Examining Changes in Stem Taper and Volume Growth with Two-Date 3D Point Clouds
, Ninni Saarinen 1,2,3, Ville Kankare 1,2,3, Topi Tanhuanpää 1,2,4, Harri Kaartinen 5,6, Antero Kukko 2,5, Markus Holopainen 1,2, Juha Hyyppä 2,5 and Mikko Vastaranta 2,3Round 1
Reviewer 1 Report
This study demonstrated the capability of TLS data for characterizing individual tree growth but also changes in stem form through simple attributes describing stem form. 35 trees from 4 sample plots were measured in two different time points to investigate whether measuring the change in stem volume and other stem attributes is possible by using two-date TLS-data. Measurement results for the sample tree species, DBH, h, Vol, TAP, f, q0.5, and HDR from both T1 and T2 are computed, and for all the stem attributes the absolute and relative change between T1 and T2 was calculated. Stem volume growth was identified with TLS data acquired in two points of time. Except for the rather clear change in stem volume, there was no clear trends in the other attributes describing the stem form of the sample trees.
More detailed comments are listed below:
It is indicated the TLS is high resolution settings, such as ~10 000 pts/m2 and 6.3 mm spacing at 10 m distance from the scanner. How is other indicators for example the range accuracy?
As shown in table 4, the standard deviation is larger than mean value for all cases, what is reason for this results?
How do you evaluate it comparing with traditional measurement methods?
Line 87, what is the meaning of ‘an important aspect per se’?
Line 217, space is needed before ‘The difference’. The same for line 236, where space is needed.
It is hinted that pairwise Student’s t-test gave strong evidence (p-value 0.0001) against the null hypothesis. It is better to give more details how you carry out the tests.
Author Response
REVIEW 1
Open Review
(x) I would not like to sign my review report
( ) I would like to sign my review report
English language and style
( ) Extensive editing of English language and style required
( ) Moderate English changes required
(x) English language and style are fine/minor spell check required
( ) I don't feel qualified to judge about the English language and style
Yes | Can be improved | Must be improved | Not applicable | |
Does the introduction provide sufficient background and include all relevant references? | (x) | ( ) | ( ) | ( ) |
Is the research design appropriate? | ( ) | (x) | ( ) | ( ) |
Are the methods adequately described? | ( ) | (x) | ( ) | ( ) |
Are the results clearly presented? | ( ) | (x) | ( ) | ( ) |
Are the conclusions supported by the results? | (x) | ( ) | ( ) | ( ) |
Comments and Suggestions for Authors
This study demonstrated the capability of TLS data for characterizing individual tree growth but also changes in stem form through simple attributes describing stem form. 35 trees from 4 sample plots were measured in two different time points to investigate whether measuring the change in stem volume and other stem attributes is possible by using two-date TLS-data. Measurement results for the sample tree species, DBH, h, Vol, TAP, f, q0.5, and HDR from both T1 and T2 are computed, and for all the stem attributes the absolute and relative change between T1 and T2 was calculated. Stem volume growth was identified with TLS data acquired in two points of time. Except for the rather clear change in stem volume, there was no clear trends in the other attributes describing the stem form of the sample trees.
A: Thank you for your careful review of our manuscript and the valuable comments regarding it. We have tried to further improve the manuscript based on the received comments. Our responses to the more detailed comments are presented separately after each comment. The new version of the manuscript has also been uploaded to the submission system.
In addition to changes made based on the comments below, we have edited the manuscript after receiving comments from another reviewer. The most notable edit is related to section Results. It was suggested, that Table 2 and Table 3 from the original manuscript would be moved to Appendix/Supplementary materials section. This edit was done as suggested and Tables 2 and 3 can now be found from supplementary materials entitled Table S1 and Table S2.
More detailed comments are listed below:
R: It is indicated the TLS is high resolution settings, such as ~10 000 pts/m2 and 6.3 mm spacing at 10 m distance from the scanner. How is other indicators for example the range accuracy?
