Scavenging ratios, wet deposition, and in‐cloud oxidation: An application to the oxides of sulphur and nitrogen

With special regard to the class of substances in precipitation that potentially originate from both gaseous and particulate precursors, the process of precipitation scavenging is discussed. A model relating daily average ground level scavenging ratios ( W ) to nucleation scavenging and in‐cloud chemical transformation is introduced and is used as guidance in the regression analysis of 3 years of SO 4 = and NO 3 − scavenging‐ratio observations made at six locations in eastern Canada. It was found that W of SO 4 = ‐, NO 3 − ‐, and SO 4 = ‐bearing particles is inversely proportional to the one‐third power of the precipitation amount in the event. The best regression model explained 41% of the variance in log W for SO 4 = . It included the effects of location, precipitation amount, precipitation type, and in‐cloud SO 2 oxidation. The last effect accounted for 50% of the variance explained. The analysis predicts that, on average, in‐cloud SO 2 oxidation accounts for 42–79% of the SO 4 = observed in rain and with the exception of one site, less than 20% of the SO 4 = observed in snow. These results are consistent with a mechanism of SO 2 oxidation involving photochemically produced H 2 O 2 . A similar analysis for NO 3 − supports the hypothesis that throughout the year much of the NO 3 − in precipitation originates from in‐cloud NO 2 oxidation. It suggests that, depending on location, oxidation of 0.5–1.2 ppbv of NO 2 is sufficient to explain observations. One possible mechanism of oxidation is the reaction with O 3 to form NO 3 and hence soluble N 2 O 5 . These results, combined with observations in the eastern United States, indicate that in eastern North America SO 2 oxidation within a rain storm becomes an increasingly important source of rain SO 4 = as latitude increases. Scavenging ratios are useful tools for investigating in‐cloud conversion and for parameterizing precipitation removal. However, they are limited to describing the average characteristic of an ensemble of storms. They cannot be used with confidence to analyze individual events.