A Data-Fusion Approach to Assessing the Contribution of Wildland Fire Smoke to Fine Particulate Matter in California

Round 1

Reviewer 1 Report

This study proposed a Bayesian statistical approach to assess wildfire contribution to PM2.5 concentrations. It incorporates three data sources: the satellite-derived HMS smoke plume data, the high-quality but sparse AQS monitoring data, and the abundant but less accurate PA monitoring data. This approach is then applied to two case studies in 2020 and 2021 in California, with 1%~3% increase in PM2.5 concentrations being attributed to wildfire smoke.

The manuscript is well-written, and the Bayesian approach looks sound and effective. My major concern is its alignment with the scope of the Remote Sensing journal, as its major merit lies in the newly proposed Bayesian data fusion approach that is “purely statistical”. The authors have also emphasized a lot about the benefit of incorporating the PA sensor data, which are obtained from a low-cost ground monitoring network. The only connection to remote sensing is the utilization of the widely used HMS data, which appears to be less central to the core focus of this study. Please see more detailed comments and suggestions below.

1. PA sensors can be installed both outdoors and indoors. For the purpose of this study, only data from outdoor sensors should be used. Please confirm this in the text.

2. In Fig.1, the color bar used for the HMS plume is non-conventional with a red color for low and a blue color for high, which seems counter-intuitive.

3. What is the data source of the meteorological variables including temperature and relative humidity used in this study?

4. In line 150, it’s better to add a note that j ∈ {1, 2} represents the two ground monitoring networks with the specific correspondence.

5. Section 2.3 introduces two estimation methods to quantify the contribution of wildfire smoke to PM2.5 concentrations, both of which are approximations with a few limitations. For instance, both estimators only use two weather variables (i.e., temperature and relative humidity) in covariate vectors or matching criteria without consideration of other meteorological or anthropogenic impacts on PM2.5 concentrations. They also consider the HMS data as a good indicator of near-surface PM2.5 concentrations, which is not always the case due to the miss detection of fire smoke in HMS or the absence of vertical distribution info in the HMS data. According to a previous study by Buysse et al., 2019, only 30~70% of days with overhead HMS smoke show enhanced near-surface PM2.5  concentrations, which might bias the assessment in this study. Please add more discussion about these potential limitations.

6. In Fig.4, it is recommended to overlap fire locations on the maps to aid readers in understanding the smoke contributions. Additionally, the results based on the two different estimation methods appear visually identical on the maps rendered with the current low-contrast color bar. It is suggested to use another color bar with higher contrast for better visualization.     

7. Please clarify the time scale of the fractional contributions of fire smoke to PM2.5 concentrations in the conclusion for comparison with other studies.

Reference:
Buysse, C. E., Kaulfus, A., Nair, U., and Jaffe, D. A.: Relationships between particulate matter, ozone, and nitrogen oxides during urban smoke events in the western US, Environ. Sci. Technol., 53, 12519-12528, https://doi.org/10.1021/acs.est.9b05241, 2019.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

General comments

References are indicated by numbers, however, when indicating the reference to later talk about its results as has been done in line 26 and other lines, my recommendation is to include the name of the authors and not only the number of their reference, as done in line 66 with ‘Stein et al.’ Please revise it.

Abbreviations are defined for several variables and parameters. Once they are defined, it is convenient to use this abbreviation instead the long name, otherwise it is better not to define it. This is the case of ‘Purple Air’ (already defined in l. 43) in line 65 amongst other cases. Please, revise along the text.

Since the manuscript is a paper and not a report or some longer document, the paragraph from l. 81 to l. 89 is not necessary, and it should be removed. The same for the paragraph from l. 299 to l. 309.

‘Fire season’ is mentioned along the manuscript but there is no mention to the month or month of the year when this ‘fire season’ is considered or studied.

It is not clear enough if the figures 2 and 3 are results obtained from the study. If they are results from the study the should be included in the ‘Results’ section. Please, clarify it. All the results from the study (even intermediate results) should be included in the ‘Results’ section.

Too many appendices have been included, therefore, it should be clarified in the text that the inclusion of said appendices provide necessary information for the article. Moreover, Appendix A could be included in the section 2.1.

In the ‘Discussion’ section it should be mentioned figures and tables obtained in the ‘Results’ section to link both sections. Moreover, there is no ‘Conclusion’ section and it should be included.

 

Specific Comments

l. 27: only PM2.5 has been previously defined. Revise it

l. 36-37: this line is repetitive with l. 53-54. Merge or delete one of them.

l. 70-79: in this paragraph there is no mention to the location (California) and it should be included. The same in Figure 1.