A: Further information about the TLS scanner was added to the text “The positional accuracy of the scanner was 5 mm on single measurement on range from 1 to 25 meters.”
Now in the manuscript, Lines 146-149: “For TLS, phase-shift scanner Leica HSD6100 (Leica Geosystems AG, Heerbrugg, Switzerland; now Hexagon AB, Stockholm, Sweden) was used with high resolution settings (i.e., ~10 000 pts/m2 and 6.3 mm spacing at 10 m distance from the scanner). The positional accuracy of the scanner was 5 mm on single measurement on range from 1 to 25 meters.”
R: As shown in table 4, the standard deviation is larger than mean value for all cases, what is reason for this results?
A: The reason behind these results is the large variation in the size of the sample trees. In the data there are few large trees, whose volume growth is multiple times more than the volume growth of the smaller trees, which leads to the standard deviation of stem volume growth to be larger than the average stem volume growth of all the sample trees. For other attributes describing the stem form, the change within the investigation period could be either positive or negative, which leads to that the mean is calculated from both positive and negative values, whereas in standard deviation only the amount of change is considered.
R: How do you evaluate it comparing with traditional measurement methods?
A: Previous research [1-4] has shown that TLS can perform on similar accuracy level as the traditional measurement methods considering diameter measurements and both measurement approaches have their limitations [1-4]. With traditional measurement methods, it would also be possible to measure stem volume growth for example, but instead of using several measurements from along a standing tree stem (TLS), the stem volume would be estimated based on taper or volume models [5] using attributes such as tree species, DBH, D6 and height. However, the use of models often generalizes the estimates depending on the data and attributes used to create the model and wouldn’t probably give as detailed description (i.e. measured stem diameter variation along the stem) of the stem form although the overall growth in stem volume and form could be detected.
1. Liang, X.; Kankare, V.; Hyyppä, J.; Wang, Y.; Kukko, A.; Haggrén, H.; Yu, X.; Kaartinen, H.; Jaakkola, A.; Guan, F. Terrestrial laser scanning in forest inventories. Isprs J Photogramm 2016, 115, 63-77.
2. Newnham, G.J.; Armston, J.D.; Calders, K.; Disney, M.I.; Lovell, J.L.; Schaaf, C.B.; Strahler, A.H.; Danson, F.M. Terrestrial laser scanning for plot-scale forest measurement. Current Forestry Reports 2015, 1, 239-251.
3. Wilkes, P.; Lau, A.; Disney, M.; Calders; K., Burt; A., Gonzalez de Tanago; J., Bartholomeus; H., Brede, B.; Herold, M. Data acquisition considerations for Terrestrial Laser Scanning of forest plots. Remote Sensing of Environment 2017, 196, 140-153.
4. Luoma, V.; Saarinen, N.; Wulder, M.A.; White, J.C.; Vastaranta, M.; Holopainen, M.; Hyyppä, J. Assessing precision in conventional field measurements of individual tree attributes. Forests 2017, 8, 38.
5. Laasasenaho, J. Taper curve and volume functions for pine, spruce and birch. The Finnish Forest Research Institute, Helsinki, 1982.
R: Line 87, what is the meaning of ‘an important aspect per se’?
A: Edited the text by changing “per se” to “as such”
Now in the manuscript on line 87: Maintaining trees undamaged can be seen as an important aspect as such.
R: Line 217, space is needed before ‘The difference’. The same for line 236, where space is needed.
A: Both corrected as suggested.
R: It is hinted that pairwise Student’s t-test gave strong evidence (p-value 0.0001) against the null hypothesis. It is better to give more details how you carry out the tests.
