the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 30 Jul 2024
Enhancing High-Resolution Forest Stand Mean Height Mapping in China through an Individual Tree-Based Approach with Close-Range LiDAR Data
Abstract. Forest stand mean height is a critical indicator in forestry, playing a pivotal role in various aspects such as forest inventory estimation, sustainable forest management practices, climate change mitigation strategies, monitoring of forest structure changes, and wildlife habitat assessment. However, there is currently a lack of large-scale, spatially continuous forest stand mean height maps. This is primarily due to the requirement of accurate measurement of individual tree height in each forest plot, a task that cannot be effectively achieved by existing globally covered, discrete footprint-based satellite platforms. To address this gap, this study was conducted using over 1117 km2 of close-range Light Detection and Ranging (LiDAR) data, which enables the measurement of individual tree height in forest plots with high precision. Besides, this study incorporated spatially continuous climatic, edaphic, topographic, vegetative, and Synthetic Aperture Radar data as explanatory variables to map the tree-based arithmetic mean height (ha) and weighted mean height (hw) at 30 m resolution across China. Due to limitations in obtaining basal area of individual tree within plots using UAV LiDAR data, this study calculated weighted mean height through weighting an individual tree height by the square of its height. In addition, to overcome the potential influence of different vegetation divisions at large spatial scale, we also developed a machine learning-based mixed-effects model to map forest stand mean height across China. The results showed that the average ha and hw across China were 11.3 m and 13.3 m with standard deviations of 2.9 m and 3.3 m, respectively. The accuracy of mapped products was validated utilizing LiDAR and field measurement data. The correlation coefficient (𝑟) for ha and hw ranged from 0.603 to 0.906 and 0.634 to 0.889, while RMSE ranged from 2.6 to 4.1 m and 2.9 to 4.3 m, respectively. Comparing with existing forest canopy height maps derived using the area-based approach, it was found that our products of ha and hw performed better and aligned more closely with the natural definition of tree height. The methods and maps presented in this study provide a solid foundation for estimating carbon storage, monitoring changes in forest structure, managing forest inventory, and assessing wildlife habitat availability. The dataset constructed for this study is publicly available at https://doi.org/10.5281/zenodo.12697784 (Chen et al., 2024).
- Preprint
(2864 KB) - Metadata XML
-
Supplement
(265 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on essd-2024-274', Anonymous Referee #1, 31 Aug 2024
Forest stand mean height is a critical indicator in forestry, playing a pivotal role in various aspects such as forest inventory, sustainable forest management practices, climate change mitigation strategies, monitoring of forest structure changes, and wildlife habitat assessment. However, mapping the national scale forest stand mean height across China is still a challenge, recently. In this manuscript, a tree-based approach to create spatially continuous forest stand mean height maps of China was developed by integrating high-point density, high-precision close-range LiDAR data and multisource remote sensing data. The methods employed for data acquisition, processing, and analysis are methodical and rational. The accuracy of the estimated results is commendable, thereby demonstrating significant potential for practical applications. However, some issues should be further corrected or clarified before publication.
in the part of Abstract, ‘Forest stands mean height is a critical indicator in forestry, playing a pivotal role in various aspects such as forest inventory estimation,’ forest inventory estimation is suggested to be modified to forest inventory with various scales, which is more reasonable.
In the line of 69: The height metrics from obtained from this approach is forest canopy height, which include not only the actual tree height. There is one mistake in the expression. The sentence should be corrected: The height metrics obtained from this approach is forest canopy height.
In terms of data, various types of data collected over a span of 6 years are included in this manuscript, such as ground measured samples, LiDAR data obtained from different sensors, and remote sensing images. How can these datasets be matched on a temporal scale? Additionally, how can reduce the limitations of images acquired in different years and seasons?
The formula of determining coefficients (formula 11), y ̅_iis not the mean value for the observed values. y ̅ is recommended. In the formula 16, the means of y ̅ also should be expressed.
In the manuscript, three accuracy indices were employed to evaluate the performance of models. However, when evaluating results with the same RMSE in various height forests, it is recommended to include rRMSE.
In Figure 3, it is evident that an overestimation of forest stand height occurs when the weighted average of tree height squared is applied for forest stands taller than 14 meters. Please provide the underlying reasons.
The decimal places of precision indexes in this paper should be consistent, such as Tabel 5.Citation: https://doi.org/10.5194/essd-2024-274-RC1 -
AC1: 'Reply on RC1', Haitao Yang, 05 Sep 2024
Dear Editor and Reviewer:
On behalf of my co-authors, we thank you very much for giving us an opportunity to revise our manuscript, and we also appreciate reviewers very much for their positive and constructive comments and suggestions on our manuscript entitled “Enhancing High-Resolution Forest Stand Mean Height Mapping in China through an Individual Tree-Based Approach with Close-Range LiDAR Data” (Manuscript Number: essd-2024-274).
We revised the manuscript according to these comments and suggestions. All changes were marked in highlight text in the revised manuscript. The line numbers in the response are the corresponding line numbers in the revised version.
Once again, thank you very much for your comments and suggestions.
Comment 1:
In the part of Abstract, ‘Forest stands mean height is a critical indicator in forestry, playing a pivotal role in various aspects such as forest inventory estimation,’ forest inventory estimation is suggested to be modified to forest inventory with various scales, which is more reasonable.
Reply 1: Thank you very much for your professional advice, we have changed ‘forest inventory estimation’ to ‘forest inventory’ at Line 21-22.
Comment 2:
In the line of 69: The height metrics from obtained from this approach is forest canopy height, which include not only the actual tree height. There is one mistake in the expression. The sentence should be corrected: The height metrics obtained from this approach is forest canopy height.
Reply 2: The mistake has been corrected according to your kind advices and detailed suggestions. Please refer to Line 69-70 for details.
Comment 3:
In terms of data, various types of data collected over a span of 6 years are included in this manuscript, such as ground measured samples, LiDAR data obtained from different sensors, and remote sensing images. How can these datasets be matched on a temporal scale? Additionally, how can reduce the limitations of images acquired in different years and seasons?
