Drivers of extreme burnt area in Portugal: fire weather and vegetation

Calheiros, Tomás; Benali, Akli; Pereira, Mário; Silva, João; Nunes, João

doi:https://doi.org/10.5194/nhess-22-4019-2022

Articles | Volume 22, issue 12

https://doi.org/10.5194/nhess-22-4019-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/nhess-22-4019-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 22, issue 12

Research article

|

19 Dec 2022

Research article |

| 19 Dec 2022

Drivers of extreme burnt area in Portugal: fire weather and vegetation

Tomás Calheiros, Akli Benali, Mário Pereira, João Silva, and João Nunes

Download

Final revised paper (published on 19 Dec 2022)
Preprint (discussion started on 30 Jun 2021)

Interactive discussion

Status: closed

RC1:
'Comment on nhess-2021-173', Anonymous Referee #1, 04 Aug 2021

The paper "Spatial variability in the relation between fire weather and burned area: patterns and drivers in Portugal" aims at analysing large (>100 ha) summer forest fires in mainland Portugal. Although the dataset is sound and of potential interest, I found several flaws in the methods and results interpretations. The use of statistics is not always appropriate, expecially the use of OA, UA and K (they are used for classification accuracy against a reference). Furthermore no analyses on the spatial pattern are provided, simply a description of the maps. The relationships between LULC and weather and fire occurence need to be deepen analysed. The authors need to clarify and explain better the reason why they divided in cluster the dataset, the significant differences between groups,.... Other multivariate analysis methods can be adopted considering more variables (e.g. geographical gradients, inhabitants,...) in order to obtain a strong explanatory analysis and then building relationships or models to be tested.

The discussion section is mainly a list of municipalities in the different groups or sentences related to parameters not considered in the study. The first part is belonging to methods. The conclusion is redundant, repeating results and discussion elements.

In its current version I think the manuscript needs a complete revision and is not eligible for publication.

Citation: https://doi.org/10.5194/nhess-2021-173-RC1
- AC1: 'Reply on RC1', Tomás Calheiros, 03 Dec 2021
  
  The comment was uploaded in the form of a supplement: https://nhess.copernicus.org/preprints/nhess-2021-173/nhess-2021-173-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/nhess-2021-173-AC1
RC2:
'Comment on nhess-2021-173', Anonymous Referee #2, 28 Oct 2021

The manuscript from Calheiros et al is very difficult to read. It reads as a series of bullet points linked together and the authors did not even bother at breaking the text into paragraphs or making any effort to increase readability. This is not a result of the authors not being native English speakers (I believe paragraphs also exist in Portuguese) but rather denotes a major lack of attention to detail.

The authors examine thresholds in burnt area associated with DSRp and how they differ across Portugal. The way they present the data is somewhat misleading: they make us believe that DSR has a very high correlation with burnt area. These types of correlations have been described before and they result from the ordering of values (from small to large). If that order is removed and simple scatter plot of burnt area DSR is presented, that relationship usually breaks, or is much weaker. I would thus encourage the authors to be more careful when using these types of analyses.

I’m not familiar with the clustering techniques used by the authors, and I will not comment on those. I will just point out that the results are rather shocking because pretty much all clusters are distributed across all Portugal, but it is well known that fires in N PT differ substantially from S PT (the authors actually state this in their introduction as well).

Citation: https://doi.org/10.5194/nhess-2021-173-RC2
- AC2: 'Reply on RC2', Tomás Calheiros, 03 Dec 2021
  
  The comment was uploaded in the form of a supplement: https://nhess.copernicus.org/preprints/nhess-2021-173/nhess-2021-173-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/nhess-2021-173-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

ED: Reconsider after major revisions (further review by editor and referees) (07 Jan 2022) by Yves Bühler

AR by Tomás Calheiros on behalf of the Authors (18 Feb 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (21 Feb 2022) by Yves Bühler

RR by Anonymous Referee #3 (25 Mar 2022)

Suggestions for revision or reasons for rejection

This study focuses on assessing the Daily Severity Rating (DSR), an additional component of the Canadian Fire Weather Index (FWI), to provide insight into its suitability for summer large fires in Portugal and the role of vegetation. The manuscript states that filling this knowledge gap can support the study region's fire prevention and suppression planning. Overall, the paper was informative. However, there were several issues with clarity, conciseness and structure, which makes it hard to evaluate the validity of the methods, results and the overall study.