A: The pairwise Student’s t-tests were performed for stem volume and four other attributes describing the stem form. The t-tests were performed by comparing estimates from T1 and T2. R-software was used to calculate the results. Based on a suggestion of the second reviewer the t-test results of TAP, f, q0.5 and HDR were added to the study. The t-tests are described and reported as follows:
Now in the manuscript starting at line 217: The difference in stem volume estimates was evaluated between T1 and T2. Statistical significance of the detected change was evaluated using Student’s pairwise t-test. For stem volume, the null hypothesis stated that the method wouldn’t be able to detect tree growth within the period between T1 and T2. For the other stem attributes, Student’s pairwise t-test was used to evaluate whether a statistically significant change had occurred in TAP, f, q0.5 or HDR. For these attributes the null hypothesis was formulated “No change in stem form attribute has occurred within the period between T1 and T2”.
Now in the manuscript starting at line 260: The relative growth in Vol varied from 4.3% to 240.3% during the study period. Also the results of pairwise Student’s t-test gave strong evidence (p-value 0.0001) against the null hypothesis, which means that the method was able to detect stem volume growth within the nine-year study period. In addition, Table 2 presents the pairwise Student’s t-test results for evaluating the null hypothesis “No change in stem form attribute has occurred within the period between T1 and T2” in TAP, f, q0.5 and HDR. For both TAP and q0.5 null hypothesis could be rejected with p-values 0.0023 and 0.0081, respectively. Whereas for f and HDR the null hypothesis stood with p-values of 0.0589 and 0.7606, respectively. Based on this, change in TAP and q0.5 did happen during the study period, whereas the stem form didn’t change significantly when considering relationship between h and DBH (HDR) or Vol in relation to basal area and h (f).
Table 2. The p-values of Student’s pairwise t-test, where it was tested whether the stem form attributes have changed during the study period. The null hypothesis stated that no change in stem form attribute has occurred within the period between T1 and T2. For tapering (TAP) and normal form quotient (q0.5) the null hypothesis could be rejected and for cylindrical form factor (f) and height to diameter ratio (HDR) it was upheld.
TAP | f | q0.5 | HDR | |
p-value | 0.0023 | 0.0589 | 0.0081 | 0.7606 |
Now in the manuscript starting at line 337: Except for the rather clear change in stem volume, there was no equally clear trends in other attributes describing the stem form of sample trees. However, according to Student’s pairwise t-test, statistically significant change could be detected for both TAP and q0.5 whereas similar change in HDR or f was not observed. On average, TAP decreased for the sample trees. When investigating the species specific change, it was notable that TAP decreased for coniferous species but increased for broadleaved sample trees. The largest decrease in TAP was detected for Scots pine trees. However, it needs to be pointed out, that all Scots pine sample trees were relatively small and still in a younger development phase when compared to other sample trees. To be able to draw further conclusions about the change in tapering during a lifetime of a tree, longer follow-up period with several measurements or more sample trees from various size and age classes would be required. Using TLS-based high-quality point clouds could be an option for re-considering the definition and measurement of stem tapering. For example, with the ability of observing the diameter at any height, tapering could be measured by comparing diameters from other heights than the traditional 1.3 m and 6.0 m.
On average, the values of q0.5 increased slightly. As DBH is in a divisor in the formula of q0.5, it can be stated that diameter above the 1.3 meter height (i.e., for q0.5 the height corresponding 50% of the total tree height) had increased relatively more than the DBH. Also, if the absolute increase in d6.0 is larger than in DBH, it causes TAP to decrease. Despite that small changes were detected also in f and HDR they were not statistically significant. Even though it can be stated that changes in stem form do happen, it is possible, that no significant changes in HDR or f should even be detected for all trees during the study period. This is because until a certain point of development, both DBH and h increase as the tree grows [1]. However, it also needs to be noticed, that the few exceptional observations in each TAP, f, q0.5 and HDR could be caused by trees allocating their growth differently. For example, location, the effect of surrounding trees, and other factors affecting the growth can lead to differences in increment of DBH and h [12-14]. Based on results in this study, further research on the effects of different growth conditions and development phases on stem form is needed to cover this issue more in detail.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The manuscript "Examining changes in stem taper and volume growth with two-date 3D-point clouds" investigates:
The ability of terrestrial laser scanners (TLS) to measure tree growth
Analyzes changes of taper parameters, derived from TLS measurements, at the tree level.
