A comparison between two receptor models to determine the source apportionment of atmospheric pollutants

Two frequently used factor analysis (FA) methods, Absolute Principal Components Scores (APCS) and Target Transformation Factor Analysis (TTFA). are compared by examining two cases with different complexity Oil sources and parameters. The effect of the noise contribution on the data is evaluated. By applying the methods to Simulated data matrices. it has been found that the APCS method characterizes the source inatrix better than TTFA method. APCS method also reconstructs the sample better than TTFA. In addition both methods are applied to real data collected during a monitoring campaign in Taranto city (South Italy). Copyright (c) 2006 John Wiley & Soils. Ltd.