| description abstract | The relative proportions of particles collected on the filter by the three individual mechanisms of impaction, diffusion (shallow or deep) and direct interception is shown to be important for the membrane (and Nuclepore) filter method for measurement of ice nuclei concentrations. Calculations of partial efficiencies for these mechanisms, and total efficiencies, were made for one Millipore filter and one Nuclepore filter, for four particle sizes and two airflow rates. It is suggested as a consequence that it is not the very small ice nuclei whose concentrations are likely to be significantly underestimated, as suggested by some workers, but those in the intermediate size range (r ? 100 nm). The maximum underestimate of these nucleus concentrations, due to some of them not being captured near the top surface of the filter, should not be greater than about 50% in the cases considered. It is argued that the sampled volume effect may vary with the type of filter used. There are brief discussions of possible effects due to the use of petrolatum to seal filter pores, pore clogging, electrostatic forces, variations in particle density and pore size distribution. Possible advantages to be gained in using filters of relatively low porosity, such as some Nuclepore filters, are suggested. | |