Advancing the knowledge on these topics is relevant for the forest research community. In general, the manuscript gives a good description of the proposed methodology and results and conclusions are supported by the empirical data analyzed in the study. However, I believe the manuscript needs to be revised. The quality of the collected data is very good but I think the analysis presented in the manuscript is somehow limited and needs to be extended.
I believe the manuscript is too focus on reporting tree specific measurements, while other more general results are somehow obscured by the amount of information provided for the sampled trees, or not considered. I think, tables 2 and 3, should be moved to an appendix. And more weight should be put in statistical analyses like the one presented for Volume. I believe, presenting a similar analysis for the taper indexes can be of interest, probably even including the species as a factor to analyze. Also, at least for HDR it could be possible to compare TLS-HDR vs caliper & hypsometer-HDR measurements. Maybe even fitting taper models and making comparisons by species.
As I mentioned, I think the quality of the data is very good but the analysis is limited. I believe the authors have a great dataset and a good experiment design, in my opinion the statistical if analysis is extended, this manuscript will make a great contribution to the forest research literature.
Author Response
REVIEW 2
Open Review
(x) I would not like to sign my review report
( ) I would like to sign my review report
English language and style
( ) Extensive editing of English language and style required
( ) Moderate English changes required
( ) English language and style are fine/minor spell check required
(x) I don't feel qualified to judge about the English language and style
Yes | Can be improved | Must be improved | Not applicable | |
Does the introduction provide sufficient background and include all relevant references? | (x) | ( ) | ( ) | ( ) |
Is the research design appropriate? | ( ) | (x) | ( ) | ( ) |
Are the methods adequately described? | ( ) | (x) | ( ) | ( ) |
Are the results clearly presented? | (x) | ( ) | ( ) | ( ) |
Are the conclusions supported by the results? | (x) | ( ) | ( ) | ( ) |
Comments and Suggestions for Authors
The manuscript "Examining changes in stem taper and volume growth with two-date 3D-point clouds" investigates:
The ability of terrestrial laser scanners (TLS) to measure tree growth
Analyzes changes of taper parameters, derived from TLS measurements, at the tree level.
Advancing the knowledge on these topics is relevant for the forest research community. In general, the manuscript gives a good description of the proposed methodology and results and conclusions are supported by the empirical data analyzed in the study. However, I believe the manuscript needs to be revised. The quality of the collected data is very good but I think the analysis presented in the manuscript is somehow limited and needs to be extended.
A: Thank you for your careful review of our manuscript and the valuable comments regarding it. We have tried to further improve the manuscript based on the received comments. We did split the received review in several parts during the editing process and tried to give response on all the commented topics. Our responses and edits are presented separately after each comment. The new version of the manuscript has also been uploaded to the submission system.
R: I believe the manuscript is too focus on reporting tree specific measurements, while other more general results are somehow obscured by the amount of information provided for the sampled trees, or not considered. I think, tables 2 and 3, should be moved to an appendix.
A: The Results chapter is now modified and Tables 2 and 3 have been moved to section Supplementary materials and named Table S1 and Table S2. Because of the objective of this study was to evaluate changes in stem volume and stem taper of individual trees within nine-year-long time- period, both Table S1 and Table S2 are still included in the end of the manuscript under the heading Supplementary materials. We will also contact the journal editor to receive further instructions whether an appendix section could be used in this manuscript or will the tables be moved to an external supplementary materials document. If using supplementary materials is suggested by the editor tables will be uploaded separately in the submission phase. In addition, further analysis was included as requested in the next comment by the reviewer (more details in the following reply).
R: And more weight should be put in statistical analyses like the one presented for Volume. I believe, presenting a similar analysis for the taper indexes can be of interest, probably even including the species as a factor to analyze.
