Volume 8, No. 4, December 2008, Pages 392-410 PDF(611 KB)
Comparison of Two Approaches to Modeling Atmospheric Aerosol Particle Size Distributions
Vladimír Ždímal1, Marek Brabec2,3, Zdeněk Wagner4
1 Laboratory of Aerosol Chemistry and Physics, Institute of Chemical Process Fundamentals of the AS CR, v. v. i., Rozvojová 135, Praha 6, 165 02, Czech Republic
2 Department of Biostatistics and Informatics, National Institute of Public Health, Šrobárova 48, Praha 10, 100 42
3 Department of Nonlinear Modeling, Institute of Computer Science, Pod Vodárenskou věží 2, Praha 8, 182 07, Czech Republic
4 E. Hála Laboratory of Thermodynamics, Institute of Chemical Process Fundamentals of the AS CR, v. v. i., Rozvojová 135, Praha 6, 165 02, Czech Republic
This paper compares two approaches to modeling (smoothing) aerosol particle size distribution (particle counts for specified diameter intervals): i) the semiparametric approach based on a maximum likelihood fitting of lognormal (LN) mixtures at each time separately, followed by smoothing parameter tracks, ii) the nonparametric approach based on a kernel-like smoothing as an application of the gnostic theory of uncertain data. The specific advantages and disadvantages of both the semiparametric and nonparametric approaches are discussed and illustrated using real data containing a day-long time series of size spectra measurements.
Particle size distribution; Lognormal mixture; Semiparametric modeling; Nonparametric modeling; Gnostic theory of uncertain data.