented. I suggest to add a flowchart presenting a methodology used for this study.
Reply:
A flow chart is included.
The discussion of your results need to be improved.
Reply:
The discussion has been improved as suggested.
The quality of the output maps is not good (see figure 1)
Reply:
The quality of the LULC maps is improved.
Round 2
Reviewer 3 Report
thanks to the authors for considering my comments and for the changes made to the previous version. I have to say I am partially satisfied. In my previous report, I had risen two major flaws. On one hand, the article was missing a dissertation on the positive outcomes for the district, limiting the discussion to a dry list of correlation indexes. The second was the limits of using 30m pixel maps to describe the changes.
Whereas the first one looks still not answered, to me. I have to say that the authors did a good job with the second.
In last words, my advice is that a little effort more should be done to fix it. Omitting parts because one has another paper in the pipeline does not make the presented paper better.
that's all
Author Response
thanks to the authors for considering my comments and for the changes made to the previous version. I have to say I am partially satisfied. In my previous report, I had risen two major flaws. On one hand, the article was missing a dissertation on the positive outcomes for the district, limiting the discussion to a dry list of correlation indexes. The second was the limits of using 30m pixel maps to describe the changes.
Whereas the first one looks still not answered, to me. I have to say that the authors did a good job with the second.
Reply:
Thanks for the positive support.
In last words, my advice is that a little effort more should be done to fix it. Omitting parts because one has another paper in the pipeline does not make the presented paper better.
Reply:
We tried our best in the short revising period given, even consulting a UC Berkeley graduate, a native English speaker, and majored in Environmental Sciences. Unfortunately, we are still unable to make the section “less dry” without mentioning the data to be submitted in our second paper.
