Spatiotemporal Study of Land Degradation Impacting the Oldest Mountains of the Indian Subcontinent
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsGeneral comments
The study utilizes an integrated approach combining Land Use and Land Cover (LULC) transition analysis with soil-erosion modeling using the Revised Universal Soil Loss Equation (RUSLE). The paper is well-structured, moving logically from the global problem of land degradation to specific regional findings and future policy.
Specific comments
-
The methodology is robust in its use of multi-temporal datasets, including MODIS (2001–2021) for long-term trends and high-resolution ESRI 10m data (2017–2024) for short-term dynamics. Please write if the application of the RUSLE model at a 30m spatial resolution across the entire Aravalli Mountain System (AMS) scale is a significant contribution, as such high-resolution measurements were previously lacking.
-
The research correctly identifies and parameterizes the six RUSLE factors: Rainfall Erosivity (R), Soil Erodibility (K), Slope Length and Steepness (LS), Cover Management (C), and Support Practice (P). There is no information, how to the use of Sentinel-2 derived NDVI to calculate the C-factor provides a dynamic and accurate reflection of vegetation health.
-
While the model is scientifically sound, the reliance on Inverse Distance Weighting (IDW) for rainfall interpolation at 30m resolution from 4km source data may introduce localized smoothing effects in areas with high topographic variability.
- I think that the study a little successfully correlates erosion hotspots with steep slopes, susceptible soil types, and active mining areas. You must also highlights a critical policy paradox: increased soil erosion persists despite afforestation efforts, suggesting that localized conservation cannot offset massive regional land conversion.
-
Some sections regarding the Supreme Court rulings are highly detailed and may benefit from a more concise summary to maintain the focus on the scientific modeling. Including a sensitivity analysis of the RUSLE parameters would further bolster the results, showing which factors most significantly drive the 13.8% increase in soil.
Contructive feedback
The study places the AMS within a global context, comparing its degradation pathways to other ancient systems and highlands. This strengthens the paper's argument that the interaction between LULC dynamics and soil erosion is a universal driver of degradation in old mountain systems. But, I suggest to address this issue: Comparing results from two datasets with such vastly different spatial resolutions (500m vs. 10m) introduces significant "scaling noise." An increase in built-up area might partially be an artifact of the ESRI dataset’s ability to detect small structures (rural housing, narrow roads) that MODIS simply cannot see. The paper needs to better account for this "resolution bias" in its trend analysis.
This paper is a highly valuable contribution to the field of geomorphology and sustainable land management. It provides a scalable, high-resolution framework that can be adapted for other vulnerable mountain systems globally. The methodology is technically sound, and the results are significant for future mineral governance and conservation planning.
Summary:
The work is scientifically valid as it represents the first time the RUSLE model has been applied at a 30m resolution across the entire AMS scale.
Author Response
Reviewer 1
General comments
The study utilizes an integrated approach combining Land Use and Land Cover (LULC) transition analysis with soil-erosion modeling using the Revised Universal Soil Loss Equation (RUSLE). The paper is well-structured, moving logically from the global problem of land degradation to specific regional findings and future policy.
Specific comments
Comment 1: The methodology is robust in its use of multi-temporal datasets, including MODIS (2001–2021) for long-term trends and high-resolution ESRI 10m data (2017–2024) for short-term dynamics. Please write if the application of the RUSLE model at a 30m spatial resolution across the entire Aravalli Mountain System (AMS) scale is a significant contribution, as such high-resolution measurements were previously lacking.
Response: We thank the reviewer for their review and positive suggestions. The Aravalli Mountain System (AMS) is not as high as the Himalaya, but it is unique for its heterogeneous geomorphology, including steep slopes, ridges, and valleys. The 30m spatial resolution study across the entire Aravalli Mountain System (AMS) scale is a significantly identify variation in slope, vegetation and Anthropogenic activities. A high-resolution study will assess an accurate, spatially distributed soil erosion hotspot that can be used by policymakers to develop control measures related to policy. We have included the given suggestion in the manuscript.
Comment 2: The research correctly identifies and parameterizes the six RUSLE factors: Rainfall Erosivity (R), Soil Erodibility (K), Slope Length and Steepness (LS), Cover Management (C), and Support Practice (P). There is no information, how to the use of Sentinel-2 derived NDVI to calculate the C-factor provides a dynamic and accurate reflection of vegetation health.
Response: We thank the reviewer for the suggestion; we have incorporated the clarification into the methodology.
Comment 3: While the model is scientifically sound, the reliance on Inverse Distance Weighting (IDW) for rainfall interpolation at 30m resolution from 4km source data may introduce localized smoothing effects in areas with high topographic variability.