Reply 3：Changes in forest resources tend to occur relatively slowly, and a 5-year period is a sufficiently long-time span to capture significant change trends. The temporal scale for China's national-level forest resource inventory is set at 5 years, aiming to balance the need for real-time data with long-term trend observation. This time span is long enough to detect significant changes in forest ecosystems, yet short enough to ensure that policies and management measures can be promptly adjusted based on the most recent data.
As of 2015, the application of LiDAR has not been widely adopted in forest remote sensing research in China. Considering the cost and the difficulty of data collection, it was challenging to collect extensive, high-point density and accurate data across China within a short timeframe. Considering the nationwide data coverage, the final dataset for this study spans 6 years (one year longer than the time span of the national inventory). This represents a limitation of the data used in this study, which is discussed in the paper. Please refer to Line 485-487 for details.
Comment 4:
The formula of determining coefficients (formula 11), y ̅_iis not the mean value for the observed values. y ̅ is recommended. In the formula 16, the means of y ̅ also should be expressed.
Reply 4: We have corrected the formulas. Please refer to equations 11-20 for details.
Comment 5:
In the manuscript, three accuracy indices were employed to evaluate the performance of models. However, when evaluating results with the same RMSE in various height forests, it is recommended to include rRMSE.
Reply 5: We agreed with the reviewer's comment and added the rRMSE to the Table5, which evaluating results with the same RMSE in various height forests. Please refer to Table 5 for details.
Comment 6:
In Figure 3, it is evident that an overestimation of forest stand height occurs when the weighted average of tree height squared is applied for forest stands taller than 14 meters. Please provide the underlying reasons.
Reply 6:
We greatly appreciate the reviewer’s insightful question. In response, we have explored the issue from both theoretical and empirical perspectives to provide a comprehensive answer. Please refer to Line 440-445 for details.
For a detailed mathematical and empirical proof, please refer to the supplementary materials.
Comment 7:
The decimal places of precision indexes in this paper should be consistent, such as Tabel 5.
Reply 7: We have adjusted to ensure the consistency of decimal places for the indexes. Please refer to Table 5 for details.
-
AC2: 'Reply on RC2', Haitao Yang, 05 Sep 2024
Dear Editor and Reviewer:
On behalf of my co-authors, we thank you very much for giving us an opportunity to revise our manuscript, and we also appreciate reviewers very much for their positive and constructive comments and suggestions on our manuscript entitled “Enhancing High-Resolution Forest Stand Mean Height Mapping in China through an Individual Tree-Based Approach with Close-Range LiDAR Data” (Manuscript Number: essd-2024-274).
We revised the manuscript according to these comments and suggestions. All changes were marked in highlight text in the revised manuscript. The line numbers in the response are the corresponding line numbers in the revised version.
Once again, thank you very much for your comments and suggestions.
Comment 1: Line 22: deleted estimation.
Reply 1: Thanks to reviewer for reminder, the estimation has been deleted in the Abstract. Please refer to Line 21-22 for details.
Comment 2: Line 45: The author used arithmetic mean height (ℎ𝑎) and weighted mean height (ℎ𝑤) to represent Forest Stand Mean Height. The similarities and differences between these two metrics should be explained at the beginning of the Introduction.
Reply 2: In the introduction, we described the differences in calculation methods and the similarities in application directions. The detailed similarities and differences were explained in the formula section and discussed in the discussion section.
The differences:
Forest stand height denotes the mean height of trees within a stand/plot, including arithmetic mean height and mean height weighted in proportion to their basal area (weighted mean height or Lorey’s mean height) (Laar and Akça 2007; Masaka et al. 2013). Please refer to Line 45-47 for details.
The similarities:
It serves as a key factor in assessing forest growth (Ma et al. 2023; McGregor et al. 2021), calculating forest volume (Xu et al. 2019) and carbon storage (Yao et al. 2018), as well as guiding sustainable forest management practices (Xu et al. 2023). Please refer to Line 47-48 for details.
Comment 3: Line 69: I think there's an extra 'from' written here, delete it.
Reply 3: We are very sorry for our incorrect English expression; we have made correction after checking. Please refer to Line 69-70 for details.
Comment 4: Figure 1 presents the content comprehensively; however, the four images in step 4 are not very clear, making it difficult to see the legend details. I suggest improving their clarity. I also noticed that these four subplots might be the same as the product images and uncertainty analysis figures shown later. Adjustments could be made accordingly.
Reply 4: We thank the reviewer for pointing out this issue. The legend and uncertainty analysis figures in Figure 1 have been adjusted.
Comment 5: Although UAV LiDAR point density is generally high, it still affects the extraction of forest attributes to some extent. Therefore, in Table 1, it would be helpful to add point density values under commonly used UAV flight parameters. This will provide a better introduction to the data, and I recommend adding this column.
Reply 5：We thank the reviewer for pointing out this issue, and we have done it according to your ideas. Please refer to Table 1 for details.
Comment 6: Table 2, Proportion of forest area covered by drone lidar data, is this value the ratio of the area where data was collected to the forest area in different Vegetation divisions?
Reply 6: Yes. For clearer explanation, we have further added note explanations. Please refer to Table 2 for details.
Comment 7: A figure should be added to Section 2.2 to visually present the field data distribution?
Reply 7: Thank you for the reviewer's reminder. Considering that field data and lidar data display more clearly, we have added the field data distribution in Supplementary Figure S1. Please refer to supplementary Figure S1 for details.
Comment 8: Line 156: I noticed that each plot of field data covers an area greater than 400 square meters, while your product has a resolution of 30 meters. Could this discrepancy affect the validation results?
Reply 8: In China's forest resource surveys, the differences in plot size have a minimal impact on the accuracy of stand height estimation mainly due to a sufficient number of samples, flexibility in plot size and shape, relatively stable forest structures, data standardization processes (Lohr, S. L. 2000; Gregoire, T. G., & Valentine, H. T. 2008; Paul TSH, et al.2019).
Certainly, due to time and labor cost constraints, there are some limitations in the sample data collection for this study, which have been addressed in the manuscript. Please refer to Line 165-166 for details.
References:
Lohr, S. L. (2000). Sampling: design and analysis. Technometrics, 42(2), 223-224.
