Comments on Draft ENVIRON Ozone Review Report

Rich Poirot, Air Pollution Control Div., VT DEC

December 27, 1995

Comments submitted to the:

OTAG Modeling Workgroup, through Mike Koerber, LADCO and

OTAG Data Analysis Workgroup, through Dave Guinnup, EPA OAQPS

This is in response to request for comments on the 11/95 ENVIRON draft report on Review of Recent Ozone Measurement and Modeling Studies in the Eastern United States. Also incorporated here are comments received from Alan Leston, CT DEP and Cliff Michaelsen, ME DEP. Overall, the report represents an excellent and informative summary effort which (with some modification) should be a very useful support document for OTAG.

1. It appears that the (original and) principal focus of the report is a review of modeling (rather than measurement) studies. Consequently, the comprehensiveness of the measurement section is quite limited in comparison to the modeling section. The Data Analysis Workgroup had identified a review of recent ambient measurement studies as an important task from a data analysis perspective (although no specific contractor was identified, nor was an estimated budget identified for the task). Its unclear whether (a future version of) the ENVIRON report can/should be viewed as the vehicle to achieve that measurement summary task (or what funds, if any, might be available). If this report is intended to cover a measurement summary task, substantial additional effort would be required. Possibly measurement summary activities currently underway through NARSTO could help meet this objective?

2. As the Report s Contents refer to a large number of Urban/Mesoscale Modeling and Measurement Studies (in preparation) which are not included in the current draft, it appears intended as a first installment of a work-in-progress. Given this in progress status, many of my review comments - suggesting additional areas of discussion - may be premature. At the same time, several conclusions stressed in the executive summary may also be somewhat premature. In particular, I m concerned that the current summary may preempt or misdirect OTAG policy/strategy discussions by speculating too severely on the relative ineffectiveness of NOx controls in general, and stationary source NOx controls in particular. In nearly every case, where a reviewed modeling study has indicated potential benefits from NOx control, the report includes editorial comments to the effect that these benefits were probably overestimated because... This kind of critical review is fine; I think its just a tad one-sided (don t see any suggestions that past studies may have underestimated benefits of NOx control or overestimated benefits of VOC reductions). A few counter-examples of this nature would make the report more useful for OTAG (and more credible). Several such counter-examples are suggested below.

3. The Report s stated intent is to provide ... an improved understanding of the transport of ozone and ozone precursors. However, the principal focus of the executive summary is not on transport per se, but rather on evaluating several OTAG initial premises regarding the comparative effectiveness of NOx vs. VOC reductions. The questions of which emissions species? , where do they originate? and what are transport contributions? are obviously inter-related and are each relevant to OTAG discussions. However, the report is silent on the questions of the origins and magnitude of transport contributions. This may be due, in part, to the use of the simple terms - urban and regional - in the initial OTAG premises, and to the report s specific focus on evaluating these premises. The terms urban and rural, are ambiguous at best (do we mean all urban areas, those with populations greater than X, or only urban non-attainment areas?) and may lead to over-simplification. From the perspective of each current (or potential future 8-hour mean or seasonal SUM06) ozone non-attainment area, what is the ozone contribution from precursor emissions outside that non-attainment area? Or from the perspective of each source region (or source category), what are the contributions to ozone concentrations in all downwind areas? Several of the regional or meso-scale modeling studies (for example model runs in the New England and NY domains) have provided useful information on the how much transport question by reducing all within-domain emissions to zero. Several other modeling studies reviewed in this report, and other measurement and measure/model (observation-based) studies could shed useful light on the how much transport question. Would like to see more emphasis on this aspect, as the question of which (local or upwind) emissions species should most effectively be controlled, is dependent on the relative contributions of local and upwind sources.