Response: We thank the reviewer for the suggestion; we have incorporated the clarification into the methodology. We also tried to smooth the data (Though the Aravalli Mountain System is characterised by complex topography and spatial distribution may be affected by orographic effects, but due to the non-availability of high-resolution data or regional-specific data, we have used 4km spatial resolution data. For converting this 4km spatial resolution rainfall data, we have used on Inverse Distance Weighting (IDW) for rainfall interpolation, which provides a computationally efficient approach that ensures compatibility with other RUSLE factors)
Comment 4: I think that the study a little successfully correlates erosion hotspots with steep slopes, susceptible soil types, and active mining areas. You must also highlights a critical policy paradox: increased soil erosion persists despite afforestation efforts, suggesting that localized conservation cannot offset massive regional land conversion.
Answer: We have revised the Discussion section to highlight the policy paradox that, despite afforestation, increased soil erosion persists due to concurrent large-scale land-use changes such as urbanisation and mining. The need to integrate soil-erosion dynamics into future conservation policies has now been addressed in the revised manuscript (L:500-514).
Comment 5: Some sections regarding the Supreme Court rulings are highly detailed and may benefit from a more concise summary to maintain the focus on the scientific modeling. Including a sensitivity analysis of the RUSLE parameters would further bolster the results, showing which factors most significantly drive the 13.8% increase in soil.
Answer: We thank the reviewer for his suggestions. We have carried out a sensitivity analysis of the RUSLE parameters, and the following paragraph and figure have been added to the result and supplementary materials.
Comment 6: Constructive feedback
The study places the AMS within a global context, comparing its degradation pathways to other ancient systems and highlands. This strengthens the paper's argument that the interaction between LULC dynamics and soil erosion is a universal driver of degradation in old mountain systems. But, I suggest to address this issue: Comparing results from two datasets with such vastly different spatial resolutions (500m vs. 10m) introduces significant "scaling noise." An increase in built-up area might partially be an artifact of the ESRI dataset’s ability to detect small structures (rural housing, narrow roads) that MODIS simply cannot see. The paper needs to better account for this "resolution bias" in its trend analysis.
Answer: We appreciate the reviewer's detailed, constructive comments. We have replaced the table in the text and also mentioned the limitation in the methodology. We also addressed this scale-based bias in the results.
|
Class |
2001 |
2010 |
2020 |
Change detection (2020-2001) |
P factor |
C factor |
|
Forest |
30.6164 |
37.316 |
264.6212 |
234.0048 |
0.8 |
0.008 |
|
Shrubland |
3907.87 |
970.791 |
996.683 |
-2911.187 |
0.8 |
0.1 |
|
Permanent Wetlands |
3.42189 |
16.0132 |
61.8002 |
58.37831 |
1 |
0.001 |
|
Croplands |
70719.7 |
73611.1 |
73227.7 |
2508 |
0.5 |
0.08 |
|
Urban and Built-up |
1107.31 |
1152.48 |
1256.3 |
148.99 |
1 |
0.1 |
|
Barren |
192.432 |
153.947 |
116.098 |
-76.334 |
1 |
0.45 |
|
Water Bodies |
8.98976 |
28.6928 |
47.1376 |
38.14784 |
0 |
0 |
|
total |
75970.34 |
75970.34 |
75970.34 |
|
|
To compare the rate of erosion using different LULC, we have also calculated the rate of erosion using MODIS LULC. MODIS LULC is used to calculate the C and P factor using the values assigned for each LULC class (Table 1). R factor is calculated using the rainfall of 2000, 2010 and 2020. Only the C, P, and R factors have been changed; otherwise, the L, S, and K factors are constant. The results indicate that the erosion rate in 2001 was 0.17 t/ha/yr, 0.165 t/ha/yr in 2010, and 0.179 t/ha/yr in 2020. This shows a comparison of LULC spatial resolutions used to calculate erosion rates; higher spatial resolution captures more features and yields a more accurate spatial distribution of erosion.
|
Rate of erosion using LULC from MODIS-2001 |
Rate of erosion using LULC from MODIS-2010 |
Rate of erosion using LULC from MODIS-2020 |
Rate of erosion using ESRI LULC 2017 |
Rate of erosion using ESRI LULC 2024 |
|
0.17 t/ha/yr |
0.165 t/ha/yr |
0.179 t/ha/yr |
1.59 t/ha/yr |
1.81 t/ha/yr |
|
|
|
|
|
|
Comment 7: This paper is a highly valuable contribution to the field of geomorphology and sustainable land management. It provides a scalable, high-resolution framework that can be adapted for other vulnerable mountain systems globally. The methodology is technically sound, and the results are significant for future mineral governance and conservation planning.
Answer: We are thankful to the reviewer for his positive comments.
Summary:
The work is scientifically valid as it represents the first time the RUSLE model has been applied at a 30m resolution across the entire AMS scale.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Comments and Suggestions for AuthorsComment 1. In my opinion, the term "Chronosequence" in the title of the manuscript is not entirely correct. "Multitemporal" or "Spationtemporal" would be more appropriate. I advise the authors to change the title to better reflect the content of their work.
