Authors:J. Tsai, S. Owega, G. Evans, R. Jervis, M. Fila, P. Tan, and O. Malpica
Summertime urban PM2.5 was collected on cellulose filters in downtown Toronto, using a customized air sampler (635 l/min). Mass concentrations for up to 19 trace elements/ions were measured by ICP-AES, INAA and IC. Source apportionment was performed on these results including additional carbon and total mass concentrations using positive matrix factorization (PMF). PMF factors exhibited trends that indicated soil (18%), stationary (19%), secondary (48%), and vehicle (15%) sources. Potential source contribution function (PSCF) analysis identified probable sources of the stationary and secondary PM2.5 as originating from the south and southwest of Toronto.
The minimum degree ordering is one of the most widely used algorithms to preorder a symmetric sparse matrix prior to numerical factorization. There are number of variants which try to reduce the computational complexity of the original algorithm while maintaining a reasonable ordering quality. An in-house finite element solver is used to test several minimum degree algorithms to find the most suitable configuration for the use in the Finite Element Method. The results obtained and their assessments are presented along with the minimum degree ordering algorithms overview.
The paper discusses the usefulness of self modelling multivariate curve resolution (nonnegative matrix factorization) as a chemometric tool for the analysis of inhomogeneous spots captured by densitometer in multivariate way. The discussion is based on two examples: a spot of decomposed aspirin with comparative spot of pure salicylic acid, and spots of overlapped ciprofibrate and clofibric acid. It is concluded that this approach works well in the case of thin-layer chromatography (TLC) and the algorithm finds reliable spectral and concentration profiles, even with high overlap and spectral similarity. Nonlinearity does not affect this algorithm in visible manner. The computation can be performed with free software and can be recommended as good method to analyze inhomogeneity and to obtain additional proof of spot contamination with estimates of the spectral profiles.
Authors:Yong-Sam Chung, Sun-Ha Kim, Jong-Hwa Moon, Young-Jin Kim, Jong-Myoung Lim, and Jin-Hong Lee
For the identification of air pollution sources, about 500 airborne particulate matter (PM2.5and PM10) samples were collected by using a Gent air sampler and a polycarbonate filter in an urban region in the middle of Korea
from 2000 to 2003. The concentrations of 25 elements in the samples were measured by using instrumental neutron activation
analysis (INAA). Receptor modeling was performed on the air monitoring data by using the positive matrix factorization (PMF2)
method. According to this analysis, the existence of 6 to 10PMF factors, such as metal-alloy, oil combustion, diesel exhaust,
coal combustion, gasoline exhaust, incinerator, Cu-smelter, biomass burning, sea-salt, and soil dust were identified.
Authors:Cao Lei, S. Landsberger, S. Basunia, and Yu Tao
Airborne particulate matter (APM) was collected in coarse fraction and in PM2.5 during spring of 2002 in Beijing suburban sampling site by Gent SFU sampler. More attention has been paid to the special
“events” such as dust, storm and haze. Taking advantage of the combination of thermal or epithermal neutron irradiation with
Compton suppression spectrometer system, twenty elemental (Al, Si, Ca, K, Dy, Cu, I, In, Ba, W, Sn, Sb, As, Ti, Br, V, Mn,
Cl, Na, Zn) concentration were determined. Among them, several key trace elements that cannot be accomplished by the traditional
neutron activation analysis (NAA) were determined. The analysis of trace elemental concentration in PM2.5 shows that the anthropogenic elements such as As, In, Sn, Sb have different trends than crustal elements. The back-trajectories
of the high concentration anthropogenic pollution elements revealed their source region. Six potential sources were resolved
by positive matrix factorization (PMF), two area type and four source type, as soil, limestone quarry, crop burning and mixture
of residue motor and coal burning sampling sites. Taking into account of everyday air particle back trajectories, source compositions
together with source regions were also identified.
Authors:B. Khakimov, G. Gürdeniz, and S.B. Engelsen
Acar, E., Gürdeniz, G., Rasmussen, M.A., Rago, D., Dragsted, L.O. & Bro, R. (2012): Coupled matrixfactorization with sparse factors to identify potential biomarkers in metabolomics. Proceedings of the 2012 IEEE International Conference on Data