Impact of Fire Emissions on U.S. Air Quality from 1997 to 2016–A Modeling Study in the Satellite Era
Round 1
Reviewer 1 Report
This manuscript discusses the impact of fire emissions on U.S. air quality. It is well-written for the most part, but there are a few areas that must be improved. None of the multi-panel have descriptors (a), (b), etc. which would help clarify references to them. Also, the references are not in a standard format.
Specific comments:
Why is the 2011 NEI used to create the 2014 baseline anthropogenic emissions? The 2014 NEI is available to be used. On a related note - why are the 2002, 2008, etc. NEI emissions used to provide baseline estimates for the years closest to them? The legend in Figure 1 looks a little strange. There is a number of metric tons and a unitless number. What do they represent? The lines for Figure 1 should be consistent for both the upper and lower panels. The bottom panel of Figure 3 looks to be of lower resolution than the upper panel. In Figure 4, the left panel seems to be of lower resolution. The panels seem to be shifted in Figure 6. Lines 281 and 290, Ozone is listed with two different units: ppmv and ppbv - it would be clearer if only one unit was used consistently. Figure 8 should not be on its side, it is difficult to see. Is there no Discussion and the authors went directly to Conclusions? This seems a bit strange.Author Response
Please refer to the attached file for the authors' responses.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The manuscript presents the results of a comprehensive modeling study to determine the effect of wildfires on PM2.5 and ozone concentrations throughout the continental US over a 20 year time horizon. The authors also conducted a detailed analysis on three regions, which are susceptible to wildfires. The study is scientifically sound, as the authors use well-tested and robust modeling techniques and data sources. The scope of the study is impressive; using a meteorological model and chemical transport model to simulate 20 years of surface concentrations is a considerable undertaking. A revised discussion of the policy implications and an additional post-processing step to derive changes in policy-relevant design values would strengthen the manuscript. Nevertheless, the manuscript is an important contribution to the literature and should be published subject to some minor revisions.
General comments:
Line 25 and other areas of the paper: The authors should be a bit more specific when referring to what form of the O3 and PM2.5 standard has been exceeded. There are two current PM2.5 standards (24-hour and annual) and three 8-hour ozone standards (70 ppb, 75 ppb, and 80 ppb)
Line 38: Not clear what climate change scenarios the authors are referring to
Line 121: While I understand that low spatial resolution grid is necessary considering the computational demands of a 20-year simulation, a 30-km grid is fairly large. Along with the model/measurement comparison provided later in the paper, is there any additional evidence the authors can provide showing that a 30-km grid is appropriate for the task at hand.
Figure 1 upper panel: Why is there emission tonnage for only CO and OC in the legend?
Figure 4: There are a lot of data points on these plots. Since the black color is saturated, it’s difficult to determine the distribution of points within each “blob”. The authors should use a density scatter plot as it will better reveal the measurement/model agreement.
Figure 4: Since some of the results are region-specific, separate scatter plots focused on each of the three study regions are highly recommended. The Supplementary Information may be the best place to place these plots.
Section 3.2: The prediction of secondary organic aerosols is a large uncertainty in CMAQ and will likely impact the PM2.5 model performance far downwind of fires. The authors should mention the deficiencies inherent in predicting the evolution of PM2.5 from fires and how they can bias the results.
Figure 8: This figure is stretched far too much along the ordinate, making it difficult to read. Perhaps a 2 x 2 configuration will be easier to interpret. Also, some regions of the percentage contribution maps (two right maps) are red in regions where there are no extra exceedance days. It may be better to use a different color on the percentage contribution maps in those regions as to not confuse it with areas with 100% contribution.
