The impact of optical measurement techniques on measured aerosol particle size distributions

Ambient aerosol particle size distributions measured by the Ultra-High Sensitivity Aerosol Spectrometer (UHSAS) at various sites around the world exhibit modes at optical diameters near 600 nm and 850 nm. These modes are not present in concurrent measurements with the Grimm 11-D Optical Particle Counter (OPC). Here, we argue that these modes result from the optical measurement technique itself, and we explain why they appear in measurements by the UHSAS but not in those by the Grimm OPC. We construct computer models of the UHSAS and Grimm (“digital UHSAS” and “digital Grimm”) and use these to investigate the size distribution that would result from measurements of artificial aerosol particle size distributions that do not contain modes. The appearance of modes for the structureless incoming size distributions sampled by the digital UHSAS is explained by the nonlinear behavior of partial scattering cross sections of uniform spherical particles as a function of their diameter. The absence of modes in the digital Grimm is explained by the coarser size resolution of that instrument. Detailed analysis of the relationship between optical and geometric diameters for uniform spherical particles reveals two important results. First, these diameters generally have different numerical values for the same particle, and second, the relationship is nonlinear; thus, widths of size bins in terms of optical diameter differ from those in geometric diameter. These results explain the modes observed in the ambient size distributions and highlight concerns with attempts to create a merged size distribution by combining measurements from different instruments. Size distributions of aerosol particles describe how many particles occur in different size ranges. Measurements of aerosol particle size distributions from some instruments at a variety of locations around the world have modes, or size ranges where number concentrations are greater than those in nearby ranges. These modes are not present in concurrent measurements with another similar instrument. We examine this seemingly anomalous behavior by creating simple computer models of how the instruments behave and applying these models to simple artificial aerosol particle size distributions. The presence of modes is explained by the physical basis used by the instruments to determine particle size, which is the amount of light scattered by a particle. This amount is a nonlinear function of the particle diameter because of the wave properties of light. This nonlinearity propagates through to the size distribution, where it creates ranges that contain more particles, resulting in modes. The reason that these modes do not occur in the other instrument is explained by the difference in resolution; the size ranges into which the particles are sorted by that instrument are so wide that the modes occur within a given size range and thus cannot be resolved.