Comment 2. (L179–180) The reasoning behind the authors' choice of 2017 and 2024 should be explained in more detail. They claim that these two years were chosen because they "represent current and divergent meteorological conditions", but this statement is not supported by facts. They are advised to supplement this statement with figures or a graph showing the dynamics of meteorological conditions during these periods for comparison.
Comment 3. The data sources are not described sufficiently. For example, it is unclear from the text which MODIS LULC product and which of the five LULC classification schemes were used. The authors are advised to supplement this section with a table describing the data used.
Comment 4. (L256-258) The authors state that "each LULC class (2017 and 2024) is based on published empirical data (e.g. forest = 0.8, farmland = 0.5, built-up = 1)", but they do not specify the sources. The data coefficients also appear unusual. The authors are advised to provide a table showing the P values for each LULC class, along with specific references to the sources used and justification for each value selected.
Author Response
Reviewer 2
Comments and Suggestions for Authors
Comment 1. In my opinion, the term "Chronosequence" in the title of the manuscript is not entirely correct. "Multitemporal" or "Spationtemporal" would be more appropriate. I advise the authors to change the title to better reflect the content of their work.
Answer: We appreciate the reviewer's suggestion! We have changed the title to “Spatiotemporal study of land degradation impacting the oldest mountains of the Indian subcontinent”
Comment 2. (L179–180) The reasoning behind the authors' choice of 2017 and 2024 should be explained in more detail. They claim that these two years were chosen because they "represent current and divergent meteorological conditions", but this statement is not supported by facts. They are advised to supplement this statement with figures or a graph showing the dynamics of meteorological conditions during these periods for comparison.
Answer: We appreciate the comments by the reviewer! We chose 2017 and 2024 due to the availability of the high-resolution data and “represent current and divergent meteorological conditions". We have supported it because of its importance to the soil erosion model. We have discussed this in 2.2. Data and its Analysis.
Comment 3. The data sources are not described sufficiently. For example, it is unclear from the text which MODIS LULC product and which of the five LULC classification schemes were used. The authors are advised to supplement this section with a table describing the data used.
Answer: We appreciate the comments by the reviewer; we have now highlighted this in section 2.2. Data and Its Analysis. For more information, we have also added more information in the supplementary section.
MODIS use five different classification scheme
- International Geosphere- Biosphere Programme scheme
- University of Maryland Scheme
- FPAR/LAI Biome classification
- Biogeochemical cycles classification
- Plant Functional Types scheme
|
S. No |
Data Name |
Source |
|
1. |
MODIS LULC |
MODIS/006/MCD12Q1 |
|
2. |
ESRI LULC |
Sentinel-2 10m Land Use/Land Cover Download |
|
3. |
Digital Elevation Model |
ALOS PALSAR |
|
4. |
Rainfall data |
CHRS Data Portal |
|
5. |
NDVI |
Sentinel 2 |
|
6. |
Soil texture maps |
National Bureau of Soil Survey and Land Use Planning (NBSS&LUP) |
Comment 4. (L256-258) The authors state that "each LULC class (2017 and 2024) is based on published empirical data (e.g. forest = 0.8, farmland = 0.5, built-up = 1)", but they do not specify the sources. The data coefficients also appear unusual. The authors are advised to provide a table showing the P values for each LULC class, along with specific references to the sources used and justification for each value selected.
Answer: We appreciate the reviewer's comments. Though we have mentioned all datasets in Section 2.2. Data and Its Analysis. In the Data Availability Statement section, we have provided all links to the data sources. In Table 2, mentioned below, we have already given all the references for the justification of the values.
|
S.no |
LULC class |
2017 |
2024 |
Change detection |
P factor |
Reference |
|
1 |
Waterbodies |
400.515 |
440.956 |
40.44104 |
0 |
Jain et al. (2001) |
|
2 |
Forest |
1231.85 |
1379.16 |
147.31 |
0.8 |
Pandey et al. (2007) |
|
3 |
Flooded vegetation |
5.55089 |
7.04385 |
1.49296 |
0.05 |
Brema, J., & Hauzinger, J. (2016 |
|
4 |
Cropland |
44736.3 |
43353.8 |
-1382.5 |
0.5 |
Wanielista and Yousef (1992) |
|
5 |
Builtup |
4958.84 |
7603.16 |
2644.32 |
1 |
Wanielista and Yousef (1992) |
|
6 |
Bareland |
264.784 |
163.02 |
-101.764 |
1 |
Wanielista and Yousef (1992) |
|
7 |
Rangeland |
24372.5 |
23023.2 |
-1349.3 |
0.8 |
Jain et al. (2010) |
|
Total |
75970.34 |
75970.34 |
|
|
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThank you for your response to my comments. I appreciate the clarifications and the revisions you have made. I am satisfied with your explanations and agree with the updates provided.
Author Response
NA
Reviewer 2 Report
Comments and Suggestions for AuthorsThe manuscript has significantly improved from the last revision and can be published in current form
Author Response
NA
