Data-based wildfire risk model for Mediterranean ecosystems – case study of the Concepción metropolitan area in central Chile

Jaque Castillo, Edilia; Fernández, Alfonso; Fuentes Robles, Rodrigo; Ojeda, Carolina G.

doi:https://doi.org/10.5194/nhess-21-3663-2021

Articles | Volume 21, issue 12

https://doi.org/10.5194/nhess-21-3663-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/nhess-21-3663-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 21, issue 12

Research article

|

03 Dec 2021

Research article |

| 03 Dec 2021

Data-based wildfire risk model for Mediterranean ecosystems – case study of the Concepción metropolitan area in central Chile

Edilia Jaque Castillo, Alfonso Fernández, Rodrigo Fuentes Robles, and Carolina G. Ojeda

Download

Final revised paper (published on 03 Dec 2021)
Supplement to the final revised paper
Preprint (discussion started on 12 May 2021)

Interactive discussion

Status: closed

RC1:
'Comment on nhess-2021-131', Paul Santi, 08 Jun 2021

While the writing is clear, there are a few unusual word choices and I recommend that the authors include another review for English grammar and word selection.

line 22 - what do you mean by “fire spots”? Fire locations?

line 52 - remove the comma after “windy”

line 53 - suggest changing to “…slopes rapidly lead…”

line 59 - “…categories of wildfire likelihood that correspond…”

line 65 - suggest removing the word “trace”

line 70 - change to “manages”

lines 70-74 - not a complete sentence

line 78 - change to “sprawl”

line 80 - change to “…2010) that lead to…”

line 84 - “…changes is related to more…”

line 98 - change to “…which are expanding rapidly, mostly for…”

line 99 - “Also, the region contains a variety of …”

line 101 - “characterizes”

lines 107-109 - the change from 44,123 hectares to 46,697 does not seem like much of an expansion to me. I would suggest saying that it has maintained a steady amount of forestry plantations in spite of urban growth.

line 119 - change “portraits” to “portrays”

line 121 - “frameworks”

line 123 - change “ingested into” to “included in"

line 126 - I suggest using the term “moderate” throughout the paper instead of “median,” which has connotations of calculated statistical parameters

lines 132 and following - wildfire spots term is unusual. Maybe use “wildfire locations” throughout the paper?

line 145 - “…network, were retrieved…”

line 158 - “NDII and NDVI data entered into the …”

line 169 - change “ingested” to “included” (suggest changing this word throughout the paper)

line 188 - change to “first”

Section 3.1 - this is a lot of discussion of numbers without reference to any figures or tables showing the actual data. I think it would help to show the data in some way.

line 234 - I am unfamiliar with the term “peri-urban”. Is this a standard LULC classification?

line 243 - change to “pixels”

line 250 - change to “hotspot”

line 253 - change to “indicate”

line 277 - what do you mean by “highest sectors”? Highest elevation, or largest area?

Figure 2 - Are the NDVI and maybe some other categories self-determined? For example, if an area is counted as burned, because it is in the database, wouldn’t the vegetation be a different type than the unburned land? That is, suppose pre-burn it was vegetation type A, but post-burn it is vegetation type B (since many locations revegetate with different species). Then your statistics would naturally associated vegetation type B with burned areas and assume they are more likely to burn than areas with vegetation type A.

Figure 3 - use larger font to make it readable? Also, you use the term “Medium” for the middle category here - make sure you use a consistent term throughout the paper (I suggest “moderate”).

line 317 - replace “recurrently” with “frequently”

line 334 - remove “for”

line 390 - replace “causing” with “requiring”

line 397 - use “2,1 million”

line 404 - explain what you mean by “some degree”. I would expect that to include the medium and higher categories, which is 40%. It seems like the 90% number would even include low risk, so I don’t think this statement is as strong as simply saying “this work shows that at least 40% of the CMA is subject to at least medium probability of wildfire occurrence.”

line 413 - “countries, along with other…”

line 428 - “areas) when there are more plantations.”

line 431 - replace “pointing” with “indicating”

line 435 - “two, not necessarily”

line 443 - remove contraction

line 451 - “it becomes”

line 462 - “improve or make”

Section 5 Conclusions - I suggest including a short paragraph summarizing the statistical results indicating influencing factors. The paragraph starting at line 480 belongs more in the discussion section, rather than introducing these new ideas in the conclusions.

Citation: https://doi.org/10.5194/nhess-2021-131-RC1
- AC1: 'Reply on RC1', Alfonso Fernandez, 25 Sep 2021
  
  The comment was uploaded in the form of a supplement: https://nhess.copernicus.org/preprints/nhess-2021-131/nhess-2021-131-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/nhess-2021-131-AC1
RC2:
'Comment on nhess-2021-131', Damiano Vacha, 10 Aug 2021

I really appreciated the article, which concerns a very important issue (the fire risk in WUI) and which needs to be thoroughly investigated. The overall structure of the paper is well balanced in its parts, the methods used are clearly presented and the assumptions made are justified. I agree with the observations made previously, especially as regards the need to present the results in figures/maps to make reading easier.

I would like to suggest that you specify a few points:

Line141-142: What are the characteristics of this database? How many fires? Are the data punctual or spatial? What is the extent of the burned areas? How are the fires distributed over the observation period?

Line 234-239: has the reliability of the database been assessed? Could the increased occurrence of fires near urban areas and roads be due to a sampling bias, caused by easier detection of fires in busier areas? Do the three sub-databases (Fixed terrestrial, Mobile terrestrial and Aerial detection) have the same spatial extension and resolution?

Line 269-274: Did you also evaluate the contribution of the slope and aspect themselves, not embedded in the solar radiation calculation?

Citation: https://doi.org/10.5194/nhess-2021-131-RC2
- AC2: 'Reply on RC2', Alfonso Fernandez, 25 Sep 2021
  
  The comment was uploaded in the form of a supplement: https://nhess.copernicus.org/preprints/nhess-2021-131/nhess-2021-131-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/nhess-2021-131-AC2
- AC3: 'Reply on RC2', Alfonso Fernandez, 25 Sep 2021
  
  The comment was uploaded in the form of a supplement: https://nhess.copernicus.org/preprints/nhess-2021-131/nhess-2021-131-AC3-supplement.pdf
  
  Citation: https://doi.org/10.5194/nhess-2021-131-AC3

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Publish subject to minor revisions (review by editor) (22 Oct 2021) by Mario Parise

AR by Alfonso Fernandez on behalf of the Authors (28 Oct 2021) Author's response Author's tracked changes Manuscript

ED: Publish as is (02 Nov 2021) by Mario Parise

Download

Article (2236 KB)
Full-text XML

Short summary

Wildfires pose risks to lives and livelihoods in many regions of the world. Particularly in Chile's central-south region, climate change, widespread land use change, and urban growth tend to increase the likelihood of fire occurrence. Our work focused on the Concepción metropolitan area, where we developed a model using machine learning in order to map wildfire risks. We found that the interface between urban areas and forestry plantations presents the highest risks.