No abstract

[1] Dry deposition of atmospheric particles is critically dependent on particle size and plays a key role in dictating the mass and number distributions of atmospheric particles. However, modeling dry deposition is constrained by a lack of understanding of controlling dependencies and accurate size-resolved observations. We present sizeresolved particle number fluxes for sub-100-nm particle diameters (Dp) over a deciduous forest derived using eddy covariance applied to data from a fast mobility particle sizer. The size-resolved particle number fluxes in 18 diameters between 8 and 100 nm were collected during leaf-on and are statistically robust. Particle deposition velocities normalized by friction velocity (v d + ) are approximately four times smaller than comparable values for coniferous forests reported elsewhere. Comparison of the data with output from a new one-dimensional mechanistic particle deposition model designed for broadleaf forest exhibits greater accord with the measurements than two previous analytical models, but modeled v d + underestimate observed values by at least a factor of two for all Dp between 6 and 100 nm. When size-resolved particle deposition velocities for Dp < 100 nm are normalized by friction velocity, the key controlling role of particle diffusivity is strongly manifest. On the basis of analyses of these new measurements and recently published sizeresolved particle number fluxes from a conifer forest, we present working parameterizations for size-resolved particle deposition velocities over forests that could reasonably be applied in regional and global atmospheric chemistry transport models.

Ion concentrations in relatively low-intensity precipitation were measured in southern Indiana, USA and are presented as a function of their temporal evolution during individual precipitation events with a specific focus on the first 30 min of those events. These data indicate that during individual rain events potassium concentrations in precipitation may decline by up to 70%-80% in the first 30 min of the event. The other ions exhibited less rapid concentration declines during this event which are in rank order (highest to lowest); sodium, chloride, magnesium, nitrate, calcium, sulfate and ammonium. There is some evidence that the initial declines for precipitation accumulations up to 2 mm in the concentration of chloride, calcium and sulfate in precipitation more closely approximate a power-law dependency on precipitation depth than the commonly applied exponential form which, if confirmed, may have implications for efforts to correct flux networks for under-sampling due to delay in sample collection. Scavenging coefficients (b) derived using an exponential relationship over entire events for sodium, chloride, nitrate, calcium, sulfate and ammonium indicate highest values for sodium and lowest for ammonium, but the uncertainty bounds on ion-specific values of b are sufficiently large that they are statistically indistinguishable.