Gregoire, T. G., & Valentine, H. T. (2008). Sampling strategies for natural resources and the environment. international journal of environmental analytical chemistry.
Paul TSH, Kimberley MO, Beets PN. Thinking outside the square: Evidence that plot shape and layout in forest inventories can bias estimates of stand metrics. Methods Ecol Evol. 2019; 10: 381–388. https://doi.org/10.1111/2041-210X.13113
Comment 9: Figure 3 only shows the weighting method for 𝑤2, has a comparison been made between the weighting of 𝑤1 and 𝑤2?
Reply 9: In Supplementary Table S4, we have compared the deviations between weighted mean heights with different weights (𝑤1 and 𝑤2) and Lorey’s mean height (national forest inventory data). Please refer to supplementary Table S4 for details.
Comment 10: Line 197: delete ‘those’.
Reply 10: We are very sorry for our incorrect writing and it is rectified. Please refer to Line 204 for details.
Comment 11: In Section 2.5.2, the referenced section should be Section 2.5.1, not Section 2.4.1.
Reply 11: Thank you for the reviewer's reminder, we have revised this error. Please refer to Line 248 and 258 for details.
Comment 12: In Figures 10 and 11, the uncertainty is given in percentage (%). The unit of 𝜀ℎ𝑖should be specified in the Methods section.
Reply 12: We appreciate it very much for this suggestion, and we have done it according to your ideas. Please refer to Equations 16-20 for details.
-
AC1: 'Reply on RC1', Haitao Yang, 05 Sep 2024
-
RC2: 'Comment on essd-2024-274', Anonymous Referee #2, 03 Sep 2024
This is a timely paper which addresses the importance of mapping forest stand mean height. Different with previous studies that generating a canopy height model (CHM) to calculate tree height based on the statistical relationships between plot-level LiDAR metrics. This study adopted tree-based approach to map arithmetic mean height and weighted mean height by using massive UAV LiDAR data. By combining numerous remote sensing data and ML-based mixed-effects model, high accuracy has achieved to map the wall-to-wall forest stand mean height of China. The authors did a lot of work and adopted advanced models to provide a reliable product. However, some technical details need further clarification from the authors. The expression of some key results and the structure of the paper can still be improved. The specific suggestions are as follows.
1 Line 22: deleted estimation.
2 Line 45: The author used arithmetic mean height (ℎ𝑎) and weighted mean height (ℎ𝑤) to represent Forest Stand Mean Height. The similarities and differences between these two metrics should be explained at the beginning of the Introduction.
3 Line 69: I think there's an extra 'from' written here, delete it.
4 Figure 1 presents the content comprehensively; however, the four images in step 4 are not very clear, making it difficult to see the legend details. I suggest improving their clarity. I also noticed that these four subplots might be the same as the product images and uncertainty analysis figures shown later. Adjustments could be made accordingly.
5 Although UAV LiDAR point density is generally high, it still affects the extraction of forest attributes to some extent. Therefore, in Table 1, it would be helpful to add point density values under commonly used UAV flight parameters. This will provide a better introduction to the data, and I recommend adding this column.
6 Table 2, Proportion of forest area covered by drone lidar data, is this value the ratio of the area where data was collected to the forest area in different Vegetation divisions?
7 A figure should be added to Section 2.2 to visually present the field data distribution?
8 Line 156: I noticed that each plot of field data covers an area greater than 400 square meters, while your product has a resolution of 30 meters. Could this discrepancy affect the validation results?
9 Figure 3 only shows the weighting method for 𝑤2, has a comparison been made between the weighting of 𝑤1 and 𝑤2?
10 Line 197: delete ‘those’.
11 In Section 2.5.2, the referenced section should be Section 2.5.1, not Section 2.4.1.
12 In Figures 10 and 11, the uncertainty is given in percentage (%). The unit of 𝜀ℎ𝑖 should be specified in the Methods section.Citation: https://doi.org/10.5194/essd-2024-274-RC2 -
AC2: 'Reply on RC2', Haitao Yang, 05 Sep 2024
Dear Editor and Reviewer:
On behalf of my co-authors, we thank you very much for giving us an opportunity to revise our manuscript, and we also appreciate reviewers very much for their positive and constructive comments and suggestions on our manuscript entitled “Enhancing High-Resolution Forest Stand Mean Height Mapping in China through an Individual Tree-Based Approach with Close-Range LiDAR Data” (Manuscript Number: essd-2024-274).
We revised the manuscript according to these comments and suggestions. All changes were marked in highlight text in the revised manuscript. The line numbers in the response are the corresponding line numbers in the revised version.
Once again, thank you very much for your comments and suggestions.
Comment 1: Line 22: deleted estimation.
Reply 1: Thanks to reviewer for reminder, the estimation has been deleted in the Abstract. Please refer to Line 21-22 for details.
Comment 2: Line 45: The author used arithmetic mean height (ℎ𝑎) and weighted mean height (ℎ𝑤) to represent Forest Stand Mean Height. The similarities and differences between these two metrics should be explained at the beginning of the Introduction.
Reply 2: In the introduction, we described the differences in calculation methods and the similarities in application directions. The detailed similarities and differences were explained in the formula section and discussed in the discussion section.
The differences:
Forest stand height denotes the mean height of trees within a stand/plot, including arithmetic mean height and mean height weighted in proportion to their basal area (weighted mean height or Lorey’s mean height) (Laar and Akça 2007; Masaka et al. 2013). Please refer to Line 45-47 for details.
The similarities:
It serves as a key factor in assessing forest growth (Ma et al. 2023; McGregor et al. 2021), calculating forest volume (Xu et al. 2019) and carbon storage (Yao et al. 2018), as well as guiding sustainable forest management practices (Xu et al. 2023). Please refer to Line 47-48 for details.
Comment 3: Line 69: I think there's an extra 'from' written here, delete it.
Reply 3: We are very sorry for our incorrect English expression; we have made correction after checking. Please refer to Line 69-70 for details.
Comment 4: Figure 1 presents the content comprehensively; however, the four images in step 4 are not very clear, making it difficult to see the legend details. I suggest improving their clarity. I also noticed that these four subplots might be the same as the product images and uncertainty analysis figures shown later. Adjustments could be made accordingly.
