Meteorological Conditions Associated with Lightning Ignited Fires and Long-Continuing-Current Lightning in Arizona, New Mexico and Florida

Round 1

Reviewer 1 Report

Title: Meteorological conditions associated with lightning igniting fires and long-continuing-current lightning in Arizona, New Mexico and Florida

 

Overview: 

This study compares meteorological conditions and lightning ignitions in Florida, Arizona, and New Mexico. Some of the methods are a bit unclear (see below). The paper uses a lot of acronyms, so might be useful to have them all listed in a key for readers to check rather than going back through the text. Oh, it’s at the end- I just found it!

 

Any discussion of how different the climates are between ANM and FL? Fuels burn at much higher dew points in FL than ANM. And fires in Florida are usually not as large and catastrophic as some of the fires in western states. Not clear how comparing these two locations is important. California has large wildfires, but lightning seems to be a less important ignition source there. 

 

Other comments:

No line numbers so I will use section headings to identify locations.

Abstract: What is “ERA5”? What is “ANM”? Define all abbreviations at first use. 

Introduction- near the beginning: remove “the” ahead of “80%”

Section 2.2 Are the coordinates for each fire just a center coordinate? Or are they spatially explicit perimeters? 

Section 2.3 Did the authors create the vegetation map or it was downloaded from another source? If author’s created the map, they should report accuracy. If it came from another source that should be made clearer including where the data can be accessed. 

Section 2.4, Paragraph 2 I don’t understand how it would be useful to expand the area where meteorological conditions were monitored since fires were not also being monitored there. 

Section 2.6, First line: use “compared” rather than “compare”- there are other tense errors in this paragraph (present tense when they should be past: e.g.,  “follow”, “calculate”, “compare”

Text right below Table 1: I would change “with more number of LIW” to “with a greater number of LIW”

Last paragraph before Figure 1 change “assignation” to “assignment”

Figure 2a - why is y-axis in red text?

Figure 3a - why is y-axis in red text?

Section 4.1 “ARN” is used in one place instead of “ANM”

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Investigating the characteristics of thunderstorms and lightning producing fires is essential to improve fire forecasting methods. 

In this work, authors extend the result of research in the Mediterranean Basin to regions in USA (Arisona, New Mexica and Florida) in order to corroborate the role of lightning in the production of lightning-ignited wildfires and to provide new tools to improve the parameterization of wildfires in atmospheric models. They present an analysis of the characteristics of lightning and thunderstorms that produce wildfire during the fire season. In addition, they use optical lightning data provided by the lightning imaging sensor onboard the tropical rainfall measuring mission satellite to investigate the continuing current phase of lightning discharges over the USA between 1998 and 2014. This analysis allows to propose the preferential meteorological conditions from reanalysis that are useful to parameterize the occurrence of wildfires.

There are a lot of abbreviations in the work that make it difficult to read. In some cases, when reading, you have to guess about their meaning.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Review of manuscript: Meteorological conditions associated with lightning igniting fires and long-continuing-current lightning in Arizona, New Mexico and Florida

This study investigates the meteorological conditions and type of vegetation associated with lightning-ignited wildfires (LIWs) and long continuing current (LCC) lightning flashes in three states in the United States: Arizona, New Mexico and Florida. The study is very similar to a previous one led by the main author in the Iberian Peninsula and Greece (Pérez-Invernón et al., 2021), which lessens the novelty of the present work. Nevertheless, forecasting the occurrence of LIW and LCC is important from the point of view of wildfire prevention, and the present study findings may be useful for fire practitioners in the three states involved. My main concerns regarding the manuscript are related to the distribution of some parts of the text among the different sections and to the definition of some lightning-related concepts.

General comments:

The authors should justify why they chose three different states in the US (Arizona, New Mexico and Florida) as the study area. Was there some reason for this choice (e.g. contrasting lightning regimes, fire regimes, existing vegetation, meteorology, etc.)? As a reader, the first doubt I had was the reason for choosing these three states and not others.

