Analysis of Current and Future Forest Disturbances Dynamics in Central Europe

Round 1

Reviewer 1 Report

Dear Editors and Authors,

I have carefully read your manuscript entitled "Analysis of Forest Disturbations Dynamics in Central Europe Countries and Future Development Prediction". In my opinion, it is a valuable paper, describing a wide range of forests problems in Central Europe.

• please, consider different keywords than words in the title;
• the literature cited is well selected and up to date;
• lines 104-109 - please, consider summarizing this information on forest ecosystems and citations for them in the table;
• I believe that figures 4-5 are very attractive for potential readers with no experience in using this software;
• in my opinion, the discussion section should be expanded, the work contains a lot of valuable information and it is worth discussing it properly with the available literature data.

Author Response

Dear reviewer,

Thank you for your valuable review.

Responses:

please, consider different keywords than words in the title;

the keywords have been changed

the literature cited is well selected and up to date;

lines 104-109 - please, consider summarizing this information on forest ecosystems and citations for them in the table;

The table with references has been added.

I believe that figures 4-5 are very attractive for potential readers with no experience in using this software;

in my opinion, the discussion section should be expanded, the work contains a lot of valuable information and it is worth discussing it properly with the available literature data.

The discussion section has been improved and reworked.

Author Response File: Author Response.doc

Reviewer 2 Report

The analyzed manuscript has a significant subject about the dynamic of forest distribution in Central Europe Countries. The manuscript needs a major improvement, especially in the Introduction and Discussion.

Some specific comments:

• Lines 64-81: more methodological aspects, a theoretical point of view, which is not suitable here [You have a good basis of references from where to begin the introduction].

Methods: 

• Must have a flowchart with a logical presentation of used methods and data.

Some inconsistencies: In the abstract the period 2020-2022 and at lines 98-99, the period is 1990-2000. Also, for some modeling, the period is 2000-2020. How can be an agreement in the manuscript if there are different periods for different data?

Discussion:

The discussion section is similar to an introduction because is no relation with the results of the manuscript. Only, the last lines could be a suitable part of the discussion.

Author Response

Some specific comments:

• Lines 64-81: more methodological aspects, a theoretical point of view, which is not suitable here [You have a good basis of references from where to begin the introduction].

The introduction is designed concerning the gradual justification of the importance of work. The first part generally characterizes the concept of disturbances and the harmful factors of their occurrence. The next part is devoted to general changes in ecosystems that cause a gradation of the harmful factors of disturbances. The section on lines 64-81 is then devoted to the importance of modeling the future development of disturbances and what methodological approaches to this problem are known so far. In our opinion, therefore, this part belongs exactly to the conclusion of the introduction part of the manuscript, also concerning the declared aim of the work.

Methods: 

• Must have a flowchartwith a logical presentation of used methods and data.

Some inconsistencies: In the abstract the period 2020-2022 and at lines 98-99, the period is 1990-2000. Also, for some modeling, the period is 2000-2020. How can be an agreement in the manuscript if there are different periods for different data?

We tried to resolve this inconsistency by modifying the formulations in the methodological part so that it is clear why only data from the period 2000-2020 were included as background data in the predictive analysis.

Discussion:

The discussion section is similar to an introduction because is no relation with the results of the manuscript. Only, the last lines could be a suitable part of the discussion.

The discussion part has been modified.

Author Response File: Author Response.doc

Reviewer 3 Report

This paper describes two approaches to forecast the occurrence of disturbances in Central European forests, one using the Holt-Winters methods and the other via the growth simulator SIBYLA. To do this, the authors used publicly available data to quantify the amount of “incidental fellings” that occurred between XXX and XXXX. The proportion of incidental fellings in regard to the total harvested wood volume is then used as a proxy for forest disturbances. The behavior of amount of wood obtained from incidental fellings is then described for years XXXX to XXXX, and the 2 forecasting results are presented. The paper presents insightful contributions to the field, but I recommend that it be better structured before it can be considered for publication.

I suggest that the main variable in study (incidental fellings) be better described in the manuscript, mentioning what types of wood can be considered as incidental fellings and how this information is presented in the different sources used by the authors. The main objective of this paper is to empirically model and predict development and intensity of disturbances in the future. However, different types of forest disturbances will have an impact on the amount of wood originating from incidental fellings. For instance, forest disturbances caused by wildfires can go undetected using the methods presented in this work if all the wood is burned and not salvageable. Thus, the strengths and weaknesses of using incidental fellings as a proxy for forest disturbances must be better discussed.

The last phrase before the conclusions state that “Our work should also contribute to strengthening the comparison of predictions through the statistical concept and predictive application.” However, the authors do not offer a suitable comparison between the two approaches used to forecast the amount of wood from incidental fellings presented in their own work. Figures 3 and 6, which present the volume of incidental fellings per country up to 2030 for both forecasting methods, show that the predictions are very different. This however is not pointed out or discussed by the authors, nor is any possible reason for this discrepancy given. Instead, the discussion section simply describes other similar works without any critical comments.

Another flaw that I detected in the paper is that the methods are not properly described. The input parameters used in the SYBILA growth simulator are not described. Figures 4 and 5 are not relevant to the paper, a table describing the information contained in these figures would be more appropriate, along with any additional information that can permit the reproducibility of the results. This information should be inserted in the results methods. There are other sections of the manuscript that should be inserted in the methods section, as shown in the manuscript file.

I also found that the results are lacking. The authors present the mathematical structure of the Holt-Winters method in equations 1-5, but fail to include the estimated values of the model parameters in the results section. Thus, unless these results are included in the manuscript, I don’t see the point in presenting equations 1-5, simply mentioning the source for the Holt-Winters test should be enough.

Given the disparities between the two forecasting approaches presented in this paper, the lack of discussion on this, and the fact that the growth simulator model parameters are poorly described, I suggest that the authors focus on the Holt-Winters method and omit the growth simulator results. The predictions made by the Holt-Winters method seem more plausible than the ones made by the growth simulator. However, more information on the results of the Holt-Winters method should be provided.

The conclusions of the paper read more like discussion topics and should be placed in this section.

Other specific comments on this paper were made directly on the manuscript file.

Comments for author File: Comments.pdf

Author Response

This paper describes two approaches to forecast the occurrence of disturbances in Central European forests, one using the Holt-Winters methods and the other via the growth simulator SIBYLA. To do this, the authors used publicly available data to quantify the amount of “incidental fellings” that occurred between XXX and XXXX. The proportion of incidental fellings in regard to the total harvested wood volume is then used as a proxy for forest disturbances. The behavior of amount of wood obtained from incidental fellings is then described for years XXXX to XXXX, and the 2 forecasting results are presented. The paper presents insightful contributions to the field, but I recommend that it be better structured before it can be considered for publication.

I suggest that the main variable in study (incidental fellings) be better described in the manuscript, mentioning what types of wood can be considered as incidental fellings and how this information is presented in the different sources used by the authors.

In the introductory part, we specified what we consider to be "incidental fellings" in terms of this manuscript. Approaches in different countries are different, and most of what we call incidental fellings in the sense of this paper are considered by the authors to be synonymous with disturbance. This is the felling of wood damaged by harmful factors, which was not planned (intentional) and is realized as an incidental event.

The main objective of this paper is to empirically model and predict development and intensity of disturbances in the future. However, different types of forest disturbances will have an impact on the amount of wood originating from incidental fellings. For instance, forest disturbances caused by wildfires can go undetected using the methods presented in this work if all the wood is burned and not salvageable. Thus, the strengths and weaknesses of using incidental fellings as a proxy for forest disturbances must be better discussed.

In the long term, the volume of wood destroyed by forest fires in Central Europe is not high. Moreover, in Slovakia, for example, there is a problem in accurately reporting and reporting data on this harmful factor, because it is mostly reported in area units (often inaccurate) and it is therefore not possible to use it for verified data in this manuscript. Even if the conversion to the volume of destroyed wood were performed, these data would not be accurate and this could negatively affect the statistical prediction.

The last phrase before the conclusions state that “Our work should also contribute to strengthening the comparison of predictions through the statistical concept and predictive application.” However, the authors do not offer a suitable comparison between the two approaches used to forecast the amount of wood from incidental fellings presented in their own work.

We tried to strengthen the discussion part of the manuscript and remove the claims that were not discussed.

Figures 3 and 6, which present the volume of incidental fellings per country up to 2030 for both forecasting methods, show that the predictions are very different. This however is not pointed out or discussed by the authors, nor is any possible reason for this discrepancy given. Instead, the discussion section simply describes other similar works without any critical comments.

The difference between these results is mainly due to the structure and construction of the models. One is purely statistical (Holt-Winters), the other is derived from real observations and modeling from forest ecosystems in Slovakia and Germany (Sibyla). The detailed construction of the models is described in the cited works, therefore we believe that it is not necessary to elaborate it separately in this manuscript.

Another flaw that I detected in the paper is that the methods are not properly described. The input parameters used in the SYBILA growth simulator are not described. Figures 4 and 5 are not relevant to the paper, a table describing the information contained in these figures would be more appropriate, along with any additional information that can permit the reproducibility of the results. This information should be inserted in the results methods. There are other sections of the manuscript that should be inserted in the methods section, as shown in the manuscript file.

Parts of the methods have been redesigned according to your suggestions. Images from the growth simulator should only declare the principle on which the prediction works. The image from the „Prophet“ module documents that for each regional area the input parameters are different (used areas of Slovakia and Germany). A table with an overview of indicators for each area would mean the need to expand the manuscript by several tens of pages.

I also found that the results are lacking. The authors present the mathematical structure of the Holt-Winters method in equations 1-5, but fail to include the estimated values of the model parameters in the results section. Thus, unless these results are included in the manuscript, I don’t see the point in presenting equations 1-5, simply mentioning the source for the Holt-Winters test should be enough.

These parameters are calculated automatically by the STATISTICA software. The equations are given only as a supplement to the information as the Holt-Winters test is designed.

Given the disparities between the two forecasting approaches presented in this paper, the lack of discussion on this, and the fact that the growth simulator model parameters are poorly described, I suggest that the authors focus on the Holt-Winters method and omit the growth simulator results. The predictions made by the Holt-Winters method seem more plausible than the ones made by the growth simulator. However, more information on the results of the Holt-Winters method should be provided.

Although the results of the Holt-Winters test appear to be more reliable, the results of modeling from a growth simulator are mainly influenced by the negative developments in the recent period. In our opinion, they also provide an interesting view of possible developments, which can only be objectively evaluated over time.

The conclusions of the paper read more like discussion topics and should be placed in this section.

Were changed

Other specific comments on this paper were made directly on the manuscript file.

We've tried to customize all your comments according to your suggestions.

Author Response File: Author Response.doc

Round 2

Reviewer 2 Report

Thank you for your response. The papers' quality is improved. 

Reviewer 3 Report

The manuscript has been improved after its revision. The authors did not accept all the suggestions made in the first round of revisions, however the justification for not doing so were given in the cover letter and were plausible. I only suggest the revision of the following minor points before the paper is ready for publication:

Line 93: Remove extra space.

Line 97: Correct to “Table”

Figure 6: Correct the word "Volume" in the y axis.