Long-term aerosol particle flux observations part I: Uncertainties and time-average statistics

Abstract

Long-term eddy covariance particle flux measurements for the size range starting from 10 nm were performed at a boreal forest site in Southern Finland. The large variability in turbulent flux estimates is inherent to the particle flux observations and thus long-term particle flux measurements enable to obtain statistically significant results by a suitable averaging. The particle flux random errors were estimated and a parameterisation for the integral time scale of turbulent flux was proposed. Application of flux errors for classification according to statistical significance of single flux values leads to systematically different deposition estimates on ensemble average basis. This must be avoided for determination of unbiased average deposition fluxes. The role of storage term in particle deposition evaluation was analysed. It was empirically determined that the method of storage term estimation discussed by [Finnigan, J., 2006. The storage term in eddy flux calculations. Agric. Forest Meteorol., 136, 108–113.] is not sensitive to the selection of the concentration averaging window in both ends of the flux averaging period. It is argued that the storage change in real atmospheric conditions results from boundary layer development as well as source–sink activity and therefore the filtering effect arising from averaging the concentration is of less importance. Diurnal, seasonal and annual variability of particle fluxes was analysed and it was observed that particle deposition rates are higher in winter. More detailed analysis of functional dependencies of particle deposition on environmental factors as well as dependence on size will be done in the second part of the paper.