Results section: The count of exceedances is not always the best metric to judge air quality trends as it is binary and can be heavily influenced by slight variations in concentration around the standard. In addition, areas can still meet clean air standards even with air quality exceedances. Compliance with the 8-hr ozone standard is based on the 99th percentile 8-hour daily max value at each station individually, while compliance with the 24-hour PM2.5 standard is based on the 98th percentile of 24-hour PM2.5 standards. The analysis of how wildfires have affected exceedance days is useful, but a minimal amount of additional post processing could make the conclusions much stronger. Attainment of the NAAQS is based on the design value, which is a three year average of the 99th percentile 8-hour DM ozone values and the three year average of the 98th percentile 24-hour PM2.5 standard. The trend in the difference of the designs value with the presence of wildfires and without the presence of wildfires in several key metropolitan areas would be an extremely useful metric that could be easily calculated from the model data.
Line 360: Under the US Clean Air Act, wildfires are considered “exceptional events”. Air quality data influenced from exceptional events can be removed from air quality standard attainment calculations after a demonstration is approved by the US EPA. While emissions from wildfires still pose health risks, air quality managers typically do not factor in wildfire influence when preparing clean air policies. Perhaps a conclusion related to the changing proportion of anthropogenic to wildfire-derived air pollutant concentrations and subsequent health risk would be more warranted.
Minor comments:
Line 20: should be “more” instead of “much”
Line 41: The end of this sentence is awkward
Line 46: This sentence is awkward
Line 51: “horizon” is spelled incorrectly
Line 149: “Evolution” instead of “Evolutions”
Line 176: There is an extra space after “The”
Various locations throughout the manuscript: The term “Northern California” is used colloquially much more frequently than “North California”
There are some sections that do not have any text such as Patents, supplementary materials, author contributions, and funding.
The conflicts of interest section contains instructions instead of a conflicts of interest disclosure.
Author Response
Please refer to the attached file for the authors' responses.
Author Response File: Author Response.docx
Reviewer 3 Report
This well-written paper provides emission values of important pollutants as a result of wildfires. The fact that emissions vary wildly from year to year is not new as fires vary wildly from year to year. The authors might want to take a look at the National Interagency Fire Center data of annual wildfires and prescribed fires (https://www.nifc.gov/fireInfo/fireInfo_statistics.html) to examine correlations. The authors assume that all fire emissions come from wildfires; however, prescribed burning has been increasing dramatically in the U.S., and the contributions of prescribed burns would be an important discovery. However, this would probably be impossible to parse out from the data. I have placed some comments/questions within the manuscript.
There is no discussion section, which needs added. Research should answer the questions what, so what, and now what. The paper answers the question of “What?”, but the lack of the Discussion section does not answer “So what?” The Conclusion section does not answer the question “Now what?, but only provides a summary of the paper.
Comments for author File: 
Comments.pdf
Author Response
Please refer to the attached file for the authors' responses.
Author Response File: Author Response.docx
Reviewer 4 Report
There are some overlaps with the reference nr 24, including figures and idea of the work. The sections should be rewritten and the differences with reference nr 24 should be clearly stated.The discussion should be included.
Author Response
Please refer to the attached file for the authors' responses.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The authors have addressed all my major concerns. There should be an updated version of Figure 1 as it looks unpolished. Both the upper and lower panels (should be listed as (a) and (b) ) encompass the same time scale but they are of different widths. This should be corrected to make it easier to read and interpret. Also, the same line types should be used in both panels for consistency. The legend in the upper panel (which is apparently baseline emissions) is missing for most pollutants and that should be corrected.
Also Figure 9 should not have a border around it as it makes it look less clean.
Although these are only two comments, they must be addressed before publication.
Author Response
The authors have addressed all my major concerns. There should be an updated version of Figure 1 as it looks unpolished. Both the upper and lower panels (should be listed as (a) and (b) ) encompass the same time scale but they are of different widths. This should be corrected to make it easier to read and interpret. Also, the same line types should be used in both panels for consistency. The legend in the upper panel (which is apparently baseline emissions) is missing for most pollutants and that should be corrected. Also Figure 9 should not have a border around it as it makes it look less clean.
Thanks for the comments. We modified both Figures 1 and 9 as in the attached file.
Author Response File: Author Response.pdf
Reviewer 3 Report
No further comments.
Author Response
Thank you for the time and effort to make this a better manuscript.
