Preview of 1994 Ozone Precursor Concentrations in the Norteastern U.S
The NESCAUM Ambient Monitoring and Assessment Committee is composed of Air Program monitoring staff from the States of NY, NJ, CT, RI, MA, ME, NH, VT and EPA Regions I and II. Over the past several years, the Committee members have been involved in the collection of detailed ozone precursor data at a number of PAMS stations in the Northeast. The Committee has prepared the following report to initiate the process of assessing the extensive PAMS data from sites in the NESCAUM region. The report focuses primarily on data collected at 6 northeastern PAMS sites during two 3-day ozone episodes in July, 1994. The Committee's objectives in preparing this report include:
- develop and test techniques for PAMS data acquisition, screening, processing and exchange,
- explore the use of PAMS data for applications like control strategy development and evaluation, model emissions, chemistry and meteorology, air toxics, source attribution, airmass aging and transport,
- develop and test computer tools (such as AIRS and Voyager) for PAMS data management and analysis,
- provide access to these and other PAMS data in useful formats to a broad user community,
- seek feedback, advice and assistance from other government, academic and private sector groups and individuals with interest and/or expertise in assessing results from PAMS, NARSTO and related programs.
NESCAUM (Northeast States for Coordinated Air Use Management) - Regional Air Quality Organization representing common State Air Program interests in the States of NY, NJ, CT, RI, MA, ME, NH and VT.
OAQPS - U.S. Environmental Protection Agency's (Office of Air Quality Planning and Standards).
PAMS (Photochemical Assessment Monitoring Stations) - National monitoring program for ozone, ozone precursors and meteorology, required in serious, severe and extreme ozone non-attainment areas.
NARSTO (North American Research Strategy for Tropospheric Ozone) - Long-term, international, multi-agency and private sector research and assessment program to improve the scientific basis for policy decisions for control of tropospheric ozone.
AIRS (Aerometric Information and Retrieval System) - EPA's national aerometric database, maintained by OAQPS, and containing ambient monitoring data (including PAMS) collected by State Air Programs.
Voyager - PC-based data exploration software (freeware) developed, maintained and provided by CAPITA.
CAPITA (Center for Air Pollution Impact and Trends Analysis) at Washington University, St. Louis, MO - Maintains Voyager Software and data files, available via Mosaic or anonymous ftp at: capita.wustl.edu.
AIRS/Voyager DDS (Data Delivery System) providing for direct extraction of AIRS Air Quality data in Voyager format via NCC computer (users manuals available via EPA Technology Transfer Network).
PAVLOVS (PAMS/AIRS/Voyager Link Of Value-added Systems) - a current CAPITA project, sponsored by OAQPS and NESCAUM to facilitate extraction and Voyager analysis of PAMS data from AIRS.
VOCs (Volatile Organic Compounds) - generally refers to gaseous, non-methane organic compounds with vapor pressure > 10-2 kilopascals. As measured in PAMS, the more polar VOCs are not included , and the carbon number ranges from C-2 through C-10.
NOx (Oxides of nitrogen) - The sum of (NO) nitrogen oxide and NO2 (nitrogen dioxide). Currently, PAMS NOx includes NO (quantified accurately), plus NO2 (not quantified accurately) and variably quantities of other reactive N compounds.
NMHC (Non-Methane hydrocarbons) - gaseous hydrocarbon compounds, measured in PAMS by gas chromatography (GC).
TNMHC (Total Non-Methane Hydrocarbons) - sum of targeted and unidentified NMHCs as measured by GC in PAMS.
Carbonyls - Aldehydes, Ketones, and some organic acids and sugars, in which a carbon atom has a double bond to oxygen.
As currently measured in PAMS, carbonyls include acetaldehyde, formaldehyde and acetone.
NMOC (Non-Methane Organic Compounds) - The sum of NMHC and carbonyls.
The following analysis is exploratory in nature, and based on a limited sample of only 6 days of PAMS data of uncertain quality. The intent is to attempt some preliminary interpretations, and to seek feedback and assistance from other individuals and organizations with interest and expertise in PAMS data assessment.
Results from the PAMS monitoring program promise to provide substantial new understanding of the role of specific photochemical precursors in the formation, accumulation and destruction of ozone. These data will be useful for developing and evaluating emission control strategies for ozone and air toxics; identifying local and regional sources and source categories; evaluating and refining emissions inventories; and distinguishing among and quantifying influences of changing emissions, chemical reactions and meteorology.
Assessment of the complex, detailed PAMS data will require sophisticated computerized techniques to facilitate the access, visualization, manipulation and integration of PAMS data from multiple sites with related air quality, meteorological and emissions data.
The Voyager Data Exploration Software and AIRS Voyager Data delivery System promise to provide useful tools for State Air Programs for PAMS data visualization, manipulation, integration and assessment.