Some equations lack numbering. Please, revise it.

l. 193-194: symbol w was used for ‘frequency’ (l. 150) and in these lines is denoted as a ‘scaling factor’. Please, revise it.

l. 212: indicate which appendix. Same for l. 263

l. 207 indicate that MCMC is shown or developed in Appendix B

Figure 1: since the figure 1 will be referred into the three sections (2.1.1 to 2.1.3), all the comments related to the figure 1 should be included at the end of section 2.1 instead of in each subsection.

 

Figure 4: revise ‘PM2.5’ in caption (in the rest of the manuscript ‘2.5’ is written as a subscript). Instead of explaining the meaning of 1.02 I recommend to explain the meaning of the limits of the interval (1.00 and 1.03)

Appendix A: please, revise the value in line 330 since 2000 µg/m3 it seems too high.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Dear authors,

Thank you for your exciting and comprehensive research. I found it necessary and worthy of attention. The problem of forest fires and their impact on planetary health is in constant focus, is relevant all the time and needs to be clarified for better management decisions.

I would recommend your manuscript for publication after minor revisions:

1. Line 32 - you cite a comment about previous studies, but you didn't provide specific references. In addition to the Agency's report, investigations are probably published in scientific papers. I strongly recommend that several scientific papers be included in the review regarding the example of the California fires, as well as fires across the northern hemisphere, to evaluate the already published experience (https://doi.org/10.1016/j.scitotenv.2023.165594; https://doi.org/10.1016/j.envint.2020.106143; https://doi.org/10.1016/j.envpol.2022.119324; https://iopscience.iop.org/article/10.1088/1748-9326/10/10/105001).

2. The sections of the Materials and Methods require explanation so that the reader more fully understands the data format, spatial and temporal resolution of the data, accuracy assessment, errors, and units. It is essential to provide a more detailed description of the satellite data and qualitative and quantitative reviews of the air quality measurement station. 

3. you've presented Bayesian theory in detail, and 

that's not bad. The logic of its application is revealed in the methods. Perhaps you should add a more straightforward explanation (briefly) of why you chose this approach since statistically it is the most common method... However, it is important to have a detailed description of the data samples and their formats for a complete presentation. 

I'm sure your work deserves to be published. Congratulations, it has been a pleasure to study the results of your research. Good luck 

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Please see the attached report.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 5 Report

The manuscript under review presents an intriguing study that combines data from various sources for fire smoke modeling in California, an area that is increasingly gaining attention in research. Despite its relevance, I have significant concerns that need to be addressed to improve the overall quality and effectiveness of the study.

 

Elucidation of Data Fusion Algorithm: The manuscript briefly mentions a data fusion algorithm but fails to provide a detailed explanation of the same. There exist numerous advanced data fusion methods, such as STARFM and ST-Cokriging. Therefore, it is imperative for the authors to provide a rationale for selecting the Bayesian-based approach and how it better fits the study region. The authors should also clarify whether the algorithm has been adopted from existing literature or if it has been independently developed for this study. Furthermore, evidence to support how this algorithm outperforms recognized algorithms is necessary.

 

Need for Detailed Background Information: The manuscript would greatly benefit from a more comprehensive background. Why was the 2021 fire selected for this study? How versatile is the proposed method when applied to other fire situations? As most authors seem to be located outside California, were there any in situ measurements taken? If so, they should be discussed.

 

Representation of Original Data: The presentation of original data could be improved. More maps should be incorporated to better illustrate the data. The contrast in the final results map is low, making it difficult to discern spatial patterns.

 

Additional Comments:

 

In Section 2, the authors mention the use of satellite data. It would be beneficial to specify the data source, along with its spatial and temporal resolution. The term "near-real time" lacks scientific precision and should be replaced with a more appropriate term.

 

Figure 1 would be better suited in the study area or data section, rather than in the introduction. Its legend incorrectly labels the plume as points when it should be colored polygons. The map could be improved by adding key elements such as a scale bar, a north arrow, and location indicators.

 

Figure 4 requires image enhancement for clearer pattern recognition. A method such as histogram matching is recommended for this purpose.

 

In conclusion, the manuscript presents a promising study, but it requires considerable improvements in terms of methodological clarity, background details, and data representation. Addressing these concerns would significantly enhance the overall quality and impact of the research.

English writing is good overall

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have addressed all the comments, and I have no additional remarks.

Reviewer 5 Report

Thanks for addressing my comments