Multi-Platform LiDAR for Non-Destructive Individual Aboveground Biomass Estimation for Changbai Larch (Larix olgensis Henry) Using a Hierarchical Bayesian Approach

Round 1

Reviewer 1 Report

the bias in assembly obtained when unifying the two sources of data should be clearly mentioned in the introduction, as well as the potential biases that it creates

Comments for author File: Comments.pdf

Author Response

Reviewer #1:

The bias in assembly obtained when unifying the two sources of data should be clearly mentioned in the introduction, as well as the potential biases that it creates.

 

Response: Thanks for your careful review and the suggestions. In the introduction section, we added some description about the potential bias caused by the fusion of the multi-platform LiDAR data as follow:

Lines 75-82 on page 2:

“Although the fusion of multi-platform LiDAR data presents an opportunity to address the limitations of different LiDAR platforms [24], the different densities and scanning angles of multi-platform LiDAR may bring the potential bias for data registration then influence on the individual tree parameter estimation. Nowadays, the fusion of multi-platform LiDAR data reaches a satisfied registration accuracy (root-mean square error (RMSE) <30 cm) and is insensitive to individual tree segmentation errors, paving the way for individual parameter estimation [17, 20, 24, 25].”

 

Comments in peer-review-20374077.v1.pdf:

L19: R² alone is not sufficient to evaluate the accuracy.

Response: Thank you for the comment and suggestion. Since tree height, diameter at breast (DBH) and crown projection area (CPA) have different magnitude of error, it is difficult to qualify the RMSE and rRMSE using a single value. We have revised the sentence as follow:

Line 20 on page 1:

“…with high accuracies (all R²>0.9 and relatively low RMSE and rRMSE)”.

 

L277: How biomass data was obtained?

Response: Sorry for the confusion. We revised the “aboveground biomass” to “reference aboveground biomass” in line 307 on page 9 and the “AGB” to “Reference AGB” in Table 1 from the current version. We explained the calculation as follows:

Lines 226-228 on page 5:

“The reference AGB of individual tree in this study was calculated using the additive biomass equations of larch plantations proposed by Dong et al. [59], shown in Table A2 of Appendix A. The descriptive statistics of the main variables are presented in Table 1.”

 

L388-389: In both cases, there is an increase in residuals with DBH or AGB and this is not comforting in any way!

Response: Thank you for the comment. We agree with you that the residual plot is not satisfactory. However, table 5 showed that NLME performed well than NLS according to fitting statistics (e.g., FI, RMSE, AIC, BIC). So, we established hierarchical Bayesian models based on NLME in the sequent procedure. Since table 5 is sufficient to show the basic result, the figure 4 is deleted from the current version.

Author Response File: Author Response.docx

Reviewer 2 Report

The article entitled “Aboveground Biomass Estimation of Changbai Larch (Larix ol-2 gensis Henry) Using a Hierarchical Bayesian Approach Based on Multi-platform LiDAR Data” provides an interesting overview of the use of terrestrial and aerial LiDAR data, with a robust detection of individual three parameters.

 The proposed research uses the fused LiDAR data (TLS and UAV LiDAR) with a hierarchical Bayesian method to estimate ABG with the advantage of small sample size (<50). This approach has the potential to obtain individual tree AGB modeling with high accuracy.

 In general the manuscript has a good scientific quality and is well structured. The results contribute in the field of Remote Sensing and forest AGB inventories with the advantage of non destructive method with a very aceptable accuracy.

Author Response

Response: Thanks for your careful review and the encouragement.