A: We have added to the manuscript similar statistical analysis for taper indexes which were presented for stem volume. A Student’s pairwise t-test was performed for all four other stem form attributes and the results are reported in the study. For these attributes, the null hypothesis of the t-test was however different than for the stem volume. With stem volume the null hypothesis was that the method wouldn’t be able to detect tree growth within the period between T1 and T2 but for the other stem attributes, Student’s pairwise t-test was used to evaluate whether a statistically significant change had occurred in TAP, f, q0.5 or HDR. For these attributes the null hypothesis was formulated “No change in stem form attribute has occurred within the period between T1 and T2”. This was due to the relatively large variation in size of the sample trees as well as them belonging to several tree species and with still a little knowledge about how these attributes would behave during the study period of 9 years. And, based on the previous knowledge, it could be expected that if the growth of the trees stays regular, no changes in some of the stem form attributes (i.e. HDR and f) would happen.
Our long term aim is to be able to concentrate on investigating the behavior of these stem form attributes after this study has been used as our first step of using TLS in tree growth and change detection. This is needed, because there is still not enough knowledge on how the growth and stem form of the trees reacts to different growth environments and/or disturbances in different parts of their rotation time. To be able to do this, larger data sets of sample trees from same tree species will be used.
Related to this editing, changes have been made in the manuscript in following chapters:
2.5 Evaluating the method’s suitability for detecting change in stem volume and stem taper
Now in the manuscript starting at line 225: The difference in stem volume estimates was evaluated between T1 and T2. Statistical significance of the detected change was evaluated using Student’s pairwise t-test. For stem volume, the null hypothesis stated that the method wouldn’t be able to detect tree growth within the period between T1 and T2. For the other stem attributes, Student’s pairwise t-test was used to evaluate whether a statistically significant change had occurred in TAP, f, q0.5 or HDR. For these attributes the null hypothesis was formulated “No change in stem form attribute has occurred within the period between T1 and T2”. In addition to statistical analysis, the change between T1 and T2 was evaluated by investigating the amount of the change among the sample trees. It was also studied if the potential change was similar for all tree sizes and species or could some differences be found.
3. Results
Now in the manuscript starting at line 260: The relative growth in Vol varied from 4.3% to 240.3% during the study period. Also the results of pairwise Student’s t-test gave strong evidence (p-value 0.0001) against the null hypothesis, which means that the method was able to detect stem volume growth within the nine-year study period. In addition, Table 2 presents the pairwise Student’s t-test results for evaluating the null hypothesis “No change in stem form attribute has occurred within the period between T1 and T2” in TAP, f, q0.5 and HDR. For both TAP and q0.5 null hypothesis could be rejected with p-values 0.0023 and 0.0081, respectively. Whereas for f and HDR the null hypothesis stood with p-values of 0.0589 and 0.7606, respectively. Based on this, change in TAP and q0.5 did happen during the study period, whereas the stem form didn’t change significantly when considering relationship between h and DBH (HDR) or Vol in relation to basal area and h (f).
Table 2. The p-values of Student’s pairwise t-test, where it was tested whether the stem form attributes have changed during the study period. The null hypothesis stated that no change in stem form attribute has occurred within the period between T1 and T2. For tapering (TAP) and normal form quotient (q0.5) the null hypothesis could be rejected and for cylindrical form factor (f) and height to diameter ratio (HDR) it was upheld.
TAP | f | q0.5 | HDR | |
p-value | 0.0023 | 0.0589 | 0.0081 | 0.7606 |
4. Discussion
Now in the manuscript starting at line 337: Except for the rather clear change in stem volume, there was no equally clear trends in other attributes describing the stem form of sample trees. However, according to Student’s pairwise t-test, statistically significant change could be detected for both TAP and q0.5 whereas similar change in HDR or f was not observed. On average, TAP decreased for the sample trees. When investigating the species specific change, it was notable that TAP decreased for coniferous species but increased for broadleaved sample trees. The largest decrease in TAP was detected for Scots pine trees. However, it needs to be pointed out, that all Scots pine sample trees were relatively small and still in a younger development phase when compared to other sample trees. To be able to draw further conclusions about the change in tapering during a lifetime of a tree, longer follow-up period with several measurements or more sample trees from various size and age classes would be required. Using TLS-based high-quality point clouds could be an option for re-considering the definition and measurement of stem tapering. For example, with the ability of observing the diameter at any height, tapering could be measured by comparing diameters from other heights than the traditional 1.3 m and 6.0 m.
