Model considerations for the use of passive sampling techniques to determine trace gas concentrations and flux densities

A number of theoretical and practical difficulties arise in the calculation of flux densities for chemical species which are measured using instrumentation with a low resolution in time. Whenever passive samplers, denuders or filters are used to measure the atmospheric concentrations required, the use of mean concentrations and mean wind velocities may lead to incorrect mean fluxes due to the neglection of the averaging rules for fluctuating entities. We investigated the potential biases resulting from calculations of average concentrations and vertical flux densities for three types of passive samplers, two of which are sensitive to wind velocity, using realistic data sets for concentrations, flux densities and meteorological parameters obtained from comprehensive field measurements. The detection of average concentrations with wind sensitive samplers and anemometers requires adequate 4sampling times to minimize the bias. They are applicable when the resolution in time is weeks rather than days. Flux density measurements yield unsatisfactory results with all types of samplers unless several pairs of samplers can be operated separately depending upon the atmospheric conductivity.