Some major general points:
1. The title would spark more interest keeping it shorter and removing redundancy. The authors should consider editing the title to be concise and relevant.
2. The abstract explained the aims and tasks of the research and the main findings. However, authors should improve the context and state the research's need and conclusion to make it easy to understand to less specialised audiences.
3. The introduction provided enough background and a clear knowledge gap that the study aims to address. However, to improve the understanding, the authors should reorganise the background topics into a logical, simple, and clear structure, avoiding talking about the same point multiple times throughout the introduction. For example, the authors could move from the Mediterranean context to Portugal, join fire weather and indices background, and combine vegetation with land uses changes, referring briefly to the urban-rural interface as this driver is not the focus of the study. The last paragraph of the introduction could mention the study of Calheiros et al. (2020), indicate the knowledge gap, what the authors did to fill the gap, and finally, an overview of the paper's contributions.
4. The authors could complete the study area description providing information about the vegetation across Portugal. Additionally, as the map of figure 1 shows pyroregions, the authors should refer to them in the text or remove them. Finally, the last sentence, lines 108-110, should be moved to the introduction (e.g., before mentioning the findings from Calheiros et al. (2020) in the introduction) or discussion.
5. The methods section should be reorganise to present the data and the approaches in a clear structure. For example, the author could use tables to list the data sources and the burned areas metrics. Furthermore, split the information into focused and clear subsections. A similar structure should be followed when reporting the results. The discussion should focus on the main findings.
6. The conclusion should be more general and concise. The authors should avoid repeating the results and focus on the meaning of the results and perhaps implications and perspectives. As a first step, they could focus only on the last paragraph and remove the earlier information.
7. To improve the quality of the figures, the authors should be consistent using the same map layout, the terms (e.g., r2 vs R2) and plot frames. Furthermore, they could make all the plots bigger, remove titles on the top of the plots and ensure the caption descriptions.
8. The references are relevant, and appropriate key studies are included. However, authors should review and correct the list. For example, on page 18, line 569:679, remove the link.
9. The English language is simple. However, some odd constructs that, if rephrased, would improve the quality and flow of the manuscript's text and make it easy to read. The manuscript should get proofread in English.

Some minor points:
1. On page 2, line 37, to remove statistical concept.
2. On page 2, line 23, to add, e.g., in the references.
3. On page 3, line 57:59, to provide references.
4. On page 3, line 88, to edit, for Portugal. Entire and Continental Portugal are redundant.
5. On page 3, line 89, change reveal to revealed.
6. On page 4, line 112:114, to re-edit.
7. Define the first time using acronyms in the text and all the figure captions. For example, on page 4, line 117, to define ECMWF.
8. On page 5, line 122:123, remove the last sentence.
9. On page 6, line 167:176, to move to the discussion.
10. On page 7, line 205, remove because the software was mentioned before.
11. On page 14, line 430:436, to re-edit, remove or move to the introduction, where the authors should be sure the study goal is clear.

References:
T. Calheiros, M.G. Pereira, J.P. Nunes, 2021. Assessing impacts of future climate change on extreme fire weather and pyro-regions in Iberian Peninsula. Science of The Total Environment, 754, 142233, doi.org/10.1016/j.scitotenv.2020.142233.

Hide

ED: Reconsider after major revisions (further review by editor and referees) (25 Mar 2022) by Yves Bühler

AR by Tomás Calheiros on behalf of the Authors (19 May 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (24 May 2022) by Yves Bühler

RR by Anonymous Referee #3 (06 Jun 2022)

Suggestions for revision or reasons for rejection

The authors improved the manuscript compared to the original version. However, there is still a need for clarity, consistency, and specificity to enhance the manuscript's quality. Also, the paper still benefits from English editing to remove unnecessary words and thus be more specific and easier to read.

Abstract
• On page 2, line 33, to add the context. E.g., to support fire management. Furthermore, the authors should follow a similar in the abstract and introduction for consistency. Currently, the abstract start with the link between fire weather indices and fire behaviour, while the introduction defines the term fire regime.
• On page 2, line 35, add the knowledge gap after Portugal. E.g., but it is still poorly understood to support fire management.
• On page 2, line 46, add the object of the paper itself before reporting the results. It is beneficial for readers to get overall the content of the article. Similarly, on page 11, line 236. E.g. This paper clarifies the effectiveness of the DSRp for estimating BA in Portugal and how vegetation influences it.
• On page 3, line 58, remove the space in 'KEY WORDS'.

