Winter Ozone Pollution in Utah’s Uinta Basin is Attenuating
Round 1
Reviewer 1 Report
Mansfield and Lyman have compiled extensive data for ozone concentrations in the Uintah basin to demonstrate that the long-standing problem of wintertime ozone pollution is getting better. I liked the novel approach to develop a proxy of temperature inversions in the absence of suitable measurements over the past decade.
Their main point that ozone pollution is going down because emissions from oil and gas development are decreasing is adequately supported by the data. I have some suggestions for additional useful details that could be provided in the text and thoughts on alternative explanations and discussion of the implications.
Page: 3
I suggest to add the material about organic measurements that is in Figure1 legend to the main text. The caption should only be about what is in the figure.
section 2.3. Please briefly mention some key attributes of the QA criteria rather than just citing where the criteria are documented. At least a list of what are the accuracy and precision for the measurements would be helpful here to give the reader a point of reference to judge whether reported differences are outside analytical uncertainty.
Page: 4 line 124. For the average snowdepth are you just averaging all the stations that have measurements together equally or weighting them by some measure of the area that they represent? Alternatively mention somewhere whether the locations of sampling points is random enough that the average is representative, or if there is some chance of bias in the locations that could influence the averaging. A map of the basin showing locations of measurement sites perhaps with some indication of where the oil fields are along with elevation might be helpful here as well as for understanding the characteristics of sites in addition to the nearby well densities.
line 132: Can you clarify how you averaged the data? The legend refers to it as average daily, but you have only 1 value per season. Is it the average of all the daily averages during the season? Then you might call it the seasonal average of daily (what time interval?) mean ozone. Do all the sites have about equal data coverage in each year, or are there gaps to contend with?
Line 124: In addition to the snow depth is there any variability in number of days with snow on the ground? If the mechanism is due to presence/absence of snow then the number of snow-covered days in a season rather than the average snow depth might be the best correlation.
Page: 11
Figure 8. Presenting a standard error or confidence limit for the slopes in addition to the R2 would be helpful. R2 values don't have information about whether the calculated slope is significantly different from 0.
Page: 12: line 275, Here as for elsewhere when giving a slope please include a measure of its confidence limit such as standard error or 95% CL.
Page: 13: Line 294 The comparison of TNMHC for 2012-2014 and 2018-2020 is not very convincing given the very large variability in early years. These plots are interesting, but don't put too much into the comparisons between such different data sets. I notice for Roosevelt that the medians for first three years line up with a value comparable to or below the medians in last 3 years. Also high average for first interval at Roosevelt is skewed by high average in first year, which has over 2 order of magnitude range.
Page: 14
Line 316 Rather than attributing the decline in emissions to global demand it is probably better to link directly to trends in exploration and production in the basin. I would agree that markets drive investment decisions to explore and develop new wells, but decisions on managing existing wells will be influenced by other factors. Local emission may diverge from global demand depending on where in its life cycle the basin is. e.g. production would go down even with high demand if the reserve was played out. I see a couple things in the figure 12 trends/ Drilling has declined considerably. Aren't there more emissions during the drilling phase than when the well is in production? Gas production has gone down more than oil in spite of fairly constant number of wells. Why is gas production declining? Is there less gas left in the reservoir so less gas coming out with the oil, or is gas being generated but not captured because the price is too low? Is it flared, or vented? If so, that might account for uptick in VOC.
I see in subsequent section that the distinction between drilling and production is called out. Are there other aspects of production not already mentioned in regard to pollution controls that are relevant here?
Page: 15
line 333.: More might be said about the decline in gas. As I queried above, is there simply less gas coming out with the oil, or is not being collected. Are the NOx emissions associated with gas coming from combustion to drive compressors?
The volatile content of the oil and what steps are being taken to collect them seem particularly relevant here.
Page: 16:
Line 391:I agree global markets would be an influence, but I also think you should add some mention of development stage for the oil field as well. Even if demand had stayed high would oil exploration have continued in the basin? Is a decline in emission a normal pattern in the evolution from new exploration to mature production? Considering that is relevant to whether we would expect ozone episodes to go back up if demand for oil and oil prices increase again.
