Particle deposition on water: Surface source versus upwind source

The deposition velocity for aerosols on natural water surfaces is investigated with emphasis (1) on the difference between the deposition of particles that originate above the surface (or upwind) and those which originate from a local uniform surface source such as sea-salt aerosol, and (2) on extending the deposition velocity formulation into the size range where the gravitationally induced gradients require corrections to the well-mixed case. The following cases are explored: (1) uniform surface source in equilibrium with deposition, (2) surface source with removal above the surface layer in addition to surface deposition, (3) surface deposition when the source is from above (upwind), and (4) surface deposition when the turbulent flux is downward but opposed by a surface source. The following are shown: (1) For a surface source of aerosols, the problem can be formulated such that the deposition processes occurring within the molecular interfacial layer do not enter the expression for the deposition velocity, provided the surface source function and deposition are defined for a point at or above the top of the molecular interfacial layer. This is useful because measurements of the surface source function are made above the interface layer and the microscopic processes occurring in the interface layer are poorly characterized. (2) For a surface source with removal above the surface layer (net upward flux), the deposition velocity is a function of the removal height and strength of the removal process. However, if the reference height is much lower than the removal height, the equilibrium deposition velocity is an acceptable approximation. (3) If the filling/emptying time for an aerosol is much longer than the turbulent mixing time, as is usually the case in the boundary layer, a quasi steady state is established. The deposition velocity for the quasi steady state is the same as for steady state, making the steady state deposition velocity useful for dynamic modeling. The results of this study have important implications on how sea-salt particles are introduced and removed from the lowest layer of numerical models, as well as retrieval of the surface source function from measurements of the size distribution at some reference height.