Reply 4: We thank the reviewer for pointing out this issue. The legend and uncertainty analysis figures in Figure 1 have been adjusted.
Comment 5: Although UAV LiDAR point density is generally high, it still affects the extraction of forest attributes to some extent. Therefore, in Table 1, it would be helpful to add point density values under commonly used UAV flight parameters. This will provide a better introduction to the data, and I recommend adding this column.
Reply 5：We thank the reviewer for pointing out this issue, and we have done it according to your ideas. Please refer to Table 1 for details.
Comment 6: Table 2, Proportion of forest area covered by drone lidar data, is this value the ratio of the area where data was collected to the forest area in different Vegetation divisions?
Reply 6: Yes. For clearer explanation, we have further added note explanations. Please refer to Table 2 for details.
Comment 7: A figure should be added to Section 2.2 to visually present the field data distribution?
Reply 7: Thank you for the reviewer's reminder. Considering that field data and lidar data display more clearly, we have added the field data distribution in Supplementary Figure S1. Please refer to supplementary Figure S1 for details.
Comment 8: Line 156: I noticed that each plot of field data covers an area greater than 400 square meters, while your product has a resolution of 30 meters. Could this discrepancy affect the validation results?
Reply 8: In China's forest resource surveys, the differences in plot size have a minimal impact on the accuracy of stand height estimation mainly due to a sufficient number of samples, flexibility in plot size and shape, relatively stable forest structures, data standardization processes (Lohr, S. L. 2000; Gregoire, T. G., & Valentine, H. T. 2008; Paul TSH, et al.2019).
Certainly, due to time and labor cost constraints, there are some limitations in the sample data collection for this study, which have been addressed in the manuscript. Please refer to Line 165-166 for details.
References:
Lohr, S. L. (2000). Sampling: design and analysis. Technometrics, 42(2), 223-224.
Gregoire, T. G., & Valentine, H. T. (2008). Sampling strategies for natural resources and the environment. international journal of environmental analytical chemistry.
Paul TSH, Kimberley MO, Beets PN. Thinking outside the square: Evidence that plot shape and layout in forest inventories can bias estimates of stand metrics. Methods Ecol Evol. 2019; 10: 381–388. https://doi.org/10.1111/2041-210X.13113
Comment 9: Figure 3 only shows the weighting method for 𝑤2, has a comparison been made between the weighting of 𝑤1 and 𝑤2?
Reply 9: In Supplementary Table S4, we have compared the deviations between weighted mean heights with different weights (𝑤1 and 𝑤2) and Lorey’s mean height (national forest inventory data). Please refer to supplementary Table S4 for details.
Comment 10: Line 197: delete ‘those’.
Reply 10: We are very sorry for our incorrect writing and it is rectified. Please refer to Line 204 for details.
Comment 11: In Section 2.5.2, the referenced section should be Section 2.5.1, not Section 2.4.1.
Reply 11: Thank you for the reviewer's reminder, we have revised this error. Please refer to Line 248 and 258 for details.
Comment 12: In Figures 10 and 11, the uncertainty is given in percentage (%). The unit of 𝜀ℎ𝑖should be specified in the Methods section.
Reply 12: We appreciate it very much for this suggestion, and we have done it according to your ideas. Please refer to Equations 16-20 for details.
-
AC1: 'Reply on RC1', Haitao Yang, 05 Sep 2024
Dear Editor and Reviewer:
On behalf of my co-authors, we thank you very much for giving us an opportunity to revise our manuscript, and we also appreciate reviewers very much for their positive and constructive comments and suggestions on our manuscript entitled “Enhancing High-Resolution Forest Stand Mean Height Mapping in China through an Individual Tree-Based Approach with Close-Range LiDAR Data” (Manuscript Number: essd-2024-274).
We revised the manuscript according to these comments and suggestions. All changes were marked in highlight text in the revised manuscript. The line numbers in the response are the corresponding line numbers in the revised version.
Once again, thank you very much for your comments and suggestions.
Comment 1:
In the part of Abstract, ‘Forest stands mean height is a critical indicator in forestry, playing a pivotal role in various aspects such as forest inventory estimation,’ forest inventory estimation is suggested to be modified to forest inventory with various scales, which is more reasonable.
Reply 1: Thank you very much for your professional advice, we have changed ‘forest inventory estimation’ to ‘forest inventory’ at Line 21-22.
Comment 2:
In the line of 69: The height metrics from obtained from this approach is forest canopy height, which include not only the actual tree height. There is one mistake in the expression. The sentence should be corrected: The height metrics obtained from this approach is forest canopy height.
Reply 2: The mistake has been corrected according to your kind advices and detailed suggestions. Please refer to Line 69-70 for details.
Comment 3:
In terms of data, various types of data collected over a span of 6 years are included in this manuscript, such as ground measured samples, LiDAR data obtained from different sensors, and remote sensing images. How can these datasets be matched on a temporal scale? Additionally, how can reduce the limitations of images acquired in different years and seasons?
Reply 3：Changes in forest resources tend to occur relatively slowly, and a 5-year period is a sufficiently long-time span to capture significant change trends. The temporal scale for China's national-level forest resource inventory is set at 5 years, aiming to balance the need for real-time data with long-term trend observation. This time span is long enough to detect significant changes in forest ecosystems, yet short enough to ensure that policies and management measures can be promptly adjusted based on the most recent data.
As of 2015, the application of LiDAR has not been widely adopted in forest remote sensing research in China. Considering the cost and the difficulty of data collection, it was challenging to collect extensive, high-point density and accurate data across China within a short timeframe. Considering the nationwide data coverage, the final dataset for this study spans 6 years (one year longer than the time span of the national inventory). This represents a limitation of the data used in this study, which is discussed in the paper. Please refer to Line 485-487 for details.
Comment 4:
The formula of determining coefficients (formula 11), y ̅_iis not the mean value for the observed values. y ̅ is recommended. In the formula 16, the means of y ̅ also should be expressed.
Reply 4: We have corrected the formulas. Please refer to equations 11-20 for details.