The coordinates of the ignition point are expected to have an associated error. Was this error quantified? In the same way, for each starting point, would it not be better to assign the forest type that is most common in the pixels surroundings the focused pixel?

The text in many parts of the document should be moved to improve the flow and readability of the manuscript, as in these examples:

·       P2,L1-8. The aims of the study should be included at the end of the Introduction section after discussing the state of the art.

·       P3,L6-13. These sentences should be included in the Data section below.

·       P6,L21-24. This paragraph does not report results, but rather discusses results and should therefore be moved to the Discussion section. The same applies to P8,L8-11; P9,L6-8; P9,L45-46, P10,L1-11, P10,L39-44. P10, L50.

The meaning of “typical” in the context of CG lightning flashes must be clarified. Are the authors referring here to all the CG flashes relative to LIW flashes?

What was the rationale for using a 18 ms threshold for the LCC flashes? Although the authors briefly explain how this limit was established (page 3), I do think the explanation is clear enough for the readers of a journal like Fire.

Specific comments:

·       P1, title: replace “igniting” with “ignited”

·       P1,L4-5. This sentence can be omitted because it refers to the western US, not to the study area.

·       P2,L13. Include the definition of LCC the first time the term is used.

·       P2,L16-17. The type of vegetation can mediate the ignition phase but also the survival and arrival phases. Therefore, these sentences should be included at the end of the paragraph.

·       P3,L23. Define “IC” the first time the term is used.

·       P4,L1. I assume the coordinates refer to the presumed coordinates of the ignition point. Please clarify.

·       P5,L23. “Pérez-Inverno´n” should be “Pérez-Invernón”

·       P5,L47-48. Non-forest areas constitute the main type of cover in the sites where LIW started in Arizona and New Mexico. The sentence reads as if it is the third most common type of cover.

·       P5,L35 and 39. These two sentences can be omitted. They do not provide any relevant information.

·       P6, L11 and 16. There is no figure or table where the readers can examine the monthly distribution of LIE. I suggest including this information in the revised manuscript. The same applies to the elevation of all CG strokes, which could be included in Table 2.

·       P9,L16-18. The total-totals index has already been defined. This sentence can be omitted.

·       P10,L9. Replace “…is ANM” with “…in ANM”.

·       P10,L24. The only meteorological variable for which significant differences were found was wind shear in Florida. Nevertheless, no explanation of this result is provided in the Discussion section.

·       P11,L22. “In the twor egions” should read “In the two regions”

·       P12,L15. Replace “9” with “Fig. 9”.

 

 

References cited:

Pérez-Invernón, F.J. [et al.]. 2021. Lightning-ignited wildfires and long continuing current lightning in the Mediterranean Basin: Preferential meteorological conditions. Atmospheric chemistry and physics 21:17529-17557.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

The topic of this paper is highly relevant for forest fire danger assesment and prediction. So, this paper is suitable for Fire journal. English is good.

Abstract

It seems to me, it is not correct to make emphasize on vegetation type, for example, ponderosa pine in ANM, concerning forest fire occurrence. Because coniferous cover ignitability is well know for forest fire scientists. Howewer, similar results can be obtained for another vegetation cover for definite meteorological condition within another study area. You should make emphasize on meteorological conditions. For example, it is well know in scientific periodic that water must be in all three phases to tend lightning discharge. You wrote only about ice content and relative humidity. What about liquid phase?

Introduction

I think this section can be extend using some literature sources on forest fuel ignition caused by lightning activity.

Also, I think you should inform readers not only about NLDN, but you can use some information on World Wide Lightning Location Network (WWLLN). Maybe, you can provide pros and cons to use each other lightning detection system.

Data and Methodology

Could you provide some support references to use Larjavaara proximity index for searching lightning ignited forest fires. You should provide justification of use this index for ANM and FL territory.