Introduction
• On page 4, lines 61: 66, the authors should assess if that information is needed or specifically start by stating burned area as a significant component of fire regimes as in line 67 and continue providing some context on the fire weather and vegetation drivers and why this is relevant for fire management.
• Generally, the authors could improve the introduction by following a more logical and accessible structure for clarity. E.g., burnt area, fire weather, vegetation content, knowledge gap, research goals and a statement summarising what the paper is presenting. Furthermore, as stated in the title, the authors should be consistent with the content order in all the sections. That would make the paper easier to read.

Methods
• The authors reported burnt and vegetation data in the same subsections, 2.3 and subsection 2.4. is using the clusters from subsection 2.5. Thus, the authors should simplifier further the method section into a more logical and specific structure. E.g., 2.1. Study area, 2.2. Burnt area, 2.3. Fire weather, 2.4. Vegetation, 2.5. Analysing clusters of burnt area. 2.6. Analysing burnt area and fire weather relationship, 2.7. Analysing the influence of vegetation on the fire-weather relationship.
• Furthermore, the authors should assess if the subsections 2.6 and 2.7 should focus on the data analysis (e.g., from line 403 in subsection 2.4) and thus move any data preparation to the data subsection.
• Also, they should refer to table 1 in 2.2, 2.3 and 2.4 subsections and table 2 in subsection 2.2.
• On page 20, line 403, change 'achieve the first objective' to the first objective.

Results
• The authors should follow a similar structure for clarity and consistency. E.g., 3.1 Burnt area clusters, 3.2 Burnt area and fire weather relationship (at the national and municipal level), 3.3 Influence of vegetation on the fire-weather relationship.
• Once a short name is defined in the text, the authors should consistently use it. For example, on page 33, line 587, change 'burnable area (BNA)' to BNA, as it was defined before.

Discussion
• As for the previous sections, the discussion should follow a similar structure. E.g., 4.1 Burned area and fire weather relationship, 4.2 Influence of vegetation on the burnt area and fire weather relationship, 4.3 Considerations and implications for management.
• On page 47, line 788, to re-edit for clarity. The sentence would be more appealing, stating a clear, meaningful message. The authors should do with every main message in each subject in the discussion. E.g., the main message on page 48, lines 806:807.
• On page 51, line 898, change 'a very' to 'an'.

Conclusions
• On page 54, line 952, remove 'usually'.
• On page 54, line 953, change 'resolution' to 'resolutions.
• On page 54, lines 955:959, to simplify the message and change 'As far as we know' to 'To our knowledge.
• On page 55, lines 963:967, change the order of the last two sentences. The implications should go to the end of the conclusions.

Figures and tables:
• On table 1, show first burnt area metrics, then fire weather indices and finally vegetation for consistency.
• In table 2, define the two following burn area metrics: BNAF/BNAS and TBAF/TBAS, for example, as the ratio of forest BNA and shrubland; the ratio of forest TBA and shrubland TBA. To keep consistency, edit the area of Forest in capital letters.
• In figure 2, add a space between 'Nº' and 'Fire' and ':'.
• In figure 3, add spaces in the information provided in the plot area and change 'p-value' to 'p-value' for consistency.
• On figure 11 caption, line 746, add '(BA)' after 'Burnt area'. Also, in line 747, change 'Area' to 'area'.
• To add and define all the short names (e.g., ‘DSR’) in figure captions. E.g., on line 777, to specify 'BA'.

References
• On page 59, line 179:1080, double-check the reference.
• On page 65, line 1226, remove '(Table 1),'.
• On page 66, line 1247, remove ','.

Hide

ED: Reconsider after major revisions (further review by editor and referees) (07 Jun 2022) by Yves Bühler

AR by Tomás Calheiros on behalf of the Authors (12 Jul 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (14 Jul 2022) by Yves Bühler

RR by Anonymous Referee #4 (11 Nov 2022)

ED: Publish subject to minor revisions (review by editor) (11 Nov 2022) by Yves Bühler

AR by Tomás Calheiros on behalf of the Authors (21 Nov 2022) Author's response Author's tracked changes Manuscript

ED: Publish subject to technical corrections (22 Nov 2022) by Yves Bühler

AR by Tomás Calheiros on behalf of the Authors (24 Nov 2022) Author's response Manuscript

Short summary

Fire weather indices are used to assess the effect of weather on wildfires. Fire weather risk was computed and combined with large wildfires in Portugal. Results revealed the influence of vegetation cover: municipalities with a prevalence of shrublands, located in eastern parts, burnt under less extreme conditions than those with higher forested areas, situated in coastal regions. These findings are a novelty for fire science in Portugal and should be considered for fire management.