Methane Emissions in Boreal Forest Fire Regions: Assessment of Five Biomass-Burning Emission Inventories Based on Carbon Sensing Satellites
Round 1
Reviewer 1 Report
Comments for author File: 
Comments.pdf
Some sentences may be a bit long, but they can be slightly shorter and can still be further polished.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
General Comment
Zhao et al. (2023) have assessed methane emissions in boreal forests using remote sensing and ground station data. The thematic involved in this Article is within the scope of Remote Sensing and would be interesting to the readers of the journal. However, this piece needs to be heavily improved prior to publication. Please consider the following comments:
(i) The aim of the study is unclear, and the Introduction section is sometimes confusing.
(ii) A better characterization of the global biomass burning inventories, comparisons, describing their methods, and the two approaches (burned area or FRP) is required. This is the core of your study and needs further development. Moreover, how were they processed?
(iii) The bilinear resample of the estimates does not make sense to me. Why not summing the values of the finer spatial resolution inventories at the coarser spatial resolution to avoid creating interpolated values?
(iv) It is not clear what the authors considered as “trends of changes” in the Results section.
(v) Methods and results are often confused in the Results section.
(vi) Directly correlate the inventories (mass emitted) with concentration data does not make sense to me. You are comparing locally emitted methane (within each grid cell) with concentrations that include emissions originated in other regions but transported to the grid cell.
(vii) Authors must discuss the difficulties and inaccuracies of MODIS-derived FRP in higher latitudes. This can actually help explaining the low values found by QFED and others, which are often much higher than GFED.
(viii) The manuscript has too many figures. Perhaps remove or transfer some of them to the supplementary materials.
(ix) The manuscript would benefit from a minor English review and editing.
Specific Comments
Line 45: Northern biome of where?
Lines 45-46: Please define the location of the boreal forests in here.
Line 47: Missing reference.
Lines 47-49: Are these related only to boreal forests or worldwide?
Line 49: hm2?
Line 50: forest.
Line 55: 30% to 45% of what? Total emissions?
Line 71: Please add references to this inventory.
Line 86: Prevent?
Line 96: Remove “forest areas with”.
Lines 96-99: Very confusing sentence. Please clarify.
Line 101: Which ones?
Line 102: . In order.
Lines 101-110: The aim of the study is unclear.
Line 115: How?
Lines 115-116: Where? Globally? I do not agree with this sentence. Please add references.
Lines 116-118: This sentence is unclear.
Lines 116-118: Will you discuss the uncertainties in these products later in the text?
Line 135: Are these daily data? Monthly? Hourly?
Line 155: Global biomass burning, not biomass combustion.
Line 167: Fire Radiative Power
Lines 197-215: Not clear why these metrics were chosen and to what they were applied (comparing the inventories?). Also, please add references that successfully used these metrics in similar studies.
Line 221: The meaning of “trends of changes” is unclear.
Line 229: Please choose a word more appropriate than “speculated”.
Line 249: “Stie” in the legend of the figure. What is “D-value”?
Lines 255-256: This is part of the methods.
Lines 309-310: Why? Usually, we have the opposite when comparing these inventories.
Lines 315-316: Authors must discuss in depth these significant differences. Why is there a peak in 2018 for GFAS in RUS but not to the other inventories?
Lines 337: What about GFAS?
Line 349: Show.
Line 362: Will these results be discussed later in the text?
Line 394: Please improve the quality of this figure.
Line 394: Is it ok to directly correlate the inventories (mass emitted) with concentration data?
Lines 451-453: This is not the case of your study area.
Lines 456-457: So, how do you explain the distinct patterns between them?
Lines 463: I do not agree. The values used to convert FRP to burned biomass and emissions depend on the vegetation type (for example, see the values adopted by GFAS).
Lines 486-488: By which sensor?
Line 519: Please clarify the axis of this figure.
Line 516-543: This is a summary of the results instead of conclusions and pointing a future direction of your work.
The manuscript would benefit from a minor English review and editing.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
This study addresses a very important research question, that is, how accurate is the estimation of fire methane emissions from existing inventories? The authors applied several methods to investigate the reliability of five global fire emission inventories and discussed the consistencies and differences among them. Although the research question proposed in this study is important, the authors fail to address this question in a clear and appropriate way as in the current version. Thus I suggest the authors make substantial improvements on the following aspects before consideration for publication in Remote Sensing.
Major comments:
(1) The most concern I have is the conclusion made from this study. The authors stated that QFED2.5 was found to be more reliable compared to the other four emission inventories. The authors reached this conclusion only based on the results of Euclidean distance analysis. I doubt the representativeness of this method in making such a strong conclusion of which emission inventory is most reliable. Being the most important method in this study, the authors did not provide a detailed description of this Euclidean distance method. How is it calculated? What does the result mean? Does the distance value only indicate the temporal similarities between an emission inventory and the satellite observations? In other words, the Euclidean distance cannot tell which emission inventory is accurate in terms of the magnitude as well as the spatial distribution. It could be possible that the QFED2.5 inventory is most accurate in terms of monthly variations but least accurate in terms of the emission magnitudes (as shown in Table 2, the QFED2.5 shows the lowest emission estimates). As also mentioned by the author (lines 468-469), “the level of the values cannot be used as a standard to measure the accuracy of inventory estimation”. In summary, the most important conclusion from this study fails to stand with the current analysis. Therefore, without adding more solid analysis to support this finding, the conclusion made from the current study is not robust, thus making the novelty and contribution of this study minimal.
