Characterizing Tree Spatial Distribution Patterns Using Discrete Aerial Lidar Data

Round 1

Reviewer 1 Report

The paper fills journal aims and deal with theme of great interest for forestry sector. I found the use of second-order statistic (pair corr. fun.) really clever and ALS data used coherent with other experiences in forest environments.
Despite this, some major issues must be solved before publication.

The length of the manuscript is considerable: it makes the reading demanding and the actual comprehension of each step hard. The two study areas (WPA and PC sites) are characterised by simplified forest structure (coniferous forests or mixed, not broadleaves forests) that makes tree detection by ALS easier than in deciduous forests or if understory density is considerable. How the proposed methods (mainly the procedure described in par. 2.3) can be generalised and applied on broadleaves forests or in more complex forests? The method you proposed is based on Li et al. 2012, but: I do not find in the text any clear reference to tree detection rate (have you measured tree position on the ground or only by visual-based measurements?). Have you compared your results with other techniques? See for example lidR package https://cran.r-project.org/web/packages/lidR/lidR.pdf with experience from dalPonte et al. or Silva et al. I do not understand par. 4.1.2 and 4.1.3: they seems additional results. Please clarify the need for these two sections.

 

Additional comments

row 146: "each spatial distribution pattern was parameterized using two different sets of parameters". Please add the parameters you used.

row 157: "For each segmented individual tree points". What do you mean?

row 377: is WPA-LM a broadleaf? In table 1 you have defined this plot as "Mixed". Please clarify.

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The paper is written very well, it is clear and well organized. The subject is relevant and the objectives are clearly explained. Overall, I found it very interesting and easy to read, even if I believe that it is very (probably too) long. I have only two specific comments on the figures.

Fig. 1: please, include the coordinates, the N arrow and a larger bar that explains the height values

Fig. 3: I believe this figure contains too many insets. The font chosen for the names is too small and it is difficult to read. I suggest to decrease the number of insets and increase their dimensons. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

General comments

The manuscript describes the implementation of a method to characterize quantitatively different types of tree spatial distribution patterns using aerial lidar data (ALS).

The rationale of the study is well structured and the paper is well written and balanced.

 

Listed below are a few tips to make some implementation steps clearer. Some additional information and comments on methods and discussion would be useful for readers interested in implementing the selected models in other areas. As an example, a flow-chart will help readers to follow the different tests of the model in the various subsections of the manuscript. As well as a Section of Recommendations for ASL users showing the issues point-by-point would allow readers to easily account for your suggestions.

 

After these simple adjustments, the manuscript will be ready to be published in Remote Sensing.

 

 

Specific comments

 

Introduction: there is no recent literature on the ASL applications for tree distribution patterns, at least on tree structure characterization (Lines 76-86).

Methods: A flow-chart in the methods section will help readers to follow the different subsections of the manuscript. The flow-chart will contain the model implementation as well as the different tests performed.

 

Figure 1: The images of aerial and lidar data are too small to provide an idea of forest characteristics. Please increase the resolution of the images or add zoom-in details.

 

Line 144: Please add some info and characteristics of the adopted tree spatial distribution model.

 

Lines 160-162: Information on how the errors vertical and horizontal slicing are evaluated have to be added.

 

Line 185: Please add a reference for “In this case, the g(r) function suggests a clustered tree spatial distribution pattern across all spatial scales”.

 

Figure 2: Please add also the picture of Rij line segment. If necessary, you can remove the scheme a) being clearly represented also in b).

 

Line 252: The description of the unknown parameter (y) is missing.

 

Figure 4: The circles of segmentation data are not clearly visible in green. Please use a more distinguishable color (e.g. light blue).

 

Tables 3 and 4: These tables can be moved to a Supplementary Material file.

 

Line 528: Please add the observed average crown sizes.

 

Line 550: Please add to the caption the setting values of the other model parameters while the spacing threshold is varying.

 

Tables 5 and 6: These tables can be moved to a Supplementary Material file.

 

Line 550: Please add to the caption the setting values of the other model parameters while the search radius is varying.

 

Line 666: “4.3. Spatiotemporal variations of tree spatial distribution patterns”. There is no assessment of temporal variations. Please modify in “Spatial variations …”

 

Line 740: At the end of the discussion a Section of Recommendations for ASL users showing the issues point-by-point would allow readers to easily account for your results. Alternatively, you can add a table summarizing results/effects and suggestions.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

All the required revisions have been satisfied. Thanks.