Reviewer 3 Report

Review: Aboveground Biomass Estimation of Changbai Larch (Larix ol-2 gensis Henry) Using a Hierarchical Bayesian Approach Based 3 on Multi-platform LiDAR Data

 

Major revision

1.      This paper focuses on a Bayesian approach based on LiDAR data to estimate tree biomass for a species. First, the Bayesian approach does not add any value compared to the classical approach. The results are not statistically different between the two approaches. To perform a statistical test, the authors can perform a paired t-test for each estimated coefficient between the two approaches. Furthermore, why do the authors choose a method with different sample sizes? If the sample size affects the result, it should be mentioned in the introduction. Second, tree biomass estimation has been well documented and much more robust estimates are currently available for specific species.  The authors should include biomass estimates at the plot scale for multi-species. I believe that the addition of LiDAR data in this study is an added value but focused on the tree scale and not at the plot scale, this kind of data is insufficient for the validation of the method used and the conclusions given in this study.

2.      The experimental setting and sampling are not well described in this study. The authors present 8 plots of 0.09 ha with varying forest stages (A1: middle-age forest; A2: near-mature forest; A3: mature forest), but it is not known how the plots were set up? and each plot contains how many trees and species? what is the location of each plot? The trees that were measured with the dendrometric tools were selected according to what type of sampling? How are these trees distributed according to forest stages? etc....

3.      Section 2.3 method remains the highlight of this paper. However, this method is not well explained in the manuscript. The authors summarized their method in Figure 2 according to each objective. The description of the method does not allow the reader to understand the purpose of the study and how the authors used LiDAR and dendrometric data to improve tree biomass estimates. I think the authors need to separate two key aspects: the first aspect is the processing of LiDAR data in conjunction with dendrometric data, the second aspect is the data analysis that needs to be done according to each specific objective.

 

Minor revision

Title: please reword

Abstract: please start with a sentence on the need of tree AGB estimation

L24-27: Why the authors mention such a conclusion in the abstract when the results of their study do not show this. The authors should draw a conclusion based on the tree biomass estimate and not on the study methodology and especially not on the sampling. The methodology and sampling of this study do not allow for this conclusion (see major revisions).

Introduction

L42-56: This paragraph should be moved to a place where the authors give the interest of the study. The first three paragraphs of the introduction are paragraphs that give general information about tree biomass estimation. The allometric models to be used, the most significant predictors and the use of LiDAR data to improve these estimates.

L57-61: these two sentences should be removed in this paragraph which is focused on TLS approaches.

L75-87: this paragraph should be moved to the previous paragraph based on TLS approaches.

L99-123: I think these two paragraphs do not add any information on the quality of paper, since the Bayesian approach is not more important than the classical approach. Please remove or reword these paragraphs.

L141: please add the family of the larch species

 

Material and methods

L162-165: please add the new section on the plot sampling

Figure 1: please add the legend of color different in each plot

Table 1: please remove because it does not add any information which can be add in the text.

Reference data: please reword this section in dendrometric data

L196: please add the sampling description of 370 trees

L197: the authors measured tree diameter and height. Crown dimensions did not measure by the authors, I suggest that the authors use an allometric equation to estimate the crown projection area. I think this variable could be useful to fit allometric models between AGB and dendrometric variables

Table 2: please remove because it does not add any information which can be add in the text.

L210 The 2.3 Methods should be improved on its structure and content (see the major revision)

Results

L326-327: please remove

L340-L343: please reword these sentences

Table 5: why rRMSE (%) is 607.3 for CPA in CSP algorithm

Eq (15) please add estimated coefficients

Figure 4: please define the estimated AGB

Table 8: please move to appendices

 

Discussion

Any changes made in this manuscript should affect the discussion.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report

Title

Please reworded the title: "Multi-platform LiDAR for non-destructive individual aboveground biomass estimation in xxxxx"

 

Abstract

Please add the study site and sampling 

 

Introduction

L31-75: I think the first paragraph is ok, but I have a problem with the references that are used. This first paragraph informs about the importance of forests and the estimation of biomass and carbon stocks, the authors should use the references from a general consideration. Please see Pan Y. et al., 2011. A large and persistent carbon sink in the world’s forests. Science, 333, 988-993.

 

L76-78: This sentence showed that tree height and crown measurements come from forest inventory and it is difficult to obtain accurage of tree measurements. However, the authors could add two sentences on tree modelling including tree height and crown allometries in forests that can help to obtain the accurate of tree measurements. 

 

L85-87: There is now a large database of field crown data. Please see Jucker, T., et al. "Tallo: A global tree allometry and crown architecture database." Global change biology (2022). 

 

L168-178: I think this paragraph is a good addition to this revised version. But, I think this paragraph is not complete, because it lacks some information about the importance of using LiDAR and Bayesian method in biomass estimation, since this is already done using allometric equations. 

 

Materials and Methods

L209: please reworded "Study area and sampling". In this section, please have two distinctes paragraphes. Please add the sentences of Fig. 1c.

L254: Field Inventory data. Here, the authors present the data collection and I think the method of reference AGB estimation could be moved in the data analysis.

L299: I had a problem with the Figure 2. For example: the authors showed that the field inventory data is in the objective 1, I do not understand. The authors could better conceive this Figure. 

 

L404: please add the section data analysis that can inclure all analysis according to each study objective.

 

Results

L426-427: please add the values of this accurate

L440-442: please remove this sentence

Please merge Table 6 and Table 7

 

Discussion

L527: what do mean the mature method?

Please see Brede, B., Terryn, L., Barbier, N., Bartholomeus, H. M., Bartolo, R., Calders, K., ... & Herold, M. (2022). Non-destructive estimation of individual tree biomass: Allometric models, terrestrial and UAV laser scanning. Remote Sensing of Environment, 280, 113180.

Author Response

Manuscript ID: remotesensing-1793152

Aboveground Biomass Estimation of Changbai Larch (Larix olgensis Henry) Using a Hierarchical Bayesian Approach Based on Multi-platform LiDAR Data

 

Responses to comments

 

Response: We appreciate again the encouragement, careful review of the manuscript, and constructive comments and suggestions of the reviewer. After carefully considering all the comments, we have addressed every point of reviewers and made a number of revisions. All the changes are marked up using the “Track Changes” function in the text and also listed below.

 

Reviewer #3:

Title

Please reworded the title: "Multi-platform LiDAR for non-destructive individual aboveground biomass estimation in xxxxx"

Response: Thank you for the suggestion. We have already reworded the title of the article as: “Multi-platform LiDAR for non-destructive individual aboveground biomass estimation for Changbai Larch (Larix olgensis Henry) using a hierarchical Bayesian approach”.

 

Abstract

Please add the study site and sampling

Response: Thank you for the suggestion. Due to the word limit (200 words maximum) of abtract, we could only add some information in the abstract as follow:

Lines 19-21 on page 1:

“… the individual tree AGB estimation of Changbai larch (Larix olgensis Henry) of eight plots from three different regions in Maoershan Forest Farm of Heilongjiang, China, using nonlinear mixed effect model with hierarchical Bayesian approach.”

 

Introduction

L31-75: I think the first paragraph is ok, but I have a problem with the references that are used. This first paragraph informs about the importance of forests and the estimation of biomass and carbon stocks, the authors should use the references from a general consideration. Please see Pan Y. et al., 2011. A large and persistent carbon sink in the world’s forests. Science, 333, 988-993.

Response: Thank you for the encouragement and suggestion. We have added the reference of Pan Y. et al., 2011 as follows.

Lines 36-37 on page 1:

“Forest aboveground biomass (AGB) serves as the basis for monitoring and accounting for carbon stock and plays a crucial role in regulating the global carbon balance [1-23].”

Lines 590-591 on page 18:

“3.  Pan, Y.; Birdsey, R.A.; Fang, J.; Houghton, R.; Kauppi, P.E.; Kurz, W.A.; Phillips, O.L.; Shvidenko, A.; Lewis, S.L.; Canadell, J.G.; et al. A Large and Persistent Carbon Sink in the World’s Forests. Science. 2011, 333, 988-993.”

 

L76-78: This sentence showed that tree height and crown measurements come from forest inventory and it is difficult to obtain accurage of tree measurements. However, the authors could add two sentences on tree modelling including tree height and crown allometries in forests that can help to obtain the accurate of tree measurements.

