Abstract. An expression for the production rate of 03, P(O 3), is derived based on a radical budget equation applicable to low and high NOx conditions. Differentiation of this equation with respect to NO or hydrocarbons (HC) gives an approximate analytic formula in which the relative sensitivity of P(O3) to changes in NO or HC depends only on the fraction of radicals which are removed by reactions with NOx. This formula is tested by comparison with results from a photochemical calculation driven by trace gas observations from the 1995 Southern Oxidants Study (SOS) campaign in Nashville, Tennessee.

Trace gas measurements pertinent to understanding the transport and photochemical formation of 03 were made at a surface site in rural Georgia as part of the Southern Oxidant Study during the summer of 1991. It was found that there was a strong correlation between 03 and the oxidation products of NOx: O3(ppb) = 27 + 2 11.4 (NOy(ppb) -NOx(ppb)), r = 0.78. This fit is similar to that observed at other rural sites in eastern North America and indicates a nominal background 03 level of 27 ppb; values higher than 27 ppb are due to photochemical production in the recent past, which varied from near zero to -•50 ppb. The origin of the 03 above background was investigated by using a free radical budget equation to calculate an in situ 03 production rate in terms of measured concentrations of NO and free radical precursors (03, HCHO, peroxides, and other carbonyls). A comparison of observed and predicted diurnal trends in 0 3 indicates significant 03 production in the afternoon at a time when 03 concentration is either steady or decreasing. The afternoon near-surface layer is thereby a source region for 03 which can be exported. In situ production accounts for approximately one half of the morning increase in 03 concentration on days with high 03; the remainder is due to entrainment of dirty air aloft by the growing convective boundary layer. Additional evidence for the role of vertical transport in controlling the hour-to-hour changes in 03 is found in the diurnal cycles of SO2 and HNO3 which also have rapid increases in the morning. The day-to-day variability of 03 was investigated using a back trajectory model. NOy concentration at the measurement site could be reasonably accounted for by considering NOx emission sources located within 1-day transport distance. In as much as there is a strong correlation between 03 and NOy, the coincidence between trajectory location and NOx emission sources appears to t•e an important factor influencing midday 03 concentration. Hydrocarbon measurements are consistent with NOx being the limiting factor for formation of 03. 20-30 ppb. During pollution episodes, 03 levels in excess ofthe 120 ppb National Ambient Air Quality Standard have been measured at rural sites [Meagher et al.southeastern United States differs from more industrialized and populated regions in that NOx emissions are lower and natural HC emissions are higher. In addition to precursor emission rates the formation of 03 depends on meteorological conditions. An active photochemistry is favored by high solar intensity, temperature, and absolute humidity which are common summertime conditions in the southeastern United States. Stagnation episodes, which are also common, allow emitted pollutants and their photochemically produced reaction products to accumulate over a several day period. Considerable progress has been made in using photochemical models to simulate the production of 03 and the effects of emission changes [e.g., Seinfeld, 1988; McKeen et al., 1991a, b; NRC, 1991; Roselle et al., 1991]. However, the coupled em...

[1] Aerosol chemical composition, size distribution, and optical properties were measured during 17 aircraft flights in New England and Middle Atlantic States as part of the summer 2002 New England Air Quality Study field campaign. An Aerodyne aerosol mass spectrometer (AMS) was operated with a measurement cycle of 30 s, about an order of magnitude faster than used for ground-based measurements. Noise levels within a single measurement period were sub mg m À3 . Volume data derived from the AMS were compared with volume measurements from a Passive Cavity Aerosol Spectrometer (PCASP) optical particle detector and a Twin Scanning Electrical Mobility Spectrometer (TSEMS); calculated light scattering was compared with measured values from an integrating nephelometer. The median ratio for AMS/TSEMS volume was 1.25 (1.33 with an estimated refractory component); the median ratio for AMS/nephelometer scattering was 1.18. A dependence of the AMS collection efficiency on aerosol acidity was quantified by a comparison between AMS and PCASP volumes in two high sulfate plumes. For the entire field campaign, the average aerosol concentration was 11 mg m À3 . Compared with monitoring data from the IMPROVE network, the organic component made up a large fraction of total mass, varying from 70% in clean air to 40% in high concentration sulfate plumes. In combination with other optical and chemical measurements, the AMS gave information on secondary organic aerosol (SOA) production and the time evolution of aerosol light absorption. CO is taken as a conservative tracer of urban emissions and the ratios of organic aerosol and aerosol light absorption to CO examined as a function of photochemical age. Comparisons were made to ratios determined from surface measurements under conditions of minimal atmospheric processing. In air masses in which the NO x to NO y ratio has decreased to 10%, the ratio of organic aerosol to CO has quadrupled indicating that 75% of the organic aerosol is secondary. Also, the ratio of light absorption to CO has more than doubled, which is interpreted as an equivalent increase in the light absorption efficiency of black carbon due to aerosol ageing.

[1] Ozone production efficiency can be defined as the number of molecules of oxidant (O 3 + NO 2 ) produced photochemically when a molecule of NO x (NO + NO 2 ) is oxidized. It conveys information about the conditions under which O 3 is formed and is an important parameter to consider when evaluating impacts from NO x emission sources. We present calculational and observational results on ozone production efficiency based on measurements made from aircraft flights in the Phoenix metropolitan area in May and June of 1998. Constrained steady state box model calculations are used to relate a ratio of O 3 production rate to NO x consumption rate (i.e., P(O 3 )/P(NO z )) to a VOC to NO 2 ratio of OH reactivity. Lagrangian calculations show how this ratio generally increases with time due to oxidation chemistry and plume dilution. City to city differences in ozone production efficiency can be attributed to corresponding differences in VOC to NO 2 reactivity ratio which in turn reflect emission patterns. Ozone production efficiencies derived from aircraft measurements in 20 plumes show a dependence on NO x concentration similar to that calculated for P(O 3 )/P(NO z ). Calculations are based on data from a single location but are believed to be applicable to a wide range of plumes from different areas.

