Preview of 1994 Ozone Precursor Concentrations in the Norteastern U.S.
The data featured in this report have been primarily from sites in the New England Region. Some PAMS sampling was also conducted during the studied episode periods at sites in New Jersey (Ryder University) and New York City (New York Botanical Gardens). The NJ data were limited to only one day during each of the studied 3-day episode periods - and have not been included in most of the preceding regional analyses. Sampling at the NY site was conducted on a forecast basis, and included sampling on 5 of the 6 days examined here (good forecasting!). These data were not available when the regional data sets were prepared for this report. However, the NY data for these and other 1994 episode periods have recently become available, and are posted in AIRS, and on the NESCAUM Bulletin Board and CAPITA ftp server (ny94pams.zip). While a detailed analysis and regional comparisons with these NY data are not possible at the present time, a quick review of the NY data provides an interesting contrast to concurrent measurements from the New England PAMS sites.
During the regional episode days, 3-hour VOC canister samples for the NY Botanical Gardens (NYBG) site did not include sampling between Midnight and 6 AM, and from Noon to 3 PM (times frequently associated with relatively high and relatively low VOC levels respectively). With this caveat, the NYBG samples collected on July 7, 8, 20, 21 and 22, 1995 are approximately comparable to the data from the New England sites during these dates. Figure 8.1 compares the average NYBG hydrocarbon values during the studied episodes with New England regional concentrations (from Table 5.1). Only compounds with average New England Concentrations in excess of 1.0 ppbc are included. While this comparison is crude, several interesting observations can be drawn:
Figure 8.1 Comparison of VOC Concentrations at NY Botanical Gardens with New England Regional Averages During July, 1994 Episode Periods
In comparison to the New England averages, the NYBG data fall into 3 general categories:
- NYBG concentrations roughly equal to New England concentrations,
- NYBG concentrations roughly 3 times New England concentrations,
- NYBG concentrations roughly 6 times New England concentrations,
For 4 compounds - commonly not associated with automotive-related sources - the NYBG averages are quite similar to the New England values. Average levels of styrene, methylcyclohexane, propane and isoprene were roughly the same at NYBG and the New England sites during these episodes. For isoprene, the NYBG averages are missing values from Midnight to 6 AM (typically a low isoprene period) and also from Noon to 3 PM (typically a high isoprene period), but the NYBG isoprene levels during other time periods are surprisingly comparable to the New England data.
It might be anticipated that isoprene emissions in a heavily urbanized area like NY City would be relatively limited. However, recent SOS research (Cardelino and Chameides, 1990)12 suggests that a 20% reduction in Atlanta's forests over the 1979-1988 decade actually increased isoprene emissions. The loss of trees was offset by increased emission effects of (urban heat island) temperature increases. Local biogenic emissions from the Botanical Gardens themselves may exert a significant influence here. It should also be noted that (unlike the automated GC sampling at the New England sites) the 1994 NYBG samples were collected in canisters. It has been suggested that (highly reactive) isoprene may be depleted in canisters - such that the 1994 NYBG values may represent low-bound estimates. Planned 1995 NY sampling will include concurrent sampling with auto GC and canisters - and should provide valuable insight into potential problems (if any) associated with canister sampling.
For most compounds - particularly those associated with automotive-related emissions - levels at NYBG were approximately 3 times higher than the New England averages. This is not surprising, given relatively high NY City automotive emissions densities. While the NYBG levels for these compounds are higher than the New England measurements, it is noteworthy that they are consistently (3x) higher for many compounds. Figure 8.2 compares benzene and toluene at NY and NE sites during the selected episode periods. The NE data excludes results from Chicopee, MA, where a local, non-auto-related toluene source is suspected (Section 6.3). While the NY levels are generally higher than NE, benzene and toluene are highly correlated in both data sets, with nearly identical B/T ratios.
Figure 8.2 Benzene vs. Toluene at New York and New England Sites During 7/94 Episodes
The slightly higher slope for the NE data, may be due to the inclusion of the rural Stafford, CT and Cape Elizabeth, ME sites in the plotted NE data. Effects of aging at these rural sites would tend to increase the B/T ratios (note how the highest NE B/T ratios in Figure 8.2 occur at the lowest concentrations).
With this exception, the benzene/toluene relationships among all of these sites appear to be of a common distribution - with a slope consistent with automotive (and other gasoline combustion) sources.
Acetylene and ethene have historically been cited as relatively unique tracers for auto tailpipe emissions. These compounds frequently exhibit among the strongest correlations of any paired species in VOC measurement programs at many sites and time periods. Given this, and the apparently strong influence of automotive emissions at all the New England sites, and at the NYBG site in particular, it is surprising that acetylene and ethene were uniquely higher (about 6x) at NYBG in comparison to the NE average.
Figure 8.3 compares acetylene to ethene (A/E) ratios from Atlanta, 1990 (5 sites) and from a Chicago 1993 PAMS site. While both data sets exhibit strong E/T correlations, the Chicago slope is notably lower. Acetylene emissions are substantially reduced by current generation catalysts compared to pre-1986 and non-catalyst vehicles.13
Figure 8.3 Acetylene vs. Ethene from Atlanta, 1990 and Chicago PAMS,, 1993
Conceivably the shift in slopes may be related to different sampling years. Acetylene emissions from newer model autos are enhanced by cold start operations13 - so sites with strong influence from cold start emissions might show consistently Higher A/E ratios than sites more heavily influenced by highway emissions. Ethene is more reactive than acetylene, such that relative concentrations could also be influenced by the degree of photochemical aging.
Acetylene and ethene can also pose measurement challenges for current generation GC technology. Figure 8.4 shows A/E ratios for the 1994 NE sites, with E. Hartford values displayed separately. The A/E ratios are quite consistent (R2 = 0.77) from all sites except E. Hartford. The New England A/E slope (A = 0.77E) falls almost exactly midway between the (Figure 8.3) values of 1.05 for 1990 Atlanta and 0.64 for 1993 Chicago. The A/E ratios from E. Hartford are clearly different from the other NE sites, and in this case, a measurement error (or misidentification) for ethene is clearly indicated - by the high number of low or zero values for ethene at a wide range of acetylene concentrations.
Figure 8.4 Acetylene vs. Ethene at NE PAMS Sites During 7/94 Episodes
The changing slopes of A/E ratios among different sites and time periods raise questions about use of these compounds as quantitative tracers for automotive exhaust.13 At the same time, the strong A/E correlations at widely scattered sites and time periods, and the relative absence of other substantial large sources of these compounds suggest that they may provide excellent qualitative tracers of auto exhaust at specific sites and time periods. In the example above, the clear difference in relative concentrations at the East Hartford site helps identify a measurement error, and illustrates the way in which exploratory analysis of PAMS data can contribute to the quality assurance process.
Figure 8.5 compares the relatively high A and E values at the NYBG site with the comparable NE data (with E. Hartford excluded). While the slopes are quite similar, the correlation is much poorer at NYBG. In the case of one extreme 7/7/94 outlier (A = 7.3E), local welding activities at NYBG are the suspected source of the high acetylene levels. Whether this, other local sources, or measurement errors are responsible for the high levels and poor A/E correlation at NYBG is a question for future study.
Figure 8.5 Acetylene vs. Ethene at NY and NE sites During July, 1994 Episodes