Comment 5:
In the manuscript, three accuracy indices were employed to evaluate the performance of models. However, when evaluating results with the same RMSE in various height forests, it is recommended to include rRMSE.
Reply 5: We agreed with the reviewer's comment and added the rRMSE to the Table5, which evaluating results with the same RMSE in various height forests. Please refer to Table 5 for details.
Comment 6:
In Figure 3, it is evident that an overestimation of forest stand height occurs when the weighted average of tree height squared is applied for forest stands taller than 14 meters. Please provide the underlying reasons.
Reply 6:
We greatly appreciate the reviewer’s insightful question. In response, we have explored the issue from both theoretical and empirical perspectives to provide a comprehensive answer. Please refer to Line 440-445 for details.
For a detailed mathematical and empirical proof, please refer to the supplementary materials.
Comment 7:
The decimal places of precision indexes in this paper should be consistent, such as Tabel 5.
Reply 7: We have adjusted to ensure the consistency of decimal places for the indexes. Please refer to Table 5 for details.
-
AC2: 'Reply on RC2', Haitao Yang, 05 Sep 2024
Status: closed
-
RC1: 'Comment on essd-2024-274', Anonymous Referee #1, 31 Aug 2024
Forest stand mean height is a critical indicator in forestry, playing a pivotal role in various aspects such as forest inventory, sustainable forest management practices, climate change mitigation strategies, monitoring of forest structure changes, and wildlife habitat assessment. However, mapping the national scale forest stand mean height across China is still a challenge, recently. In this manuscript, a tree-based approach to create spatially continuous forest stand mean height maps of China was developed by integrating high-point density, high-precision close-range LiDAR data and multisource remote sensing data. The methods employed for data acquisition, processing, and analysis are methodical and rational. The accuracy of the estimated results is commendable, thereby demonstrating significant potential for practical applications. However, some issues should be further corrected or clarified before publication.
in the part of Abstract, ‘Forest stands mean height is a critical indicator in forestry, playing a pivotal role in various aspects such as forest inventory estimation,’ forest inventory estimation is suggested to be modified to forest inventory with various scales, which is more reasonable.
In the line of 69: The height metrics from obtained from this approach is forest canopy height, which include not only the actual tree height. There is one mistake in the expression. The sentence should be corrected: The height metrics obtained from this approach is forest canopy height.
In terms of data, various types of data collected over a span of 6 years are included in this manuscript, such as ground measured samples, LiDAR data obtained from different sensors, and remote sensing images. How can these datasets be matched on a temporal scale? Additionally, how can reduce the limitations of images acquired in different years and seasons?
The formula of determining coefficients (formula 11), y ̅_iis not the mean value for the observed values. y ̅ is recommended. In the formula 16, the means of y ̅ also should be expressed.
In the manuscript, three accuracy indices were employed to evaluate the performance of models. However, when evaluating results with the same RMSE in various height forests, it is recommended to include rRMSE.
In Figure 3, it is evident that an overestimation of forest stand height occurs when the weighted average of tree height squared is applied for forest stands taller than 14 meters. Please provide the underlying reasons.
The decimal places of precision indexes in this paper should be consistent, such as Tabel 5.Citation: https://doi.org/10.5194/essd-2024-274-RC1 -
AC1: 'Reply on RC1', Haitao Yang, 05 Sep 2024
Dear Editor and Reviewer:
On behalf of my co-authors, we thank you very much for giving us an opportunity to revise our manuscript, and we also appreciate reviewers very much for their positive and constructive comments and suggestions on our manuscript entitled “Enhancing High-Resolution Forest Stand Mean Height Mapping in China through an Individual Tree-Based Approach with Close-Range LiDAR Data” (Manuscript Number: essd-2024-274).
We revised the manuscript according to these comments and suggestions. All changes were marked in highlight text in the revised manuscript. The line numbers in the response are the corresponding line numbers in the revised version.
Once again, thank you very much for your comments and suggestions.
Comment 1:
In the part of Abstract, ‘Forest stands mean height is a critical indicator in forestry, playing a pivotal role in various aspects such as forest inventory estimation,’ forest inventory estimation is suggested to be modified to forest inventory with various scales, which is more reasonable.
Reply 1: Thank you very much for your professional advice, we have changed ‘forest inventory estimation’ to ‘forest inventory’ at Line 21-22.
Comment 2:
In the line of 69: The height metrics from obtained from this approach is forest canopy height, which include not only the actual tree height. There is one mistake in the expression. The sentence should be corrected: The height metrics obtained from this approach is forest canopy height.
Reply 2: The mistake has been corrected according to your kind advices and detailed suggestions. Please refer to Line 69-70 for details.
Comment 3:
In terms of data, various types of data collected over a span of 6 years are included in this manuscript, such as ground measured samples, LiDAR data obtained from different sensors, and remote sensing images. How can these datasets be matched on a temporal scale? Additionally, how can reduce the limitations of images acquired in different years and seasons?
Reply 3：Changes in forest resources tend to occur relatively slowly, and a 5-year period is a sufficiently long-time span to capture significant change trends. The temporal scale for China's national-level forest resource inventory is set at 5 years, aiming to balance the need for real-time data with long-term trend observation. This time span is long enough to detect significant changes in forest ecosystems, yet short enough to ensure that policies and management measures can be promptly adjusted based on the most recent data.
As of 2015, the application of LiDAR has not been widely adopted in forest remote sensing research in China. Considering the cost and the difficulty of data collection, it was challenging to collect extensive, high-point density and accurate data across China within a short timeframe. Considering the nationwide data coverage, the final dataset for this study spans 6 years (one year longer than the time span of the national inventory). This represents a limitation of the data used in this study, which is discussed in the paper. Please refer to Line 485-487 for details.
Comment 4:
The formula of determining coefficients (formula 11), y ̅_iis not the mean value for the observed values. y ̅ is recommended. In the formula 16, the means of y ̅ also should be expressed.
Reply 4: We have corrected the formulas. Please refer to equations 11-20 for details.
Comment 5:
In the manuscript, three accuracy indices were employed to evaluate the performance of models. However, when evaluating results with the same RMSE in various height forests, it is recommended to include rRMSE.
