|
Aerosols play an important part in determining optical properties of the atmosphere over the ocean. Problems of measuring aerosol properties and the need to forecast them, have led to the development of model simulations. At present, the models rely on standard meteorological variables as input and produce aerosol size distribution. These models are generally dependent on the knowledge of wind, relative humidity, visibility and altitude of the observation site, however they do not take into account the history of the airmass and the associated generation and transport processes of aerosol. As a result such predicted aerosol size distributions have a high variability when compared to individual measurements. Their value is therefore limited to climatological applications. Current studies consider the generation, transport and decay processes. In the generation process bursting bubbles during white cap conditions are considered. This process is important for particles larger than about 1 micron diameter. Smaller particles are generated mainly through gas-to-particle conversion processes over continents. Because of their relatively long life time in the atmosphere, these particles cannot meteorologically be traced to their origin. Transport processes are fairly well described through dynamic boundary layer models, bulk formulas for vertical transport, and general circulation models for horizontal air trajectories. Dissipation processes include coagulation for small particles, gravitational settling for particles larger than 5 micron diameter and turbulent transport through the upper boundary of the mixed layer. These studies serve to delineate appropriate meteorological variables which can serve as inputs to a dynamic aerosol model. For a practical use such dynamic models are still too complex and simplifications are needed. A compromise is dictated by the type and accuracy of available meteorological input data. At present white cap coverage, depth of mixed layer, airmass type and travel time of the air over water are considered as additional appropriate model inputs besides wind and relative humidity.
|
