OH and HO2 chemistry in the urban atmosphere of New York City

Observed hydroxyl (OH) and hydroperoxy (HO2) radicals, collectively called HOx, were compared with OH and HO2 calculated by a box model that used the regional atmospheric chemistry mechanism and was constrained to the ancillary measurements during the PM2.5 Technology Assessment and Characterization Study-New York (PMTACS-NY) summer 2001 intensive in New York City. The measurements are described in the companion paper, Ren et al. (HOx concentrations and OH reactivity observations in New York City during PMTACS-NY2001, Atmospheric Environment, this issue). This comparison enables an investigation of HOx chemistry in this polluted urban atmosphere. For HO2, the observed concentrations and diurnal variation were usually well reproduced by the model calculations, with an observed-to-modeled ratio of 1.24, on average, for day and night. For OH, the model was generally able to match the measured concentrations during daytime with an observed-to-modeled ratio of about 1.10, but the calculations significantly underestimated OH during nighttime. The budgets of HOx show that its production was dominated by the photolysis of HONO, accounting for similar to56% of HOx production on average, during daytime due to relatively high HONO concentrations, while nighttime HOx production was mainly from the O-3 reactions with alkenes. The OH reactivity measurements agree with the calculations to within 10% for both the composite diurnal variation and individual days. Calculations indicate that the reactions of OH with NO2, hydrocarbons, CO, NO, and carbonyls accounted for about 32%, 25%, 12%, 10% and 7% of total OH loss, respectively, in this urban area. Modeled instantaneous O-3 production from HO2 and RO2 reactions with NO was 150+/-100 ppbv day(-1). O-3 production rates from measured HO2(P(O-3)(obs)(HO2)) was greater than modeled HO2(P(O-3)(calc)(HO2)) at higher values of NO. Average daily cumulative P(O-3)(obs)(HO2) was similar to140 ppbv day(-1), a factor of 1.5, greater than average daily P(O-3)(calc)(HO2). (C) 2003 Elsevier Ltd. All rights reserved.