The rate of uptake of NO2 by liquid water according to (R1), 2NO2(g) + H2O(1) → 2H+ + NO3− + NO2−, is shown to be unaffected by O2 (0.2 atm). Hence the rate constant and Henry's law solubility constant of NO2 previously obtained may be employed to evaluate the rates of aqueous phase reactions of NO2 in the ambient atmosphere. Reactions (R1) and (R2), NO2(g) + NO(g) + H2O(1) → 2H+ + 2NO2−, are quite slow at representative atmospheric partial pressures and cloud liquid water content; the characteristic times range upward from 103–104 hours at 10−7 atm, increasing with decreasing partial pressures of the gases. Direct acidification of cloud liquid water by (R1) or (R2) is also unimportant. Catalytic enhancement of (R1) is potentially important for catalyst concentrations of order 10−7 M, assuming sufficiently fast rate constants (∼108 M−l s−1). Iron‐catalyzed reaction in particular, however, is found to be unimportant. Reaction of NO2 with dissolved S(IV) is potentially mportant, based upon an assumed upper limit rate constant of 2.5×107 M−1 s−1. Deposition of NO2 to surface (ocean or lake) water is shown to be controlled by aqueous phase mass transport and/or reaction and is much slower than heretofore assumed.

Atmospheric concentrations of a series of carbonyl compounds known as formaldehyde (FA), acetaldehyde (AC), acetone (AN), glycolaldehyde (GA), glyoxal (GL), methylglyoxal (MG), glyoxylic acid (GD), and pyruvic acid (PD) were measured at a rural site in Georgia in summers of 1991 and 1992. The midafternoon median concentrations, in parts per billion, determined for 1991–1992 were FA, 3.6/3.1; AC, 0.58/0.74; AN, 1.7/1.8; GA, 0.21/0.26; GL, 0.02/0.09; MG, 0.03/0.08; GD, 0.46; and PD, 0.11, the latter two for 1992 only. All of the carbonyls except AC and AN exhibited a strong diurnal dependence, with maxima in the midafternoon and minima during the night, consistent with a rapid in situ photochemical production in the daytime and a loss by dry deposition in a shallow inversion during the night. FA correlated well with O3, GA and MG, consistent with their photochemical production near the surface at the measurement site. GL and MG showed the strongest correlation among all species, suggesting common origins as well as similar atmospheric lifetimes. The presence of GA, MG, and GL along with FA at the observed relative concentrations are consistent with laboratory developed isoprene oxidation mechanism and the expectation that isoprene represents a major reactive hydrocarbon in this rural region. At the concentrations observed, these carbonyls serve as important radical sources. The contribution of FA accounts for half of that by O3 and the higher carbonyls approximates half of that by FA. With respect to production of peroxyacetyl nitrate, isoprene contributes as much as acetaldehyde. These results lend further credence to the notion that isoprene plays a pivotal role in photochemical processes, especially in rural environments.

Steady state photochemical calculations were performed using observed or estimated trace gas concentrations as constraints. According to these calculations the local rate of 03 production P(O3) in all four plumes is VOC sensitive, sometimes strongly so. The local sensitivity calculations show that a specified fractional decrease in VOC concentration yields a similar magnitude fractional decrease in P(O3). Imposing a decrease in NO•, however, causes P(O3) to increase. The question of primary interest from a regulatory point of view is the sensitivity of 03 concentration to changes in emissions of NO• and VOCs. A qualitative argument is given that suggests that the total 03 formed in the plume, which depends on the entire time evolution of the plume, is also VOC sensitive. Indicator ratios O3/NO z and H202/NO z mainly support the conclusion that plume 03 is VOC sensitive.

The concentration of formaldehyde at Mauna Loa Observatory, Hawaii, was determined during four Mauna Loa Observatory Photochemistry Experiment 2 (MLOPEX 2) measurement intensives between September 1991 and August 1992. The observed diurnal variations, 200-900 parts per trillion by volume (pptv) during daytime and 60-200 pptv during nighttime, resulted mainly from the local air circulation pattern whereby island modified marine boundary layer air prevailed during the day and free tropospheric air dominated during the night. A seasonal variation was also observed; the median/mean values of all data points are: 149/196, 129/149, 143/178, and 181/211 pptv for autumn, winter, spring, and summer intensives, respectively. During nighttime downslope flow periods which brought in free tropospheric air to the measurement site, the formaldehyde concentrations (median/mean) were 122/123, 110/112, 120/125, and 140/137 pptv for autumn, winter, spring, and summer, respectively. This seasonal dependence may be attributable to changes in solar insolation and NO concentrations. A simple box model calculation constrained by the experimentally determined concentrations of CH3OOH yielded a formaldehyde concentration (without/with heterogeneous removal) for free tropospheric air, at 7øC, of 155/140, 125/115,210/195, and 220/205 pptv for autumn, winter, spring and summer, respectively. The calculated values are in good agreement with the measured concentrations for winter (within 27/15%, without/with heterogeneous removal) and fall (within 14/5%), but are significantly higher for spring (75/63%) and summer (57/46%).