(2) The authors spent quite some effort describing the inter-annual, seasonal, and monthly variations of the satellite observations (section 3.1) and methane emissions (section 3.2). I don’t see how these analyses could be useful in addressing the question of which emission inventory is most reliable.
(3) The English writing of this manuscript needs to be improved significantly. Many places throughout the manuscript need re-writing, clarification, or corrections. Here I am giving only a few examples. The authors should carefully go through the manuscript and make necessary changes. Otherwise, the current writing causes confusion as I am reading it.
a) Line 54: “When severe forest fires occur, fire emissions from methane emission sources can account for 30% to 45%...”. This sentence does not make sense. Should it be methane emissions from fires can account for 30% to 45%?
b) Line 57. What do you mean by “destructive potential”? Being destroyed or destroying other things?
c) Line 59. Inaccurate description of “84 times higher warming potential”. On which time scale?
d) Line 62: “a growth rate of 15 ppb yr-1 in the past year”. Over how many years or which period? Please specify.
e) Line 102: No period before new sentence “In order to …”
f) No need for capital words. For example, Line 121 “Temperate coniferous forest”, Line 115 “high Forest cover”
g) Line 267. Should “carbon dioxide” be “methane”?
h) What do you mean by “lighting factors” in Line 280?
i) Line 329: “gas temperature” should be “air temperature”?
Minor comments:
(1) Please provide information on the percentage of fire methane emissions from the study region to global fire methane emissions. By giving a percentage of the global totals, it would make more sense why the authors chose the study region.
(2) It would be better to provide a summary table showing the details of the five emission inventories discussed in the study, for example, the spatial resolution, the temporal resolution, the underlying estimation method, etc. With this kind of table, the readers are much easier to know the differences and similarities among different emission inventories.
(3) Can you explain why the spatial distribution from GFED4.1 is different from other inventories?
(4) The title of sections 3.2 and 3.3 are exactly the same.
(5) What’s “satellite inversion of CH4 concentration data” in Figure 13? Is it the same as the XCH4? The data source for Figure 13 is unclear. Please clarify.
(6) What does the “D-value” in Figure 4 mean? The difference between satellite observations and ground observations? The authors should clarify.
(7) What is Ce in Line 453?
The English language of this manuscript needs to be improved substantially. Please refer to the above comments.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 4 Report
The research focused on assessing the impact of forest fires on greenhouse gas emissions, specifically CH4, through a verification approach using satellite remote sensing data. The article is relevant to publish in the journal, but I have some suggestions to improve the understanding of the text.
The abstract could clearly and directly state the research objectives and proposed innovations, making it easier for the reader to understand what the study intends to achieve. While it is mentioned that typical wildfire regions worldwide have been considered, the summary does not specify which these regions are. The abstract does not describe the methodological steps used in the article, such as numerical comparison, correlation analysis, and trend consistency analysis were performed to quantitatively assess consistency and differences in datasets on a regional scale. Insert a sentence explaining the procedure for eliminating 29 lag effects and combining time series similarity measures, as this is a differential of this research.
The introduction does not discuss how spatial and temporal scale differences might impact the results. This may be relevant to understanding the constraints of inventories and observations. The conclusion of the introduction is brief and needs to clearly emphasize what the study seeks to achieve in advancing science and contributing to the understanding of greenhouse gas emissions.
In the methodology, the text can insert a description and figure of the logical flow, helping the reader understand how the different parts fit into the overall narrative.
When discussing the differences between the GBBEI, it is necessary to consider how they impact previous studies and how the results can be interpreted in light of these differences.
Upon reading the article, it is unclear whether the analysis of the lagging effect is a discovery of this article or has already been the subject of other studies on this specific topic.
Figures
Figure 1 – The geographic coordinates are difficult to read. Maybe increase the figure.
Figure 2 – It looks distorted. What is the cartographic projection of the map?
Figure 14. The figure caption does not mention subfigures “a” and “b”.
