Estimation of Aboveground Carbon Stocks in Forests Based on LiDAR and Multispectral Images: A Case Study of Duraer Coniferous Forests

Round 1

Reviewer 1 Report

Even if the manuscript is interesting (from a scientific point of view) and I have enjoyed to read it as forester, i have found some flaws that have to be solved and i've struggled to revise it because line number were missing, several format typos are present ("Error! Reference source not found"), english is not well checked, tables are numbered uncorrectly, species are not written in italic and so on. It seems more an advanced draft than a final paper, and this decrease to me the positive thought that i had of the paper (even if with limits that must be cleared). The fact that lines are missing bring to a general problem to identify "the location" of the requests/revisions i have made.

As a general comment, several times consecutive sentences are ridundant and in order to improve readiness and catch the point they have to be merged providing a clearer and direct point of view instead of general statement. Moreover, check the references in the text, there are several problems.

Please follow paper guidelines and improve the format of the paper: e.g. use italic for scientific tree species name; use m2 as measure unit, there are too many mistakes related to that. You may omit the definition of equation parameters already presented. i.e. AGB presented in eq1 can be omitted after eq. 2; Table 1 is table 5, please check the number in the whole manuscript.

 

Specific comments

Abstract: ”in the field of forestry at home and abroad”, please make more understandable what is home. Maybe change with worldwide.

Introduction. AGB is not an estimation of carbon stored, they are directly related “However, remote sensing can provide accurate”, please change however.

“However, since optical images do not directly provide forest AGC estimates, regression models between field AGC estimates and spectral vegetation indices are re-quired to provide spatially continuous forest AGCs” Please, rephrase, it does not make sense to me.

"Optical methods"? Please, rephrase with optical data

“parametric model was higher than that of the nonparametric model”; if you state that you should present the two models take as reference with related accuracy metrics in order to support your statement.

“LiDAR data do not provide rich spectral information” please rephrase, I do not understand which spectral information may provide.

[27]. The above study, please, add first author name.

Please, increase the quality of figure 1

“DBH were measured several times to find the average value to reduce the error”, how many times, are there relevant errors in DBH measure? there are several ways accepted by the scientific community to avoid errors in dbh  measure.

How many trees are measured? Please provide some measured information also related to different tree species. These information are discussed in 4.2 without data values.

Figure 2. Please explain each boxes.

2.1.3 add the spatial resolution of each band.

2.2., Please present here all variables extracted from point cloud data

“The measured AGC of the forest was calculated from the field measurement data as a reference value.” How is it calculated? Please rephrase and explain clearer the workflow.

2.2.2. Is the presented method used for both UAV-lidar and terrestrial lidar data?

2.2.3. please explain how multispectral metrics are selected.

Please, explain how multispectral data are related to each plot

Table 4, EDVI change as NDVI

3.1. “In this study, we selected relevant and accurate independent variables to avoid the effects of covariance and poor correlation among multiple variables for fitting the prediction model”. How?

Please, present in the text the model tested such as: a DBH^b, a + DBH b, etc. then add the parameters calculated. It is enough move the table after the next paragraph.

Figure 6. Please, provide a graph with square dimension and add the line x=y

Figure 2, please, check figure number, and make the same of previous scatter plot. Moreover, is the map referred to a particular field plot? The dimension is 20x40 more or less, while the inventory data were 10x40. Do you provide a map for the whole study area?

Please use the same legend for both distribution maps

To prove the effectiveness of backpack lidar some timing comparison with traditional field measure are needed. Moreover, some comparison in terms of accuracy per measured variables.

Our results further suggest that combining LiDAR and hyperspectral data is essential to improve the accuracy of AGB accuracy and AGC estimation. Which Hyperspectral data?

Why are not tested non parametric methods. i.e. random forest algorithm can be trained with all available metrics.

Finally, the paper can not be accepted for publication without major revision related to solve the above mentioned questions/suggestions/needs and without a very detailed improvement of the document that take into account the fact it must be a final paper ready for the readers and not a draft to be checked by a "draft corrector".

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors presented a process for verifying LiDAR and multispectral data for assessing carbon (C) accumulation. Critically evaluated the combined data to estimate AGC, LiDAR data are both active and include spectral characteristics of optical images.

The dependence of the AGC method pd structural features of stands that can be combined with LiDAR data is indicated. The levels of obtained carbon coefficients, which can be reflected in multispectral information, are discussed.

 Therefore, both presented LiDAR data and multispectral data, the fusion of LiDAR with multispectral data can be a good method for estimating forest AGC in the future.

This study can be a valuable resource for scientists and forestry workers to obtain more accurate AGC data.

Errors due to the quality of the measurement equipment used ( including GPS signal readings) during data acquisition by portable LiDAR can directly affect data quality. This leads to misinterpretation of results with significant absolute coordinates.

 In addition, it is not indicated what factors affect the accuracy of point cloud acquisition.  This affects the precision of absolute coordinates in the description of individual trees. At the same time, how does the compilation of results over the full sample area of the LiDAR survey translate? Therefore, how to effectively and accurately acquire data from field surveys.

The vegetation cover index should significantly affect the assessment of the AGC (kg C) index. The authors rightly pointed out the need to consider the horizontal structure of forests on possible regression prediction models with the studied LIDAR parameters for predicting the severity of biomass in the field.

 

Multivariate regression is a deceptive technique, just "include" enough variables, and usually for some of them will turn out to be significant.

A paper with a supported structure, including full methodology and presentation of results, lacks reference to the source data of the processed research results. How does species structure affect the assessment of changes in AGC prediction?

 

Additional Notes:

In the body of the article, there were some indications of errors in the reference citation "Error! Reference source not Found"

Frequent errors of writing "R2 ="

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

I want to thank the authors because they have improved their paper very much, taking into account almost all the suggestions and answering deeply every questions that have been made. Nevertheless, as a researcher, I do not like the "superficiality" about typos, mistakes, grammar, when a paper is submitted. The paper is not ready for a publication and I require minor revisions because some grammar mistakes are still there (f.i. Table 4: vegetation indices not index), formatting (there are, at least, 3 different fonts within the paper, just take a look to table, or authors affiliation), and generally (typos, missing space, wrong link "Error! Reference source not found") are still present within the whole paper (e.g. line 6 - missing space, line 10 - paragraph spaces; line 31 remove dot; table 1; line 183; figure 4; etc.). I am sorry but the paper can not be accepted like that.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf