HETEROGENEOUS NIGHTTIME PRODUCTION AND DEPOSITION OF PARTICLE NITRATE AT A RURAL SITE IN NORTH-AMERICA DURING SUMMER 1988

Measurements at a rural site in central Ontario, Canada, in August 1988 showed gradual increases in particle nitrate (pNO3-) concentration from 1800 hours to midnight and a decrease thereafterward toward the early morning. A time-dependent Lagrangian box model, which included homogeneous and heterogeneous chemical reactions of nitrogen species, dry deposition Of 03, N02, HNO3 and pNO3-, and sources of NO(x), was used to simulate the observed nighttime increase of pNO3- for four nights in August. The model results indicate that the nighttime pNO3- was mainly formed by reactions of the nitrate radical (NO3) and N2O5 with particles, with these species contributing on average more than 80% and about 10% of the total nightly pNO3- formation, respectively. Conversion of HNO3 to pNO3-was found to be negligible compared to these two processes, contributing less than 5%. By assuming steady-state concentrations, a simplified approach to fit nighttime pNO3- concentrations using only NO(x) and 03 observations gave results similar to the box model. As expected, the pseudo-first-order rate constants for the formation of pNO3- on particles from N03 and N2O5 were found to be linear functions of the particle mass. When assuming an extended Junge distribution for the particles, the rate constants indicated a near unity sticking coefficient (alpha1) for N03 but a substantially smaller sticking coefficient (alpha2) of about 0.02 for N2O5, consistent with results from laboratory studies. Sensitivity tests showed that the ranges of 0.8-1.0 for al and 0.01-0.04 for alpha2 gave reasonable model fits for pNO3- and HNO3. The box model showed that the heterogeneous reactions were the major sinks for N03 and N2O5. The dry deposition velocities for 03, N02, pNO3-, and HNO3 derived from the model simulations were similar to observed values over similar vegetation. The dry deposition rate constant of pNO3- appeared to be a function of the near-surface wind speed, although the dry deposition velocity of pNO3- was quite variable. The box model predicted similar amounts of dry deposition for HNO3 and pNO3- during each of the four nights.