Submission of PAMS data to, and retrieval from, AIRS imposes a major obstacle to data assessment for the NESCAUM States. The alternative - exchange of PAMS data in variable formats among the States, with one State providing the service of common formatting - is also extraordinarily difficult with current tools.
The two ozone episode periods examined here both exhibited atypically high regional NOx levels on the mornings of the days with highest afternoon ozone (although other Summer 1994 ozone episodes did not).
Data collected for this report are too limited, and the validity of "simple" VOC to NOx ratios is too uncertain to draw meaningful conclusions on the relative contributions of NOx and VOC compounds to ozone formation. The current scientific paradigm suggests that both VOC and NOx reductions may be effective throughout much of the Northeast.
Targeted hydrocarbon compounds accounted for from 25% (rural) to 75% (urban) of TNMHC at Northeastern PAMS sites. Unidentified compounds accounted for 25-35% of TNMHC averages at Northeastern urban sites, and 55-75% at rural sites.
Although large sample-to-sample variations were recorded, the mix of hydrocarbon species across the region was surprisingly uniform. The most abundant anthropogenic hydrocarbons at most sites included isopentane, toluene, propane, ethane, n-butane, m/p-xylene and n-pentane. Isoprene, a biogenic hydrocarbon, was also among the most abundant targeted species at most Northeastern sites.
On a ppbc basis, acetone was the most abundant targeted carbonyl compound at all sites, followed by formaldehyde and acetaldehyde. Among the different sites, acetone levels were highest at Chicopee, MA; formaldehyde highest at Lynn, MA; and acetaldehyde highest at East Providence, RI.
On a ppbv basis, formaldehyde was the most abundant carbonyl at all sites, and exhibited a strong mid-afternoon peak, indicative of secondary (photochemical) production as the predominant production path.
On the basis of Carter's maximum incremental reactivity scheme, the most active ozone-forming VOCs in the region were formaldehyde, isoprene, acetaldehyde, m/p-xylene and toluene. Together, these 5 compounds accounted for more than 75% of the ozone formation potential during the studied episodes.
Unlike most other VOCs, the 3 most active VOC species (formaldehyde, isoprene, and acetaldehyde) often exhibit peak concentrations near mid-day. Consequently, their contributions to forming or enhancing peak afternoon ozone levels may be particularly important. Since these three compounds result from either secondary photochemistry or biogenic sources, they may not respond to traditional control strategies.
Of the targeted biogenic VOC compounds, only isoprene (emitted predominantly by deciduous vegetation) is currently quantifiable by most PAMS sites in the Northeast. Isoprene levels during the 2 (high temperature) episodes examined here were higher than anticipated (up to 25 - 35 ppbc) at several Northeastern sites. These levels are similar to concentrations reported during similar high temperature periods in the Southeast, and are considerably higher than current model estimates.
Diurnal isoprene variations at 3 NESCAUM sites peak in early afternoon, and are consistent (in timing but not magnitude) with a simple temperature dependance algorithm employed in current models. However, diurnal isoprene patterns at the other 3 NE sites are different - exhibiting morning and late afternoon peaks. This pattern is not consistent with model algorithms, but is similar to diurnal patterns observed at southeastern sites, and may be related to leaf stomatal closure during periods of mid-day moisture stress.
Both benzene and formaldehyde are ubiquitously present in urban areas in the Northeast (and other regions) at average levels which represent estimated cancer risks in a range from 1/100,000 to 1/10,000.
Short-term exposures to high levels of ozone are frequently characterized by relatively low levels of benzene, and relatively high levels of formaldehyde and particulate matter.
Many of the most abundant HCs observed at most sites during the studied episodes appear to be related to mobile source emissions, including exhaust, headspace evaporation and whole gasoline evaporation.
The PAMS data have sufficient detail to distinguish separate influences from these similar source types.
July, 1994 ratios of selected HC species at most NE urban sites are consistent with mobile source-related emissions profiles and source attribution estimates from 1990 measurements in Atlanta.
PAMS data have sufficient detail to quantify effects of airmass aging, as more reactive VOC species are photochemically destroyed during daytime transport from urban centers to downwind sites.
Two unique, non-automotive, local sources can be identified with the HC data from Chicopee, MA, and add considerably to the HC totals at this site.
Distinguishing among influences from NOx sources or source types poses an extreme challenge for the existing PAMS monitoring configuration. There is some indirect indication, through ratios of NOx (PAMS-required) to CO (not PAMS required), that stationary NOx sources contributed significantly to ozone formation during both the studied 1994 episodes.
The July 20-22/94 episode provides clear evidence of the transport of both ozone and ozone precursors (reactive HCs and NOx) along the coast of Maine from sources to the southwest. In this case, the transport of ozone itself appears to be more important than the transport of precursors.
