The Signature of Climate in Annual Burned Area in Portugal

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. The introductory section (lines 23-33) provides a detailed background on wildfire activity in Portugal, including historical data and the impact of extreme years. However, there is a need to ensure that all literature cited is up-to-date and covers all key research in the field. The authors are advised to check that the cited literature includes the most recent research and that there is a need to add more references to support the background information.

 

2. The article proposes the use of 44 years of data (1980-2023) to study the distribution of annual area burned, which is a reasonable time span to reveal long-term trends and cyclical changes. It is suggested that the specific steps for data collection and analysis be further detailed in the Methods section. Specific information is needed on where the data on the area of forests burned in Portugal come from.

 

3 The methods section (lines 104-165) describes in detail the datasets and statistical models used. However, it is recommended that the authors provide more details on the data pre-processing, model selection and validation process to ensure transparency and reproducibility of the methodology.

 

4 The results section (lines 171-300) presents the results of the time series analysis and model fitting through graphs and tables. These results clearly demonstrate the distribution and variability of annual burned area. It is recommended that the authors ensure that all graphs and tables have clear titles and detailed descriptions to make them easier for the reader to understand.

 

5 The concluding section (lines 392-453) provides a discussion based on the results and presents the impacts of climate variability and climate change on wildfire activity in Portugal. The conclusion seems to be closely related to the results, but it is recommended that the author further discusses the implications of the results for existing theory and practice.

 

6. The overall quality of the English language in the article is good, but there are still some minor grammatical and spelling errors. The authors are advised to proofread the article carefully to ensure its professionalism.

 

7. The article presents a novel model that relates the distribution of annual burned area to climatic conditions. The statistical method of combining climate data and wildfire activity is less innovative, and it is recommended that artificial intelligence-related methods be introduced to explore the factors influencing wildfire activity as a result of climate change.

 

8. The content of the study has important implications for understanding and predicting wildfire activity, especially in the context of climate change. This is essential for developing effective fire management and prevention strategies.

 

9. The methodology and conclusions of the study appear to be scientifically sound. However, the authors are advised to further discuss possible limitations and directions for future research.

 

10. The overall strength of the article lies in its in-depth analysis of wildfire activity in Portugal and its exploration of the impacts of climate change. The study provides valuable insights and may have a positive impact on fire management practices.

 

Comments on the Quality of English Language

The overall quality of the English language in the article is good, but there are still some minor grammatical and spelling errors. The authors are advised to proofread the article carefully to ensure its professionalism.

Author Response

1. The introductory section (lines 23-33) provides a detailed background on wildfire activity in Portugal, including historical data and the impact of extreme years. However, there is a need to ensure that all literature cited is up-to-date and covers all key research in the field. The authors are advised to check that the cited literature includes the most recent research and that there is a need to add more references to support the background information.
   Answer:
   This is a relevant issue. References in the introductory section encompass a wide spectrum of subjects related to the topic of the paper, namely atmospheric conditions and wildfire activity [2-12], extreme fire events [13-20], fire-land-atmosphere relationships [21-25], remote sensing applied to retrieval of land surface parameters, monitoring of wildfire activity and discrimination of burned areas [26-42], operational services oriented to wildfire prevention and firefighting [43-52], indices of fire danger [53,54] and impact of climate change [55-59]. Most of these references (43 papers) were cited in the last 3 years (37 papers in 2024, 3 in 2023 and 3 in 2022). One of the papers [21] has currently 521 citations, another one [3] has currently 432 citations, 11 have more than 100 citations, 6 have citations between 50 and 99, and 9 have citations between 25 and 49. I do believe that the references listed, together with the references therein, provide a faithful overview of wildfire activity and related issues involved, in particular fire weather, climate variability and climate change.
   There is however an additional point that helps putting in perspective the unusually list of papers provided in the Introduction. Most of the papers listed have my co-authorship, and this relates to the peculiar nature of this paper. As explained in the Acknowledgements, this paper resulted from the ‘2023 Academic Career Award in Meteorology’ that I received from the Portuguese Association of Meteorology and Geophysics (APMG). The prize included an offer by MDPI to publish an article in Climate. Given the circumstances, I chose the topic not only because of the novelty of the work developed, but also because the topic offered room for revisiting my career by citing, in a harmonious way, the papers I consider as more representative, and especially for paying tribute to colleagues and students that significantly contributed to make up my career. In order to reinforce this sensible point, I have decided to address it in the last paragraph of the Introduction section.
   