Response: Thank you for the comment and suggestion. We have added the sentences as follows:

Lines 53-55 on page 2:

“Remote sensing-based forest inventory can provide a practical and economical approach to AGB estimation with the help of the allometries of the individual tree parameters such as tree height and crown.”

 

L85-87: There is now a large database of field crown data. Please see Jucker, T., et al. "Tallo: A global tree allometry and crown architecture database." Global change biology (2022). 

Response: Thank you for the comment. We have revised the sentence and added the reference of Jucker, T., et al. (2022) as follow:

Lines 62-66 on page 2:

“Although a tree allometry and crown architecture database on a global scale has been published in a recent study [22], UAV-LiDAR can still conveniently and efficiently provide more accurate crown information than field measurements and ensure the spatial integrity and time consistency of data but lacks tree trunk information [23, 24].”

Lines 628-629 on page 19:

“22.       Jucker, T.; Fischer, F.J.; Chave, J.; Coomes, D.A.; Caspersen, J.; Ali, A.; Loubota Panzou, G.J.; Feldpausch, T.R.; Falster, D.; Usoltsev, V.A.; et al. Tallo: A global tree allometry and crown architecture database. Global Change Biol. 2022, 28, 5254-5268.”

 

L168-178: I think this paragraph is a good addition to this revised version. But, I think this paragraph is not complete, because it lacks some information about the importance of using LiDAR and Bayesian method in biomass estimation, since this is already done using allometric equations.

Response: Thank you for the comment and suggestion. We have added the sentences to emphasize the importance of using LiDAR and Bayesian method in biomass estimation as follow:

Lines 146-149 on page 3:

“However, it is still worth exploring how to obtain individual-tree AGB estimation of larch with low cost, non-destructive samples and high accuracy using multi-platform LiDAR data and hierarchical Bayesian method [51].”

 

Materials and Methods

L209: please reworded "Study area and sampling". In this section, please have two distinctes paragraphes. Please add the sentences of Fig. 1c.

Response: Thank you for the comment and suggestion. We have reworded the title of Section 2.1 as “Study Area and Sampling” and adjusted the structure into two paragraphs and added the sentences of Fig. 1c as follows:

Lines 164-177 on page 4:

“2.1. Study Area and Sampling

The study area is located in the Maoershan Experimental Forest Farm, Shangzhi City, Heilongjiang Province, China, from 127°29’ E to 127°44’ E and 45°14’ N to 45°29’ N (Figure 1a). The slope ranges from 5° to 25°, and the terrain is mountainous, rising from south to north with an average elevation of about 300 m. This region has a temperate continental monsoon climate. Maoershan is typical natural secondary forests in northeastern China surrounded by various broadleaved trees, such as white birch (Betula platyphylla Suk.), Mongolia oak (Quercus mongolica Fisch. ex Ledeb.), and Korean aspen (Populus davidiana), and a few coniferous trees, such as Changbai larch (Larix olgensis Henry), Mongolian pine (Pinus sylvestris var. mongolica Litv.), and Korean pine (Pinus koraiensis Sieb. et Zucc.).

The eight sample plots of 0.09 ha (30×30m) were selected from three larch plantations regions according to different site conditions and stand ages (A1: middle-age forest; A2: near-mature forest; A3: mature forest) (Figure 1b). The normalized UAV-LiDAR and TLS point data of the eight sample plots were shown in Figure 1c.”

 

Figure 1 is updated as follow:

Figure 1. The location of study area: (a) Maoershan Experimental Forest Farm in Heilongjiang Province, P.R. China (China Map Examination No. is GS (2019) 1831) and three larch plantations regions (A1-A3) with varying forest stages; (b) the specific location of eight plots (plot 1-plot 8) in three regions; (c) The normalized UAV-LiDAR and TLS point data of the eight sample plots.

 

L254: Field Inventory data. Here, the authors present the data collection and I think the method of reference AGB estimation could be moved in the data analysis.