On average, the values of q0.5 increased slightly. As DBH is in a divisor in the formula of q0.5, it can be stated that diameter above the 1.3 meter height (i.e., for q0.5 the height corresponding 50% of the total tree height) had increased relatively more than the DBH. Also, if the absolute increase in d6.0 is larger than in DBH, it causes TAP to decrease. Despite that small changes were detected also in f and HDR they were not statistically significant. Even though it can be stated that changes in stem form do happen, it is possible, that no significant changes in HDR or f should even be detected for all trees during the study period. This is because until a certain point of development, both DBH and h increase as the tree grows [1]. However, it also needs to be noticed, that the few exceptional observations in each TAP, f, q0.5 and HDR could be caused by trees allocating their growth differently. For example, location, the effect of surrounding trees, and other factors affecting the growth can lead to differences in increment of DBH and h [12-14]. Based on results in this study, further research on the effects of different growth conditions and development phases on stem form is needed to cover this issue more in detail.
R: Also, at least for HDR it could be possible to compare TLS-HDR vs caliper & hypsometer-HDR measurements.
A: During the study, we compared the TLS-based HDR with traditional measurement methods. The results of the analysis showed that the change was very minimal on both methods and there was almost no difference between them. For TLS-based HDR, the mean change was 0.01 and std.dev 0.10 and for HDR measured with traditional methods, the mean change was -0.001 and std.dev 0.10.
Also, earlier research [1-4] has shown that TLS is able to reach the same level of accuracy in DBH measurements as the traditional measurement methods. That is why this comparison was left away from this manuscript.
1. Liang, X.; Kankare, V.; Hyyppä, J.; Wang, Y.; Kukko, A.; Haggrén, H.; Yu, X.; Kaartinen, H.; Jaakkola, A.; Guan, F. Terrestrial laser scanning in forest inventories. Isprs J Photogramm 2016, 115, 63-77.
2. Newnham, G.J.; Armston, J.D.; Calders, K.; Disney, M.I.; Lovell, J.L.; Schaaf, C.B.; Strahler, A.H.; Danson, F.M. Terrestrial laser scanning for plot-scale forest measurement. Current Forestry Reports 2015, 1, 239-251.
3. Wilkes, P.; Lau, A.; Disney, M.; Calders; K., Burt; A., Gonzalez de Tanago; J., Bartholomeus; H., Brede, B.; Herold, M. Data acquisition considerations for Terrestrial Laser Scanning of forest plots. Remote Sensing of Environment 2017, 196, 140-153.
4. Luoma, V.; Saarinen, N.; Wulder, M.A.; White, J.C.; Vastaranta, M.; Holopainen, M.; Hyyppä, J. Assessing precision in conventional field measurements of individual tree attributes. Forests 2017, 8, 38.
R: Maybe even fitting taper models and making comparisons by species.
A: Our intention is to concentrate on creating and fitting taper models for several tree species in the near future and at the same time investigate the effects of silviculture and development phase of the trees on tree growth and stem form. The findings of this study will be used in support of the future work. Based on our previous experiences from tree stem taper modeling and its requirements, we felt that including the creation of taper models to the current study is not feasible. And it would have required slightly different data acquisition design with more variability and larger number of sample trees to do comprehensive evaluation.
R: As I mentioned, I think the quality of the data is very good but the analysis is limited. I believe the authors have a great dataset and a good experiment design, in my opinion the statistical if analysis is extended, this manuscript will make a great contribution to the forest research literature.
A: Thank you for the constructive comments to improve our manuscript. We have tried to widen the analysis and improve the manuscript in general following the suggestions given in the review process.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The manuscript has been revised according to comments.
Reviewer 2 Report
The authors made the suggested modifications.