Reply 5: We agreed with the reviewer's comment and added the rRMSE to the Table5, which evaluating results with the same RMSE in various height forests. Please refer to Table 5 for details.
Comment 6:
In Figure 3, it is evident that an overestimation of forest stand height occurs when the weighted average of tree height squared is applied for forest stands taller than 14 meters. Please provide the underlying reasons.
Reply 6:
We greatly appreciate the reviewer’s insightful question. In response, we have explored the issue from both theoretical and empirical perspectives to provide a comprehensive answer. Please refer to Line 440-445 for details.
For a detailed mathematical and empirical proof, please refer to the supplementary materials.
Comment 7:
The decimal places of precision indexes in this paper should be consistent, such as Tabel 5.
Reply 7: We have adjusted to ensure the consistency of decimal places for the indexes. Please refer to Table 5 for details.
-
AC2: 'Reply on RC2', Haitao Yang, 05 Sep 2024
Dear Editor and Reviewer:
On behalf of my co-authors, we thank you very much for giving us an opportunity to revise our manuscript, and we also appreciate reviewers very much for their positive and constructive comments and suggestions on our manuscript entitled “Enhancing High-Resolution Forest Stand Mean Height Mapping in China through an Individual Tree-Based Approach with Close-Range LiDAR Data” (Manuscript Number: essd-2024-274).
We revised the manuscript according to these comments and suggestions. All changes were marked in highlight text in the revised manuscript. The line numbers in the response are the corresponding line numbers in the revised version.
Once again, thank you very much for your comments and suggestions.
Comment 1: Line 22: deleted estimation.
Reply 1: Thanks to reviewer for reminder, the estimation has been deleted in the Abstract. Please refer to Line 21-22 for details.
Comment 2: Line 45: The author used arithmetic mean height (ℎ𝑎) and weighted mean height (ℎ𝑤) to represent Forest Stand Mean Height. The similarities and differences between these two metrics should be explained at the beginning of the Introduction.
Reply 2: In the introduction, we described the differences in calculation methods and the similarities in application directions. The detailed similarities and differences were explained in the formula section and discussed in the discussion section.
The differences:
Forest stand height denotes the mean height of trees within a stand/plot, including arithmetic mean height and mean height weighted in proportion to their basal area (weighted mean height or Lorey’s mean height) (Laar and Akça 2007; Masaka et al. 2013). Please refer to Line 45-47 for details.
The similarities:
It serves as a key factor in assessing forest growth (Ma et al. 2023; McGregor et al. 2021), calculating forest volume (Xu et al. 2019) and carbon storage (Yao et al. 2018), as well as guiding sustainable forest management practices (Xu et al. 2023). Please refer to Line 47-48 for details.
Comment 3: Line 69: I think there's an extra 'from' written here, delete it.
Reply 3: We are very sorry for our incorrect English expression; we have made correction after checking. Please refer to Line 69-70 for details.
Comment 4: Figure 1 presents the content comprehensively; however, the four images in step 4 are not very clear, making it difficult to see the legend details. I suggest improving their clarity. I also noticed that these four subplots might be the same as the product images and uncertainty analysis figures shown later. Adjustments could be made accordingly.
Reply 4: We thank the reviewer for pointing out this issue. The legend and uncertainty analysis figures in Figure 1 have been adjusted.
Comment 5: Although UAV LiDAR point density is generally high, it still affects the extraction of forest attributes to some extent. Therefore, in Table 1, it would be helpful to add point density values under commonly used UAV flight parameters. This will provide a better introduction to the data, and I recommend adding this column.
Reply 5：We thank the reviewer for pointing out this issue, and we have done it according to your ideas. Please refer to Table 1 for details.
Comment 6: Table 2, Proportion of forest area covered by drone lidar data, is this value the ratio of the area where data was collected to the forest area in different Vegetation divisions?
Reply 6: Yes. For clearer explanation, we have further added note explanations. Please refer to Table 2 for details.
Comment 7: A figure should be added to Section 2.2 to visually present the field data distribution?
Reply 7: Thank you for the reviewer's reminder. Considering that field data and lidar data display more clearly, we have added the field data distribution in Supplementary Figure S1. Please refer to supplementary Figure S1 for details.
Comment 8: Line 156: I noticed that each plot of field data covers an area greater than 400 square meters, while your product has a resolution of 30 meters. Could this discrepancy affect the validation results?
Reply 8: In China's forest resource surveys, the differences in plot size have a minimal impact on the accuracy of stand height estimation mainly due to a sufficient number of samples, flexibility in plot size and shape, relatively stable forest structures, data standardization processes (Lohr, S. L. 2000; Gregoire, T. G., & Valentine, H. T. 2008; Paul TSH, et al.2019).
Certainly, due to time and labor cost constraints, there are some limitations in the sample data collection for this study, which have been addressed in the manuscript. Please refer to Line 165-166 for details.
References:
Lohr, S. L. (2000). Sampling: design and analysis. Technometrics, 42(2), 223-224.
Gregoire, T. G., & Valentine, H. T. (2008). Sampling strategies for natural resources and the environment. international journal of environmental analytical chemistry.
Paul TSH, Kimberley MO, Beets PN. Thinking outside the square: Evidence that plot shape and layout in forest inventories can bias estimates of stand metrics. Methods Ecol Evol. 2019; 10: 381–388. https://doi.org/10.1111/2041-210X.13113
Comment 9: Figure 3 only shows the weighting method for 𝑤2, has a comparison been made between the weighting of 𝑤1 and 𝑤2?
Reply 9: In Supplementary Table S4, we have compared the deviations between weighted mean heights with different weights (𝑤1 and 𝑤2) and Lorey’s mean height (national forest inventory data). Please refer to supplementary Table S4 for details.
Comment 10: Line 197: delete ‘those’.
Reply 10: We are very sorry for our incorrect writing and it is rectified. Please refer to Line 204 for details.
Comment 11: In Section 2.5.2, the referenced section should be Section 2.5.1, not Section 2.4.1.
Reply 11: Thank you for the reviewer's reminder, we have revised this error. Please refer to Line 248 and 258 for details.
Comment 12: In Figures 10 and 11, the uncertainty is given in percentage (%). The unit of 𝜀ℎ𝑖should be specified in the Methods section.
