High Ethylene and Propylene in an Area Dominated by Oil Production
Round 1
Reviewer 1 Report
The paper is well written and flows quite well.
The authors presents a unique set of data on hydrocarbons, alcohols and carbonyls in the Uinta Basin, Utah oil and gas fields. The analyses present convincing evidence that high concentrations of ozone precursor NOx are produced by various combustion engines used in the gas and oil fields. The authors' in situ measurement data suggest that official emission inventories of alkenes+acetylene are lower than reality as are overall inventoried emissions. Their data support earlier observations that reducing NOx would be a more effective strategy that reducing organics to control excessive wintertime ozone production in the Uinta Basin.
I find the maps confusing and hard to visualize the central focus of the data presented therein. I strongly suggest that a new Figure be added as Figure 1 showing just the terrain gradation, some elevation contours, rivers and locations of Horsepool, Roosevelt and Castle Peak (note: not annotated on present map). Then on subsequent maps, show the gas and oil wells, facilities and sampling locations on a map with a white background without the terrain showing. Do this on subsequent Figures without the terrain and vegetation data that confuses the measurement data presented such as in present Figures 3,4, 7 and S1-S11.
I suggest that the section on "Permanent Monitoring Stations" (lines 117-137) be moved up to appear before "Portable Sampling Stations" (lines 89-116).
I strongly suggest that a Figure of the ozone concentrations and ambient surface temperatures over time be presented with particular focus on the period February 22-29, 2019.
In present Figure 2, it would be valuable to know the locations of the sites plotted clearly on a map.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This paper describes VOC measurements made in the Unita basin of Utah, among oil and gas facilities, and the finding that the most reactive species, ethylene and propylene, are emitted mostly in the oil producing area of the valley and seem to be arising from the engines used to drive process equipment. The results are potentially useful and the paper could be published if several aspects of the presentation are improved upon. I have the following general and specific comments and suggestions that the authors need to attend to.
General Comments:
I found the units used here (mg/m3) to be quite awkward, especially because compounds are being lumped into classes and different compound classes are being compared. Unless this some kind of a Journal requirement, I recommend using parts-per-billion-carbon (ppbC), which will make it much easier to understand the analyses that are presented.
The data presented here lack context. A fair amount of previous work has been done in this air basin, but no quantitative comparisons have been presented. Also, the authors mention emissions inventories for this area, but only relative comparisons are alluded to. The things that would be useful to know are: 1) Are the absolute emissions changing in the Basin? this would be useful to know given there have been some regulations instituted in this Basin: 2) How well do the current inventories represent the emissions in an absolute sense since this is what drives ozone production? 3) What are the VOC measurements relative to methane levels? This could be very useful way to compare with past measurements, for example from the UBWOS 2012-2014 data.
In several places the authors refer to “ozone reactivity”. From context, this appears to be a misleading term. “Ozone reactivity” refers to the product of an ozone rate constant times a compound concentration, in the same way that “OH reactivity” is understood to be an OH rate constant times a compound concentration. The authors present the concept of incremental reactivity as a way to quantify the tendency of a compound to make ozone because of: 1) its reactivity in the atmosphere (usually with OH) and: 2) the propensity of the products of a compound to propagate radicals. I believe this is the feature the author intend to convey by the term “ozone reactivity”, and they are correct that the alkenes are more the most potent at this (per molecule). If this is what they authors intend, then it needs to be described properly. How about “ozone formation potential”?
Specific Comments:
Line 10. The important feature here is not the mountainous topography, it’s that the area is a basin, surrounded by mountains or higher ground, except for the Green River canyon that cuts through it at either end, N.E. to S.W.
Line 17. The nature of the data justifies at most 3 significant figures, e.g. 1145.7 should be 1150.
Line 19. The first example of the inappropriate use of “ozone reactivity”.
Lines 42-43. I thought much of the problem was that the inventories were not based on actual measurements of the equipment, but based on literature estimates – so-called “paper inventories”.
Lines 108. What were the box temperatures?
Lines 147-153. A diagram of the system would be nice here.
Line 172. We need details on what these inventories are based on. For example, do any of them incorporate results from previous measurements, e.g. the 2012-2014 UBWOS projects?
Line 187. Are these datasets log-normal?
Line 242-243. How did your absolute concentrations compare with those of Helmig et al,?
Line 248-249. I think you mean to say that the locations of oil and gas sources are also anti-correlated with elevation within the Basin.
Line 260. I assume this is percentage by mass? It would be best to do it by carbon, which is close for some species, but not for others, e.g. oxygenates like methanol.
Line 271. What do the dashes indicate in Tables S1-S3?
Line 301. How close are the data to a log-normal distributions?
Line 349. This is an overly-broad statement and is not true for some air basins such as Houston, where petrochemical facilities are the main sources of alkenes.
Figure 6. (and the data file I downloaded from the website). These numbers are labeled wrong, they should be micrograms/m3, not mg/m3.
Line 433. Do you mean this over-prediction is on an absolute basis or relative to the other compounds?
Line 445 and Line 466. “ozone reactivity” is again used inappropriately here.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
The paper is much improved. I have only the following comments and suggestions:
If you keep the mass-based units, please include the conversion (ie. mass to mixing ratio) for a few of the common compounds when you first present concentration numbers.
The authors are wrong about the appropriate number of significant figures to include in the numbers they present. It has nothing to do with detection limit- quote that in the methods section for sure. The appropriate number of sig figs reflects all the uncertainties in the measurement. If they present 5 sig figs, for example 1145.7, they are saying they know that number with an uncertainty of one part in 10000. There is no way that is correct, their standard is probably only known to +/- 5%, maybe a little bit better, and then there are all the other uncertainties associated with the method. Please change the numbers to reflect their estimated uncertainty.
Author Response
See attachment
Author Response File: Author Response.pdf
