Traditional source apportionment analysis can be distinguished into two categories: 1. Pure mathematical iterative updating method aim for minimizing the difference between measured data matrix and reconstructed data matrix. 2. A geometric approach which identify sources through determining the edges and/or vertices of the measurements in a multidimensional eigenspace. The pure algorithm method has inherent shortcoming due to matrix rotation, while the geometric method suffers from large data uncertainty. A new approach is showing here firstly using the eigenspace projection to properly determine number of factors and to eliminate the effect of data rotation, then a mild matrix updating technique is applied to find the optimized solutions.

New idea in source apportionment analysis: A combination of iterative updating algorithm with eigenspace projection

244th National Meeting (2012)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.