Reply 12: We appreciate it very much for this suggestion, and we have done it according to your ideas. Please refer to Equations 16-20 for details.
-
AC1: 'Reply on RC1', Haitao Yang, 05 Sep 2024
-
RC2: 'Comment on essd-2024-274', Anonymous Referee #2, 03 Sep 2024
This is a timely paper which addresses the importance of mapping forest stand mean height. Different with previous studies that generating a canopy height model (CHM) to calculate tree height based on the statistical relationships between plot-level LiDAR metrics. This study adopted tree-based approach to map arithmetic mean height and weighted mean height by using massive UAV LiDAR data. By combining numerous remote sensing data and ML-based mixed-effects model, high accuracy has achieved to map the wall-to-wall forest stand mean height of China. The authors did a lot of work and adopted advanced models to provide a reliable product. However, some technical details need further clarification from the authors. The expression of some key results and the structure of the paper can still be improved. The specific suggestions are as follows.
1 Line 22: deleted estimation.
2 Line 45: The author used arithmetic mean height (ℎ𝑎) and weighted mean height (ℎ𝑤) to represent Forest Stand Mean Height. The similarities and differences between these two metrics should be explained at the beginning of the Introduction.
3 Line 69: I think there's an extra 'from' written here, delete it.
4 Figure 1 presents the content comprehensively; however, the four images in step 4 are not very clear, making it difficult to see the legend details. I suggest improving their clarity. I also noticed that these four subplots might be the same as the product images and uncertainty analysis figures shown later. Adjustments could be made accordingly.
5 Although UAV LiDAR point density is generally high, it still affects the extraction of forest attributes to some extent. Therefore, in Table 1, it would be helpful to add point density values under commonly used UAV flight parameters. This will provide a better introduction to the data, and I recommend adding this column.
6 Table 2, Proportion of forest area covered by drone lidar data, is this value the ratio of the area where data was collected to the forest area in different Vegetation divisions?
7 A figure should be added to Section 2.2 to visually present the field data distribution?
8 Line 156: I noticed that each plot of field data covers an area greater than 400 square meters, while your product has a resolution of 30 meters. Could this discrepancy affect the validation results?
9 Figure 3 only shows the weighting method for 𝑤2, has a comparison been made between the weighting of 𝑤1 and 𝑤2?
10 Line 197: delete ‘those’.
11 In Section 2.5.2, the referenced section should be Section 2.5.1, not Section 2.4.1.
12 In Figures 10 and 11, the uncertainty is given in percentage (%). The unit of 𝜀ℎ𝑖 should be specified in the Methods section.Citation: https://doi.org/10.5194/essd-2024-274-RC2 -
AC2: 'Reply on RC2', Haitao Yang, 05 Sep 2024
Dear Editor and Reviewer:
On behalf of my co-authors, we thank you very much for giving us an opportunity to revise our manuscript, and we also appreciate reviewers very much for their positive and constructive comments and suggestions on our manuscript entitled “Enhancing High-Resolution Forest Stand Mean Height Mapping in China through an Individual Tree-Based Approach with Close-Range LiDAR Data” (Manuscript Number: essd-2024-274).
We revised the manuscript according to these comments and suggestions. All changes were marked in highlight text in the revised manuscript. The line numbers in the response are the corresponding line numbers in the revised version.
Once again, thank you very much for your comments and suggestions.
Comment 1: Line 22: deleted estimation.
Reply 1: Thanks to reviewer for reminder, the estimation has been deleted in the Abstract. Please refer to Line 21-22 for details.
Comment 2: Line 45: The author used arithmetic mean height (ℎ𝑎) and weighted mean height (ℎ𝑤) to represent Forest Stand Mean Height. The similarities and differences between these two metrics should be explained at the beginning of the Introduction.
Reply 2: In the introduction, we described the differences in calculation methods and the similarities in application directions. The detailed similarities and differences were explained in the formula section and discussed in the discussion section.
The differences:
Forest stand height denotes the mean height of trees within a stand/plot, including arithmetic mean height and mean height weighted in proportion to their basal area (weighted mean height or Lorey’s mean height) (Laar and Akça 2007; Masaka et al. 2013). Please refer to Line 45-47 for details.
The similarities:
It serves as a key factor in assessing forest growth (Ma et al. 2023; McGregor et al. 2021), calculating forest volume (Xu et al. 2019) and carbon storage (Yao et al. 2018), as well as guiding sustainable forest management practices (Xu et al. 2023). Please refer to Line 47-48 for details.
Comment 3: Line 69: I think there's an extra 'from' written here, delete it.
Reply 3: We are very sorry for our incorrect English expression; we have made correction after checking. Please refer to Line 69-70 for details.
Comment 4: Figure 1 presents the content comprehensively; however, the four images in step 4 are not very clear, making it difficult to see the legend details. I suggest improving their clarity. I also noticed that these four subplots might be the same as the product images and uncertainty analysis figures shown later. Adjustments could be made accordingly.
Reply 4: We thank the reviewer for pointing out this issue. The legend and uncertainty analysis figures in Figure 1 have been adjusted.
Comment 5: Although UAV LiDAR point density is generally high, it still affects the extraction of forest attributes to some extent. Therefore, in Table 1, it would be helpful to add point density values under commonly used UAV flight parameters. This will provide a better introduction to the data, and I recommend adding this column.
Reply 5：We thank the reviewer for pointing out this issue, and we have done it according to your ideas. Please refer to Table 1 for details.
Comment 6: Table 2, Proportion of forest area covered by drone lidar data, is this value the ratio of the area where data was collected to the forest area in different Vegetation divisions?
Reply 6: Yes. For clearer explanation, we have further added note explanations. Please refer to Table 2 for details.
Comment 7: A figure should be added to Section 2.2 to visually present the field data distribution?
Reply 7: Thank you for the reviewer's reminder. Considering that field data and lidar data display more clearly, we have added the field data distribution in Supplementary Figure S1. Please refer to supplementary Figure S1 for details.
Comment 8: Line 156: I noticed that each plot of field data covers an area greater than 400 square meters, while your product has a resolution of 30 meters. Could this discrepancy affect the validation results?