4. The exclusive focus of the Measurement Analysis Studies section is on temporal trends (it might be more accurately entitled Measurement-Based Trends Analysis Studies). The principal use of the (historical trend) measurement studies in the report (executive summary) is to confirm the effectiveness of the (historical experiment) VOC-only control approach: Historical reductions in VOC emissions have produced significant reductions in ozone concentrations in the Northeast corridor and other urban regions. At a minimum, this generalization might be modified to describe general reductions in peak 1-hour concentrations. When longer averaging times are considered (as indicated in the GM analysis and elsewhere by Lefohn, and/or for 8-hr or seasonal SUM06 - as per currently considered EPA standard revisions), ozone levels may be unchanged or even increasing in many of these sites/regions (despite substantial VOC-only reductions). Also, the historical trend in NOx emissions remains somewhat uncertain, and is otherwise temporally and spatially complex. From the recent EPA Air Quality Trends Report (9/95), we see that over the past 10 years, estimated national NOx emissions have increased by 3% (and by about 6 % during the summer). However, over the same 10 year period, measured national average NO2 concentrations have decreased by 9% (similar to the estimated 10% reduction in national anthropogenic VOC emissions). Measured NO levels have exhibited similar (or greater) declines in many eastern urban areas (despite presumed reductions in scavenging from declining ozone levels). Over approximately this same time period, there has also been a downward trend in measured wet nitrate deposition in the eastern US, as reported in the 1995 NADP report on Precipitation Chemistry Trends in the United States: 1980-1993. Thus there would appear to be a substantial directional discrepancy between emission estimates and ambient measurements.

5. From the recent EPA National Air Pollutant Emissions Trends, 1900-1994 report, we see an estimated 1985-94 NOx emissions increase in the OTAG region (approximately EPA Regions I through VII) of +2.5% (similar to the estimated increase in national NOx emissions). However, on a sub-regional basis, NOx emissions have decreased in the Northeast (EPA regions I (-10%), II (- 8%), III (- 2%)), while increasing in the Southeast (region IV (+ 8%)) and Midwest (regions V (+5.5%) and VII (+ 10%). Thus, the sub-regions where the largest improvements in met.-adjusted peak ozone concentrations have been observed, are also regions where the largest estimated NOx emission reductions have occurred. In the 1995 Rao et al. Study (JAWMA, 45:57-61), some of the largest, most consistent statewide 1983-92 ozone reductions occurred in CT (improvements at all sites and improvements of more than -2%/yr at 6 of 10 sites). Rao et al. estimated annual ozone reductions of -3.2% and -4.2% at Bridgeport, CT and New Haven, CT. Recent CT DEC evaluation of long-term ambient NO2 and NO concentrations at these sites during the summer (JJA) ozone season indicates 1981-95 average annual reductions of about -1%/yr for NO2 and about -3%/yr for NO levels at these 2 sites (Al Leston, CT DEP, pers. com.). Thus, the historical trends and spatial patterns in estimated VOC and NOx emissions and in measured ozone and NOx, do not completely support the summary statement that historical ozone reductions have been exclusively due to VOC reductions (nor is there strong evidence in these historical records of the alleged disbenefits of NOx control).

6. The report cautions on several occasions that regional model results (often characterized by coarse grid sizes) tend to overstate benefits of NOx controls - near urban areas, by artificially mixing rural, biogenic VOC with urban NOx - and in rural areas by artificially mixing elevated point source NOx with ground level biogenics. These points are well-taken, but may be overly speculative, given our very limited, measurement-based understanding of biogenic VOC emissions, concentration patterns and reaction chemistry, and the exceedingly poor performance of current generation models in handling and reproducing ambient levels of biogenic species. In other words, we may not know enough about biogenic VOCs to state definitively that models are treating them artificially. Conceivably, errors caused by coarse grid sizes may be relatively small compared to (or in opposite directions from) errors in model emission estimates or chemistry. Recent measurements in SOS region indicate that Isoprene is the most important reactive hydrocarbon in the Atlanta urban area (its not an exclusively rural species). Measurements at several suburban sites around Atlanta indicate that relatively high isoprene levels often persist over night (its not necessarily destroyed in a few hours). SOS estimates of historical isoprene levels in Atlanta suggest that a net effect of deforestation in Atlanta has been to increase isoprene emissions - due to increased temperatures (historical VOC trends are not necessarily declining). Recent PAMS measurements during ozone episodes in the Northeast (1994 NESCAUM PAMS report) indicate that isoprene is also one of the most important reactive hydrocarbon species at several northeastern urban sites, including Hartford, CT, Chicopee, MA and New York City. In those rural areas where NOx emissions may be predominantly from elevated point sources, surely there are occasions - such as during sunny, mid-summer afternoons with high solar potential for ozone formation - when vertical mixing of these emissions with surfaced-based biogenic or anthropogenic VOCs is rapid and efficient. Conversely, when such mixing does not occur, those elevated point source NOx emissions have a greater potential for long-range transport - contributing to ozone production in areas further downwind.