Response: Thank you for the comment and suggestion. We have already removed the description of the reference AGB to “2.3.2. Establishment of Individual-tree AGB Model Based on U-T LiDAR” as follows:

Lines 293-295 on page 8:

“The reference AGB of individual tree in this study was calculated using the additive biomass equations of larch plantations proposed by Dong et al. [61], shown in Table A2 of Appendix A.”

 

L299: I had a problem with the Figure 2. For example: the authors showed that the field inventory data is in the objective 1, I do not understand. The authors could better conceive this Figure.

Response: Sorry for the confusion. We have updated Figure 2 in the current version and revised the sentence above Figure 2 as follow:

Lines 222-228 on page 5 and 6:

“Firstly, based on the U-T LiDAR data, two algorithms (CSP and RHCSA) were used to delineate individual trees and obtain the optimal LiDAR-derived individual tree parameters (i.e., DBH, TH and CPA) based on the field inventory data (objective 1). Secondly, five widely used AGB model forms (model I-V) were selected and compared by NLS and Bayesian approach based on the reference AGB of individual tree, and the corresponding mixed-effects LiDAR-AGB model was developed with random effects for different regions to further improve the model fitting (objective 2).

 

Figure 2 is updated as follow:

Figure 2. Flowchart of the proposed methodology in this study.

 

L404: please add the section data analysis that can inclure all analysis according to each study objective.

Response: Thank you for the comment and suggestion. We have introduced methods according to each study objective in the first paragraph of Section 2.3 (see above response or Lines 220-232 on page 5 and 6). To clarify, we added the subsection names in the method section to the corresponding objective in figure 2 (see the updated figure 2 above) and also added some sentences to explain this as follows:

Lines 232-233 on page 6:

“…A flowchart of this study is depicted in Figure 2. Section 2.3.1, 2.3.2, and section 2.3.3 and 2.3.4 explained the procedure for objective 1, 2, and 3, respectively.”

 

Results

L426-427: please add the values of this accurate

Response: Thank you for the suggestion. We have added the values of the accurate as follow:

Lines 363-364 on page 10:

“In addition, the accuracies of DBH estimated by CSP and RHCSA were very similar (R²:0.983 v.s. 0.990, RMSE: 1.017 v.s. 1.024; rRMSE: 4.9 v.s. 4.8).”

 

L440-442: please remove this sentence

Response: Thank you for the suggestion. We have removed the sentence.

 

Please merge Table 6 and Table 7

Response: Thank you for the suggestion. These two tables are merged into the new Table 6 in the current version.

 

Discussion

L527: what do mean the mature method?

Please see Brede, B., Terryn, L., Barbier, N., Bartholomeus, H. M., Bartolo, R., Calders, K., ... & Herold, M. (2022). Non-destructive estimation of individual tree biomass: Allometric models, terrestrial and UAV laser scanning. Remote Sensing of Environment, 280, 113180.

Response: Thank you for suggestion and reference. We agree with the statement "…but estimation of individual tree AGB based on these data has been challenging so far, especially in dense tropical canopies…" in this study (Terryn, et al., 2022). The sentence was revised as follows and the Terryn et al.’s study was cited as the reference [78].

Lines 439-444 on page 14:

“Although it is a widely used method to estimate individual tree AGB using an allometric equation based on field inventory data (e.g., DBH, TH), its efficiency and accuracy are still unsatisfactory, particularly for large-scale forest inventory [78].”

 

Lines 739-741 on page 21:

“78.  Brede, B.; Terryn, L.; Barbier, N.; Bartholomeus, H.M.; Bartolo, R.; Calders, K.; Derroire, G.; Krishna Moorthy, S.M.; Lau, A.; Levick, S.R.; et al. Non-destructive estimation of individual tree biomass: Allometric models, terrestrial and UAV laser scanning. Remote Sens. Environ. 2022, 280, 113180.”

Author Response File: Author Response.docx