Hydroxyl and peroxy radical chemistry in a rural area of Central Pennsylvania: Observations and model comparisons

Atmospheric hydroxyl (OH), hydroperoxy (HO2), total peroxy (HO2 and organic peroxy radicals, RO2) mixing ratios and OH reactivity (first order OH loss rate) were measured at a rural site in central Pennsylvania during May and June 2002. OH and HO2 mixing ratios were measured with laser induced fluorescence (LIF); HO2+RO2 mixing ratios were measured with chemical ionization mass spectrometry (CIMS). The daytime maximum mixing ratios were up to 0.6 parts per trillion by volume (pptv) for OH, 30 pptv for HO2, and 45 pptv for HO2+RO2. A parameterized RACM (Regional Atmospheric Chemistry Mechanism) box model was used to predict steady state OH, HO2 and HO2+RO2 concentrations by constraining the model to the measured OH reactivity and previously measured volatile organic compound (VOC) distributions. The averaged model calculations are generally in good agreement with the observations. For OH, the model matched the observations for day and night, with an average observed-to-modeled ratio of 0.80. In previous studies such as PROPHET98, nighttime NO was near 0 pptv and observed nighttime OH was significantly larger than modeled OH. In this study, nighttime observed and modeled OH agree to within measurement and model uncertainties because the main source of the nighttime OH was the reaction HO2+NO -> OH+NO2, with the NO being continually emitted from the surrounding fertilized corn field. The observed-to-modeled ratio for HO2 is 1.0 on average, although daytime HO2 is underpredicted by a factor of 1.2 and nighttime HO2 is over-predicted by a factor of similar to 2. The average measured and modeled HO2+RO2 agree well during daytime, but the modeled value is about twice the measured value during nighttime. While measured HO2+RO2 values agree with modeled values for NO mixing ratios less than a few parts per billion by volume (ppbv), it increases substantially above the expected value for NO greater than a few ppbv. This observation of the higher-than-expected HO2+RO2 with the CIMS technique confirms the observed increase of HO2 above expected values at higher NO mixing ratios in HO2 measurements with the LIF technique. The maximum instantaneous O-3 production rate calculated from HO2 and RO2 reactions with NO was as high as 10-15 ppb h(-1) at midday; the total daily O-3 production varied from 13 to 113 ppbv d(-1) and was 48 ppbv d(-1) on average during this campaign.