The PAMS program would benefit from additional resources to support additional measurements, data management and data assessment tools. It is anticipated that such support might be provided by pooling resources and expertise from EPA, the States, private industry, and the academic community.
Additional emphasis should be placed on speciation of unidentified VOC compounds at sites where non-targeted compounds compose a high fraction of total VOC. In addition, the feasibility of adding additional toxic compounds to the target species list (1,3-butadiene, for example) should be considered.
Additional emphasis should be placed on quantification of biogenic VOCs, additional reactive nitrogen species, and aldehydes.
More accurate measurement methods for NO2 and NOy are needed (at least a subset of PAMS sites).
CO measurements provide a useful non-reactive tracer for mobile source exhaust, and could be helpful in the QA process, and in distinguishing among effects of mobile, area and stationary sources. CO measurement methods with improved detection limits need to be developed and deployed at (at least a subset of) PAMS urban, upwind and downwind sites.
Current PAMS measurement requirements are not well-suited to quantifying influences of stationary source NOx emissions. The program would benefit from co-located contemporaneous measurements of SO2, and fine particle concentration, composition (SO4, NO3, organics, elemental carbon, Se, Vi, As, Ni, etc.), and temporal variation (nephelometer, TEOM, etc.). These latter additions at PAMS sites would also provide valuable contributions to the understanding of fine particle sources, concentrations and effects.
Current PAMS network configuration in the Northeast is not well-suited to quantifying influences from sources to the west/southwest of the East Coast urban corridor. Establishing additional PAMS-like sites to the west of the corridor (as currently proposed by NARSTO) should be given high priority.
Upper air meteorological measurements within, and upwind of PAMS areas are critical to data assessment, and should be given high priority. Measurements of ozone and precursors at upper elevations (either surface measurements at high elevations or aircraft measurements) are also highly desirable.
The sheer volume and complexity of PAMS data demands an efficient data management system. While AIRS may ultimately provide an efficient data repository, many States continue to struggle with the efficient submission of PAMS data to AIRS. More efficient software for PAMS/AIRS data transmissions needs to be developed and provided to States by PAMS equipment vendors or EPA.
Efficient mechanisms are needed for the extraction and dissemination of multi-site PAMS data (and related air quality and meteorological data) from AIRS in formats which are useful for data analysis. The AIRS/Voyager DDS and current PAMS/AIRS/Voyager project are quite promising in this regard, but require additional modifications to extract multi-site data with mixed time interval and method codes.
It would be useful if data acquisition mechanisms similar to the AIRS/Voyager DDS could be developed to provide similar access to non-PAMS data, including:
- Surface and Upper air meteorological measurement data and calculated 3-dimensional wind fields,
- Emissions data ( with aggregation /disaggregation to assure compatibility with ambient data,
- PAMS-like air quality and met data from NARSTO sites (which may not be submitted to AIRS).
It would be useful if States could visualize and screen their data in a Voyager environment prior to AIRS submittal (i.e. develop an automated Voyager compiler to accept AIRS submittal files).
Although an important potential application of PAMS data is evaluation of emission inventories, the process of comparing ambient measurements with emission estimates is not straightforward. Both monitoring and inventory staff would benefit from better access to emission estimates which are:
- directly (or approximately) relevant to specific times and locations of ambient PAMS measurements,
- composed of VOC species (or VOC groups) which are comparable to PAMS species or groups,
- developed to include some non-reactive compounds, to help distinguish effects of different sources from effects of airmass aging.
Measurement-based source profiles (such as those developed for auto-related sources in Atlanta) are most useful for assessing PAMS data. States would benefit from better access to measurement-based emissions profile information of this nature. Ideally, characteristics of these profiles would include:
- local measurements for each PAMS area, periodically updated as justified by fuel changes etc.,
- measurement methods and targeted species similar to those required at PAMS sites,
- representative range of local (and if relevant, distant) mobile, area, and stationary source types.
Although an important potential application of PAMS data is improving the meteorology, chemistry and emissions components of photochemical grid models, such work will require many years of development, and is well beyond the technical and financial capabilities of most State Air Programs. Less complex, less costly methods for evaluating control strategy alternatives from PAMS data need to be developed and provided to the States.
As larger data sets are collected, PAMS data may be used to empirically derive simple estimates of the relative effectiveness of VOC and NOx controls. However, several key uncertainties require clarification before the use of VOC to NOx ratios can provide useful information for control strategy development:
- How large (important) are uncertainties in the accuracy and precision of current NOx measurements?
- Exactly which VOC compounds should be considered for comparison with NOx concentrations?
- Exactly which time intervals should be included in a comparison of VOC and NOx concentrations?
- How should differential reactivities of different compounds be included in calculation of total VOC?
Last modified Wednesday, September 06 1995