Line 396: I think you can be more definitive about whether these trends account for NOx decline by telling us more about the prevalence of heavy duty diesel or other sources that power drilling rigs and gas compressors
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Overall comments:
The submitted paper by Mansfield and Lyman has analyzed trends of the surface ozone and ozone precursor concentrations of 38 measurement stations over 11 winters (from 2010 to 2020) in the Uinta Basin of western Utah, USA (UBU USA). The values of the trends and their statistical significance under different meteorological conditions (snowpack depth, temperature inversion strength and wind speeds for ozone-precursors) are evaluated. Authors found and used successfully the combination of the meteorological parameters (snowpack existence and vertical atmospheric stability) to divide all winter seasons with the ozone measurements into two goops: (1) the first with the snowpack and high vertical stability of the atmosphere; (2) the second without the snow and with the atmosphere that is more close to neutral conditions. Such division allowed them to mark the winter seasons when the ozone concentrations are high and rather sensitive to increasing the ozone-precursor species (grope 1 with statistical significant negative ozone trends). Meanwhile the ozone trends from the second winter grope were found more often positive, but with low statistical significance. Also the ozone concentrations into the second grope were close to their background values. All this allowed the authors to suggest some reasonable explanation of the observed winter ozone trends into the UBU USA over previous decade.
Thus the suggested case study is of interest and close to the journal thematic range.
However it is necessary to make following remarks:
- Maybe make sense to divide the second section of the paper (Methods) into two separate parts. The first one might be devoted to the description of the utilized data and their quality/reliability. The methods of the data treatment and estimation could be discussed into the second part.
- Testing statistical significance of trend values (Section 3.8, Table 2) need to be described more thouroghly. Also together with the author’s testing approaches it would be useful to assess the significance within the framework of some statistical model consisting of a linear trend plus noise. (For example: whether a trend is significant (no equal zero) is verified by the ratio between the estimated trend and its standard error under the assumption that the ratio is distributed as Student’s-t).
Also I would like to make a number of miner comments.
Specific comments:
- Line 97: The discussion for Figure 1 needs to be moved from the figure caption to the text of the paper.
- Line 152: It would be useful to add a name of station to the caption of Figure 3.
- Lines 147: It is not clear what do the authors mean “a correlation coefficient” here and in what follows (square, absolute value)? It is positive (0.77) but Figure 3 shows a negative correlation of the ozone exceedance count with time.
- Lines 159: The vertical axis of Figure 4 should be described.
Optinal comments:
- Lines 45-47: It would be useful to give some references/arguments for these statements.
- Line 224: Instead “…already provide evidence…..” may be better to say to “ ….have already provided some evidence….”.
- Line 461: Instead “…Utah.” it is better to put “ …Utah,….”
Some misprints
- Line 234: Instead “…NOx titrates…“ better to say “…NOx titrate….”.
- Line 274: Instead “….a deline to only 91%….” should be “….a decline to only 91%….”.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
General comment
In general the manuscript is interesting, useful and good written; provides informative and good investigation. However, the manuscript should be corrected and updated.
Specific comments:
In the abstract and Introduction, it is necessary to indicate what kind of ozone is being studied, tropospheric or surface or ground-level ozone? Are there also local measurements of stratospheric ozone - total ozone column?
Line 71:
Should be:
(2) Our group at Utah State University (USU) has for many years measured
Line 97: It looks like the body text is not separated from Figure caption:
From [16].Measurements of organic compounds have been less plentiful and less consistent…
Should be:
From [16].
Measurements of organic compounds have been less plentiful and less consistent…
Line 124: ‘studies. [9–11].’ Should be ‘studies [9–11].’
Line 182: Correlations are not presented in Table. In title, should be like ‘Number of correlations …’
Line 317: and (2)many pollution should be ‘and (2) many pollution
In the manuscript Section 5 should be included - Conclusions - where the main results should be summarized. Perhaps a sentence is needed about the general significance in life of those concentrations that the authors study.
In the Supplement, Table S2 - please explain again - what you mean with word "season"? E.g., in Antarctica season means all period of "summer" expedition, from November till March. What is in your case? Table S2 - red and black shading - what that mean??
===========
More comments are included in the pdf text of the manuscript.
My opinion - the manuscript can be published after minor revision and corrections mentioned above.
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.docx
