An examination of the algorithm for estimating light extinction from T IMPROVE particle speciation data

Prenni, A., J.L. Hand, W.C. Malm, S. Copeland, G. Luo, F. Yu, N. Taylor, L.M. Russell, and B.A. Schichtel (2019), An examination of the algorithm for estimating light extinction from T IMPROVE particle speciation data, Atmos. Environ., 214, 116880, doi:10.1016/j.atmosenv.2019.116880.
Abstract

The Interagency Monitoring of Protected Visual Environments (IMPROVE) network is the basis for monitoring visibility in Class I Areas throughout the United States. Monitoring is conducted by collecting PM2.5 and PM10 samples every third day at over 150 remote and rural sites nationwide, with PM2.5 samples being analyzed for chemical composition. Light extinction is reconstructed using an algorithm that relates speciated mass to light extinction based on aerosol mass and size (second IMPROVE equation). In addition, the National Park Service directly measures light scattering at a subset of IMPROVE sites using Optec NGN-2 integrating nephelometers. The optical measurements serve as a check of the reconstructed scattering, in that measured scattering from the nephelometers and reconstructed scattering from the second IMPROVE equation should be equivalent. During its development, the second IMPROVE equation was shown to accurately estimate light scattering for a broad range of aerosol compositions and loadings for samples collected through 2003. Here reconstructed scattering is assessed from the second IMPROVE equation by comparing measured light scattering from nephelometers to reconstructed light scattering at 11 collocated sites over a 16 year period (2001–2016). The comparisons suggest that the relationship between measured and reconstructed light scattering has changed over time and that in recent years the second IMPROVE equation has underestimated light scattering at many sites. This shift toward poorer agreement corresponds to periods with relatively large decreases in sulfate and organic mass concentrations. These decreases lead to biases in reconstructed scattering calculated with the second IMPROVE equation, due to the assumed relationship between mass concentration and size distribution. This relationship, referred to as the split component algorithm, appears to be flawed as currently implemented. A potential approach is explored that scales the split component algorithm each year and at each site, based on measured mass concentrations at the site. The proposed approach appears to reduce the biases in the second IMPROVE equation.

PDF of Publication
Download from publisher's website
Research Program
Atmospheric Composition Modeling and Analysis Program (ACMAP)