An evaluation of reaction probabilities of sulfate and nitrate precursors onto East Asian dust particles

Abstract

We investigate the chemical evolution of dust particles and examine the magnitudes of reaction probability (γ) of sulfate and nitrate precursors (such as H2SO4, SO2, N2O5, and HNO3) onto dust particles in East Asia. For this investigation, three data sets from ACE‐Asia U.S. National Science Foundation (NSF) C130 flights 6, 7, and 8 and three data sets from measurements in Seoul were analyzed. During the selected dust storm periods, large amounts of CO32− still remained in fine‐mode dust particles (Dp < ∼1.3 μm). The average fractions of CO32− (based on [Ca2+] + [Mg2+] equivalences) for the three C130 flights were 0.87, 0.68, and 0.39, respectively, and the average CO32− fractions for the three data sets in Seoul ranged from 0.43 to 0.86. The degrees of chemical evolution of mineral dust indicated by the CO32− fractions were significantly smaller than those reported in previous dust chemistry transport modeling studies conducted with the assumption of aerosol internal mixing. Our analysis suggests that this could be due to the use of excessively high γ values in the model simulations, as further confirmed in this study with Lagrangian photochemical model simulations conducted with both the constraints observed by ACE‐Asia C130 flights and the initial concentrations obtained by U.S.‐EPA Models‐3/Community Multi‐scale Air Quality (CMAQ) modeling over East Asia. It is also found in this study that the magnitudes of γs are closely related with aerosol mixing state. In order to confirm this we conducted Lagrangian model simulations for an example case under the assumption of aerosol external mixing. Under this assumption the formations of sulfate and nitrate on/in fine‐mode mineral dust are greatly limited because of small γs onto mineral dust (γDust) and low precursor concentrations. However, it is also recommended that for more precise evaluations of γDust, a more sophisticated Lagrangian type of photochemical modeling is necessary. In addition, in this study a possible dependence of the γ values on relative humidity (RH) is also investigated. On the basis of our analysis we suggest that particular attention should be paid to the issue of the RH‐dependent γ in future dust chemistry transport modeling studies with the issue of aerosol mixing state.