Minor corrections
Lines 19, 20, 24, 154 - “biomass burning emission inventories” I suggest “biomass-burning emission inventories”
Line 94 - “sources, we” I suggest “sources, and we”
Line 122 - “forest area, dominated” I suggest “forest area dominated”
Line 130 - “and utilizes the” I suggest “and utilized the”
Line 166 - “provides 0.1” I suggest “and provides 0.1”
Line 170 - “provides global” I suggest “and provides global”
Line 174 - “emission coefficients estimated to” I suggest “emission coefficients is estimated to”
Line 202 - “analysis method, used” I suggest “analysis method used”
Lines 286 and 288 - “CH4 concentration shows” I suggest “CH4 concentration showed”
Line 308 - “CH4 while” I suggest “CH4, while”
Line 325 - “occur in July” I suggest “occurred in July”
Line 349 - “Figures 11 and 12 showed” I suggest “Figures 11 and 12 show”
Line 391 - “that lagging effect” I suggest “that a lagging effect”
Line 486 - “As shown in the Figure” I suggest “As shown in Figure”
Line 491 - “, Russia in” I suggest “, Russia, in”
Minor corrections
Minor corrections
Lines 19, 20, 24, 154 - “biomass burning emission inventories” I suggest “biomass-burning emission inventories”
Line 94 - “sources, we” I suggest “sources, and we”
Line 122 - “forest area, dominated” I suggest “forest area dominated”
Line 130 - “and utilizes the” I suggest “and utilized the”
Line 166 - “provides 0.1” I suggest “and provides 0.1”
Line 170 - “provides global” I suggest “and provides global”
Line 174 - “emission coefficients estimated to” I suggest “emission coefficients is estimated to”
Line 202 - “analysis method, used” I suggest “analysis method used”
Lines 286 and 288 - “CH4 concentration shows” I suggest “CH4 concentration showed”
Line 308 - “CH4 while” I suggest “CH4, while”
Line 325 - “occur in July” I suggest “occurred in July”
Line 349 - “Figures 11 and 12 showed” I suggest “Figures 11 and 12 show”
Line 391 - “that lagging effect” I suggest “that a lagging effect”
Line 486 - “As shown in the Figure” I suggest “As shown in Figure”
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The authors have responded to my suggestions point by point and have all been revised. This manuscript can be accepted.
Author Response
Thank you very much for your valuable feedback on our manuscript during the review period.
Reviewer 2 Report
Dear authors,
Thank your very much for your effort on revising the manuscript. I really appreciate your hard work and consider that the manuscript has improved. However, there are still important issues to be addressed:
Abstract – Better result in your study area.
Supplementary Material – I do not have access to the SM, so it was not revised (this is a MDPI fault, not yours).
Interpolation method – I am still not convinced about this method. You have to improve the justification for its use.
Line 124: “a more reliable inventory” produced by you?
Lines 498-519: This explanation is quite controversial. According to your results QFED (FRP-based) is the most reliable inventory but you explain the lower QFED results to the high inaccuracies in MODIS FRP. How can a more inaccurate input data lead to a more reliable inventory?
Line 518: fire radiative power
Line 594: “selection of the study area” makes no sense in here.
Lines 594-609: this section needs to be improved significantly. For example: what is the percentage of uncertainty involved in each variable?
Minor editing is required.
Author Response
Thank you very much for your valuable suggestions on our manuscript during the review period.
The attachment is our point-by-point response.
Author Response File: Author Response.docx
Reviewer 3 Report
The authors have addressed most of my concerns. One minor comment I have is please provide some references of existing studies that applied the Euclidean distance method (in Section 2.2.5) to address similar questions like this study.
The English language of the current version has been improved. A final thorough check would be needed.
Author Response
Thank you very much for your valuable suggestions on our manuscript during the review period.
We appreciate your comments on our research. The following are the relevant reference materials we provide, which mainly use the idea of Euclidean distance to conduct research on the similarity between time series data and data patterns. Considering that both satellite data and inventory data have undergone preliminary data research and analysis in our study, the Pearson correlation coefficient between the two is relatively high, and the time series of the data is relatively intuitive. Therefore, we use the Euclidean distance method to quantitatively analyze the similarity of time series.
- Bautista-Thompson, E.; De la Cruz, S.S. Shape similarity index for time series based on features of euclidean distances histograms. In Proceedings of the 2006 15th International Conference on Computing, 2006; pp. 60-64.
- Cheruvu, A.; Radhakrishna, V.; Rajasekhar, N. Using normal distribution to retrieve temporal associations by Euclidean distance. In Proceedings of the 2017 International Conference on Engineering & MIS (ICEMIS), 2017; pp. 1-3.
- Yu, K.; Guo, G.-D.; Li, J.; Lin, S. Quantum algorithms for similarity measurement based on Euclidean distance. International Journal of Theoretical Physics 2020, 59, 3134-3144.
- Singh, P.K. Similar vague concepts selection using their Euclidean distance at different granulation. Cognitive Computation 2018, 10, 228-241.
Reviewer 4 Report
The authors incorporated all recommended revisions. Therefore, I congratulate the authors for the research.
I recommend that the authors check the verbal tense used in the results section, as conventionally, this section requires the past tense.
Author Response
Thank you very much for your valuable suggestions on our manuscript during the review period.
We have checked the tense issues in the results section (corresponding modifications are reflected in the manuscript). We use the simple present tense in the chart description and the past tense in the conclusion presentation section.