   
2. The article proposes the use of 44 years of data (1980-2023) to study the distribution of annual area burned, which is a reasonable time span to reveal long-term trends and cyclical changes. It is suggested that the specific steps for data collection and analysis be further detailed in the Methods section. Specific information is needed on where the data on the area of forests burned in Portugal come from.
   Answer:
   Details about the Portuguese Rural Fire Database were included after the end of the sentence in line 111 of the original manuscript:
   “Statistical information about wildfire activity at the regional and national levels together with records of individual fire events for the period 1980–2023 are also available at the ICNF site [60]. Under Portuguese law, firefighters are the source of the information about each rural fire occurring in Portugal. During the period 1980-2000 location and burned area verification was done by the Portuguese “Forest Guards”, and, after 2001, the verification procedure was improved by introducing SGIF, the Information Management System for Forest Fires [61]”.
   
3. The methods section (lines 104-165) describes in detail the datasets and statistical models used. However, it is recommended that the authors provide more details on the data pre-processing, model selection and validation process to ensure transparency and reproducibility of the methodology.
   Answer:
   Lines 119-125 of the original manuscript were rewritten in order to provide more details about pre-processing:
   “Daily values of DSR are defined in a 0.5°×0.5° regular lat-lon grid, with 150 grid points covering the Portuguese territory. In this study, for each day of the period 1980–2023, averages were computed of DSR on grid points over Portugal. Then, for each year, the Cumulated Daily Severity Rate (CDSR) was obtained by adding the daily spatial averages of DSR from January 1 to December 31. For each year, daily spatial averages of DSR over Portugal were also accumulated from January 1 up to each day of the year; the notation CDSR(d) will be used to refer to the cumulative DSR from 1 January to day d”.
   The following two paragraphs were added at the end of the Data and Methods section (line 170 of the original manuscript) in order to describe the fitting of the statistical model of BA and its usage to assess the impacts of climate variability and climate change on BA.
   “The statistical model of BA is derived in two steps. First, it is shown that, without the three largest values, the sample of BA covering the period 1980–2023 follows a Rayleigh distribution with constant scale parameter σ; this is done by showing that the estimates of MMR and CV of the sample are very close to the theoretical values for a Rayleigh distribution as given by (5) and (9). In the second step, it is shown that the entire sample follows a Rayleigh distribution whose logarithm of σ linearly depends on CDSR. This is done by using quantile regression to show that, in accordance with (13a), percentiles 25 and 75 of the logarithm of BA depend linearly on CDSR with virtually the same slope; then, using quantile and least squares regressions it is shown that, in accordance with (8) and (13b), percentile 54 and the mean of the logarithm of BA linearly depend on CDSR, with the average of the respective slopes virtually coinciding with the common slope estimated for percentiles 25 and 75.
   The fitted model is finally used to assess the impacts of climate variability and climate change on BA. The impact of climate variability is characterized by computing the contribution of the time series of the annual means of the fitted model to the total variance of the time series of BA. The impact of climate change is in turn assessed by comparing the return period of several thresholds of annual burned area for two subperiods of 1980-2023, covering respectively the first thirty years (1980-2009) and the last thirty years (1994-2023)”.
   
4. The results section (lines 171-300) presents the results of the time series analysis and model fitting through graphs and tables. These results clearly demonstrate the distribution and variability of annual burned area. It is recommended that the authors ensure that all graphs and tables have clear titles and detailed descriptions to make them easier for the reader to understand.
   Answer:
   Figures and tables, and respective captions, were revised as suggested by the reviewer.
   
   
5. The concluding section (lines 392-453) provides a discussion based on the results and presents the impacts of climate variability and climate change on wildfire activity in Portugal. The conclusion seems to be closely related to the results, but it is recommended that the author further discusses the implications of the results for existing theory and practice.
   Answer:
   The following two sentences were added at the end of the Discussion section (line 453 of the original manuscript):
   “In this regard, statistical models as the one proposed in this work can be operationally applied to produce outlooks of annual burned area in Portugal by using CDSR as derived from meteorological information provided by seasonal forecasts that are currently disseminated by a variety of weather services like the European Centre for Medium-Range Weather Forecasts (ECMWF), the US National Centers for Environmental Protection (NCEP), and national centers such as the Met Office in the United Kingdom, Météo-France in France, the Deutscher Wetterdienst (DWD) in Germany and the Centro Euro-Mediterraneo sui Cambiamenti Climatici (CMCCl) in Italy. In the same line, the statistical model of BA can also be used to generate synthetic time series of BA in future scenarios of climate based on information provided by climate models”.
   

6. The overall quality of the English language in the article is good, but there are still some minor grammatical and spelling errors. The authors are advised to proofread the article carefully to ensure its professionalism.
   Answer:
   The text was proofread as suggested by the reviewer.

7. The article presents a novel model that relates the distribution of annual burned area to climatic conditions. The statistical method of combining climate data and wildfire activity is less innovative, and it is recommended that artificial intelligence-related methods be introduced to explore the factors influencing wildfire activity as a result of climate change.
   Answer:
   The aim of the paper was to assess the contributions of climate variability and climate change to the interannual variability of burned area in Portugal. The methodology proposed has proven to be adequate to reach the proposed goals and may be extended to other ecosystems. Studying the factors influencing wildfire activity as a result of climate change, as proposed by the reviewer, is certainly an important and exciting topic but I believe that such endeavor is beyond the scope of the present work. The reviewer’s suggestion is now incorporated as future work at the end of the Concluding remarks section.
   
   
8. The content of the study has important implications for understanding and predicting wildfire activity, especially in the context of climate change. This is essential for developing effective fire management and prevention strategies.
   Answer:
   I thank the reviewer for his positive view on the implications of the work performed.

9. The methodology and conclusions of the study appear to be scientifically sound. However, the authors are advised to further discuss possible limitations and directions for future research.
   Answer:
   The following sentence was added at the end of the Concluding remarks section (line 485 of the original manuscript):
   “It is however worth noting that the usage of statistical models of burned areas such as the one proposed in this work presupposes that the relationships among weather, vegetation and ignitions do not change in time. This shortcoming, that is especially relevant when addressing the impact of climate change, can be circumvented in part by exploring the roles played in wildfire activity by the above-mentioned three factors for different scenarios of future climate”.

10. The overall strength of the article lies in its in-depth analysis of wildfire activity in Portugal and its exploration of the impacts of climate change. The study provides valuable insights and may have a positive impact on fire management practices.
    Answer:
    I thank the reviewer for the positive appreciation of the work performed.

 

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript entitled “The signature of climate in annual burned area in Portugal” focuses on climatological and ecological impacts of wildfire activity. The aim is to assess the impact of climate variability and climate change on annual burnt area, considering CDRS. It is very interesting to see the methodological approach. 

There are some comments that need to be taken under consideration  

Firstly, the manuscript includes several studies carried out regarding wildfire activity in Portugal, with an excessive number of self-citations. There are 71 from 76 references, in which the author is co-author. The bibliography needs to be improved and other references need to be included in the sections: Introduction and Discussion.

Line 54 – Please paraphrase as “an amount that is twice the annual average for the period 1980-2023”

Line 122 – Please change the sentence “As mentioned in the Introduction”

Line 123-124 –Please explain more clearly the sentence “For each year,….of the year”

Line 386 -387 - Please see again the sentence as, I don’t understand what the meaning of (10) is.

Lastly, the conclusions are more general and are not drawn from the results and discussion, regarding Portugal.

Comments on the Quality of English Language

The manuscript entitled “The signature of climate in annual burned area in Portugal” focuses on climatological and ecological impacts of wildfire activity. The aim is to assess the impact of climate variability and climate change on annual burnt area, considering CDRS. It is very interesting to see the methodological approach. 

There are some comments that need to be taken under consideration  

Firstly, the manuscript includes several studies carried out regarding wildfire activity in Portugal, with an excessive number of self-citations. There are 71 from 76 references, in which the author is co-author. The bibliography needs to be improved and other references need to be included in the sections: Introduction and Discussion.

Line 54 – Please paraphrase as “an amount that is twice the annual average for the period 1980-2023”

Line 122 – Please change the sentence “As mentioned in the Introduction”

Line 123-124 –Please explain more clearly the sentence “For each year,….of the year”

Line 386 -387 - Please see again the sentence as, I don’t understand what the meaning of (10) is.

 

Lastly, the conclusions are more general and are not drawn from the results and discussion, regarding Portugal.

 

Author Response

1. Firstly, the manuscript includes several studies carried out regarding wildfire activity in Portugal, with an excessive number of self-citations. There are 71 from 76 references, in which the author is co-author. The bibliography needs to be improved and other references need to be included in the sections: Introduction and Discussion.
   Answer:
   This is a pertinent remark, but, as stated in the Acknowledgements, “this study was motivated by the ‘2023 Academic Career Award in Meteorology’ that I received from the Portuguese Association of Meteorology and Geophysics (APMG). (…) Success in an academic career is very seldom solely due to one’s effort; instead, it is generally the outcome of long-lasting cooperation with colleagues and students. The unusually long list provided in this study of papers I have co-authored is by no means to be viewed as an unrestrained desire to quote oneself but as the way to thank all of those that made up my career by listing a sample that I did my best to make it representative. The bias introduced is however easily corrected by looking into the references listed in the papers cited”.
   Given the relevance of the issue, and in order to avoid any misinterpretation, the following paragraph was added at the end of the Introduction section:
   “This paper is the outcome of the “2023 Academic Career Award in Meteorology” that the author received from the Portuguese Association of Meteorology and Geophysics (APMG). In line with this award, the topic of the paper was chosen so that a review could also be made of the author’s rather long academic career that he started on the 1st of August 1982 as a teaching assistant at the Department of Physics of the Faculty of Sciences of the University of Lisbon, Portugal. This is the only reason for the unusually long list of co-authored papers that are referred to in the Introduction and Concluding remarks sections. The list is just to be viewed as a way of paying tribute to all those that have contributed to make up the author’s career”.

2. Line 54 – Please paraphrase as “an amount that is twice the annual average for the period 1980-2023”
   Answer:
   The sentence was changed accordingly.

3. Line 122 – Please change the sentence “As mentioned in the Introduction”
   Answer:
   Lines 119-125 of the original manuscript were rewritten as follows:
   “Daily values of DSR are defined in a 0.5°×0.5° regular lat-lon grid, with 150 grid points covering the Portuguese territory. In this study, for each day of the period 1980–2023, averages were computed of DSR on grid points over Portugal. Then, for each year, the Cumulated Daily Severity Rate (CDSR) was obtained by adding the daily spatial averages of DSR from January 1 to December 31. For each year, daily spatial averages of DSR over Portugal were also accumulated from January 1 up to each day of the year; the notation CDSR(d) will be used to refer to the cumulative DSR from 1 January to day d”.

4. Line 123-124 –Please explain more clearly the sentence “For each year,….of the year”
   Answer:
   Please see answer to the previous comment.

5. Line 386 -387 - Please see again the sentence as, I don’t understand what the meaning of (10) is.
   Answer:
   Lines 386-388 of the original manuscript were rewritten as follows:
   "The displacement to the right of the cdf  curve for the last 30-year period compared to that for the first 30-year period indicates that, for a chosen threshold BA*, the respective cdf values are lower in the last period than in the first; attending to relation (10) that defines return period, such decrease implies a decrease in the respective return periods for the chosen threshold".

6. Lastly, the conclusions are more general and are not drawn from the results and discussion, regarding Portugal.
   Answer:
   Although the proposed methodology was applied for Portugal, it can be extended to other ecosystems because fire activity is the result of the “fire triangle” composed of weather-vegetation-ignitions. The purpose of this concluding section is to provide a unified view of the problem paving the way to future applications in other regions of the world. For instance, a model following a closely related approach to the one presented for Portugal is currently being developed for the Brazilian Pantanal.