Reply 8: In China's forest resource surveys, the differences in plot size have a minimal impact on the accuracy of stand height estimation mainly due to a sufficient number of samples, flexibility in plot size and shape, relatively stable forest structures, data standardization processes (Lohr, S. L. 2000; Gregoire, T. G., & Valentine, H. T. 2008; Paul TSH, et al.2019).
Certainly, due to time and labor cost constraints, there are some limitations in the sample data collection for this study, which have been addressed in the manuscript. Please refer to Line 165-166 for details.
References:
Lohr, S. L. (2000). Sampling: design and analysis. Technometrics, 42(2), 223-224.
Gregoire, T. G., & Valentine, H. T. (2008). Sampling strategies for natural resources and the environment. international journal of environmental analytical chemistry.
Paul TSH, Kimberley MO, Beets PN. Thinking outside the square: Evidence that plot shape and layout in forest inventories can bias estimates of stand metrics. Methods Ecol Evol. 2019; 10: 381–388. https://doi.org/10.1111/2041-210X.13113
Comment 9: Figure 3 only shows the weighting method for 𝑤2, has a comparison been made between the weighting of 𝑤1 and 𝑤2?
Reply 9: In Supplementary Table S4, we have compared the deviations between weighted mean heights with different weights (𝑤1 and 𝑤2) and Lorey’s mean height (national forest inventory data). Please refer to supplementary Table S4 for details.
Comment 10: Line 197: delete ‘those’.
Reply 10: We are very sorry for our incorrect writing and it is rectified. Please refer to Line 204 for details.
Comment 11: In Section 2.5.2, the referenced section should be Section 2.5.1, not Section 2.4.1.
Reply 11: Thank you for the reviewer's reminder, we have revised this error. Please refer to Line 248 and 258 for details.
Comment 12: In Figures 10 and 11, the uncertainty is given in percentage (%). The unit of 𝜀ℎ𝑖should be specified in the Methods section.
Reply 12: We appreciate it very much for this suggestion, and we have done it according to your ideas. Please refer to Equations 16-20 for details.
-
AC1: 'Reply on RC1', Haitao Yang, 05 Sep 2024
Dear Editor and Reviewer:
On behalf of my co-authors, we thank you very much for giving us an opportunity to revise our manuscript, and we also appreciate reviewers very much for their positive and constructive comments and suggestions on our manuscript entitled “Enhancing High-Resolution Forest Stand Mean Height Mapping in China through an Individual Tree-Based Approach with Close-Range LiDAR Data” (Manuscript Number: essd-2024-274).
We revised the manuscript according to these comments and suggestions. All changes were marked in highlight text in the revised manuscript. The line numbers in the response are the corresponding line numbers in the revised version.
Once again, thank you very much for your comments and suggestions.
Comment 1:
In the part of Abstract, ‘Forest stands mean height is a critical indicator in forestry, playing a pivotal role in various aspects such as forest inventory estimation,’ forest inventory estimation is suggested to be modified to forest inventory with various scales, which is more reasonable.
Reply 1: Thank you very much for your professional advice, we have changed ‘forest inventory estimation’ to ‘forest inventory’ at Line 21-22.
Comment 2:
In the line of 69: The height metrics from obtained from this approach is forest canopy height, which include not only the actual tree height. There is one mistake in the expression. The sentence should be corrected: The height metrics obtained from this approach is forest canopy height.
Reply 2: The mistake has been corrected according to your kind advices and detailed suggestions. Please refer to Line 69-70 for details.
Comment 3:
In terms of data, various types of data collected over a span of 6 years are included in this manuscript, such as ground measured samples, LiDAR data obtained from different sensors, and remote sensing images. How can these datasets be matched on a temporal scale? Additionally, how can reduce the limitations of images acquired in different years and seasons?
Reply 3：Changes in forest resources tend to occur relatively slowly, and a 5-year period is a sufficiently long-time span to capture significant change trends. The temporal scale for China's national-level forest resource inventory is set at 5 years, aiming to balance the need for real-time data with long-term trend observation. This time span is long enough to detect significant changes in forest ecosystems, yet short enough to ensure that policies and management measures can be promptly adjusted based on the most recent data.
As of 2015, the application of LiDAR has not been widely adopted in forest remote sensing research in China. Considering the cost and the difficulty of data collection, it was challenging to collect extensive, high-point density and accurate data across China within a short timeframe. Considering the nationwide data coverage, the final dataset for this study spans 6 years (one year longer than the time span of the national inventory). This represents a limitation of the data used in this study, which is discussed in the paper. Please refer to Line 485-487 for details.
Comment 4:
The formula of determining coefficients (formula 11), y ̅_iis not the mean value for the observed values. y ̅ is recommended. In the formula 16, the means of y ̅ also should be expressed.
Reply 4: We have corrected the formulas. Please refer to equations 11-20 for details.
Comment 5:
In the manuscript, three accuracy indices were employed to evaluate the performance of models. However, when evaluating results with the same RMSE in various height forests, it is recommended to include rRMSE.
Reply 5: We agreed with the reviewer's comment and added the rRMSE to the Table5, which evaluating results with the same RMSE in various height forests. Please refer to Table 5 for details.
Comment 6:
In Figure 3, it is evident that an overestimation of forest stand height occurs when the weighted average of tree height squared is applied for forest stands taller than 14 meters. Please provide the underlying reasons.
Reply 6:
We greatly appreciate the reviewer’s insightful question. In response, we have explored the issue from both theoretical and empirical perspectives to provide a comprehensive answer. Please refer to Line 440-445 for details.
For a detailed mathematical and empirical proof, please refer to the supplementary materials.
Comment 7:
The decimal places of precision indexes in this paper should be consistent, such as Tabel 5.
Reply 7: We have adjusted to ensure the consistency of decimal places for the indexes. Please refer to Table 5 for details.
-
AC2: 'Reply on RC2', Haitao Yang, 05 Sep 2024
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 432 | 102 | 110 | 644 | 25 | 11 | 12 |
- HTML: 432
- PDF: 102
- XML: 110
- Total: 644
- Supplement: 25
- BibTeX: 11
- EndNote: 12
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1