7. On several occasions, the report emphasizes that control of NOx emissions at ground level is many times more effective than control of elevated sources. This generalization may require some clarification. Do we mean effective on a per/ton or per/$ basis? Are we comparing reductions in local or upwind sources? If the goal is attainment, is it most effective to achieve this goal through the control of local, ground-level sources alone? For short transport distances, reduction of surface NOx (or VOC) emissions is likely to be more efficient on a per/ton basis than reduction of elevated emissions. But as transport distances increase, the relative efficiency of controlling NOx emissions in general, and elevated NOx emissions in particular increases. In the Rao and Mount modeling study, cited as the origin of the finding that controlling surface NOx emissions was 2-3 times more beneficial than controlling elevated NOx emissions, the tested scenarios represent complex combinations of proportionate emissions reductions from specifically defined regions within the OTR. The 2-3 times more beneficial comparison is based on reductions in peak ozone levels in and downwind of NYC from proportionate 75% reductions of area source NOx emissions within the Amtrack corridor vs. point source reductions from certain sections of the OTR outside the corridor. This does not necessarily mean that controlling surface emissions in upwind regions is more effective than controlling point sources in those regions. In any event, the incremental benefits of 75% upwind stationary source NOx reductions in this study ranged from 10 to 31 ppb reductions in peak ozone on the modeled days, and reduced concentrations much closer to standards than the 25% control scenario.

8. The recent Rao et al. Work on the ozone baseline concept is discussed on several occasions in the report, but is not mentioned in the (critical) section on Control Strategy Effectiveness. There are at least 3 important implications of this work that might be mentioned in this section: 1) Control strategy implications from any historical episode modeling applications are flawed to the extent that they predict changes which are partially based on unpredictable, stochastic processes; 2) Control strategies which appear effective in reducing (unpredictable) episodic ozone levels can result in minimal improvements in the (predictable) baseline; 3) NOx-based reductions were far more effective than VOC-based reductions in improving (a surrogate index for) the predictable baseline component .

9. In light of the above comments, I think the recommended caveat on regional NOx controls (p. xv) is not appropriate, or is at best, not useful and should be deleted. The initial premise (Regional NOx control may be effective ...) is already worded with sufficient caution, is a true statement as written, and is at least as justifiable as the perhaps-too-confidently-stated endorsements of the unquestioned efficacy of VOC reductions on p. xiv. The advice that all NOx controls need to be evaluated on a case by case basis is either too obvious (VOC controls should also be evaluated on a case by case basis, if the goal is attainment), or unmanageable (are we to make 1,000 model runs - to evaluate effectiveness of reducing each NOx source?).

10. Other recent modeling or analysis work that might be considered in future updates might include publications and summary reports from the SOS group (summarized by Chameides and Cowling, (1995) The State of the Southern Oxidants Study: Policy-Relevant Findings in Ozone Pollution Research); NESCAUM (1992 Regional Ozone Concentrations in the Northeastern US, and Preview of 1994 Ozone Precursor Concentrations in the Northeastern US); OTC (Technical Support Document: The Long Range Transport of Ozone in the Northeast Ozone Transport Region); NARSTO; LADCO; 1993 San Diego conference, 1990 Cambridge, MA Symposium on The Role and Importance of NOx Emission Control (Water, Air & Soil Pollution, 67/1-2 (1993)); 1987 Quebec, Canada, North American Oxidant Symposium , assorted recent modeling and analysis reports from eastern Canadian Provinces and Environment Canada; several recent PAMS analysis studies sponsored by EPA; results from recent Gulf of Maine/NARE field study; recent measurement or modeling studies by A. Lefohn, S Wofsy, S. Sillman, etc.


R.E. Morris: Review of Recent Ozone Measurement and Modeling Studies in the Eastern United States - Daft Nov 30 1995. ENVIRON Corp., Navato, CA 94945-5010, 415-899-0700

Submit your comments, feedback, questions, and ideas pertaining this page. Your input will be automatically added to the existing annotations. In order to add a new comment, you must be registered with the OTAG/AQA Peoples Page.