Please, provide brief description of bootstrap procedure for calculating confidential intervals. There are different variations of bootstrap procedure are known at present time.

Results

Section 3.1 

I suggest to change size of Figures 1. These pictures should be bigger.

Figure 2.c and Figure 3.c, Please clarify distance from lightning stroke candidate to reported forest fire. What spatial resolution of NLDN system? Because It is unlikely that cloud-to-ground stroke can produce forest fuel ignition in so far distance like several km. You should think about interpretation of this result.

Maybe, this statistical analysis must be supported by natural observations  in forest within study area?

Discussion

You should exclude from discussion well known statements like "These studies reported a positive correlations between the occurrence of dry thunderstorm and LIW". Your discussion should be deeper.

Conclusions

I suggest exclude conclusion about forest vegetation.

Also I suggest to rework conclusions to exclude references from this section. Maybe, you should move some text to the discussion from the conclusions.

References

I suggest to extend references with some results on forest fuel ignition and WWLLN.

General remark

I did not find results with confidential intervals. You shoul plot these intervals on the corresponding figures.

Supplement

I suggest to merge main part of the article with supplement.

This is more suitable for readers.

Non published materials

I read the another article in Atmospheric Chemistry and Physics.

I think current article can be overlapped with 12 % with the previous article.

Authors explained this overlap. I pretty confident, this is not a reason to reject current article. Such studies have significant scientific value in th context of definite study area.

In conclusion,

I suggest major revision

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 5 Report

The authors presented a manuscript on the meteorological conditions associated with lightning-igniting fires in Arizona, New Mexico, and Florida. The paper brings interesting conclusions on the meteorological conditions associated with lightning-igniting fires including;  high temperatures between ground and 800 hPa levels, low precipitation rates, and high-based clouds. The methodology and results are well addressed. In general, the manuscript is well organized and the reading is relatively easy. I have minor concerns about the paper that I described below;

 

In table 1, the authors say “The lightning-ignition efficiency (LIE) for each forest type was calculated as the ratio of total LIW to total CG strokes.”

I didn’t understand the LIE completely. The criteria to determine the number of LIW is well explained, but what are the criteria for total CG strokes in terms of temporal and spatial distribution? Do you count the CG strokes for the same 10 km criteria but for different temporal occurrences?

 

Section 4, second paragraph, please change ARN by ANM.

 

I didn’t understand why Figures, 4, 5, 6, 7,8, and 9 are at the end of the manuscript and not along with the text.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Revision- unaddressed changes below:

• Any discussion of how different the climates are between ANM and FL? Fuels burn at much higher dew points in FL than ANM. And fires in Florida are usually not as large and catastrophic as some of the fires in western states. Not clear how comparing these two locations is important. California has large wildfires, but lightning seems to be a less important ignition source there. - somewhat addressed, but still not clear why these two regions were compared. 

• Section 2.3 Did the authors create the vegetation map or it was downloaded from another source? If author’s created the map, they should report accuracy. If it came from another source that should be made clearer including where the data can be accessed. - in author’s response, but should be cited in the paper. 

• Section 2.4, Paragraph 2 I don’t understand how it would be useful to expand the area where meteorological conditions were monitored since fires were not also being monitored there. - in author’s response, but I didn’t see explanation in the paper. 

 

Also wondering why “&” is used instead of “and” in the text now when listing Arizona and New Mexico. If journal allows, it’s fine I guess.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 4 Report

Dear authors!

Thank you for your revision!

I agree with most part of your answers and changes in manuscript. But suggest again to consuder following changrs in manuscript.

1. I agree that the scope of your research is NLDN data analysis. I agree that you will not make a comparison with WWLLN. But I suggest to include a few sentences about WWLLN in Introduction for more wide background of your current study. At present time several studies on forest fire occurrence and prediction is known using WWLLN. And I think you should inform readers about this situation.

2. Maybe, sample picture with CI can be presented in Supplement?

I suggest minor revision

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf