A PRACTICAL DEMONSTRATION OF THE ABSOLUTE PCAS BIAS

This paper gives a practical demonstration that the receptor models ba sed on the absolute PCA techniques, to apportion atmospheric contamina nts to sources, might bias results. The demonstration is done with two simple examples. Results show in both cases that the mean of zeta(j) (APCS(jk)) is in fact closer to the absolute PC0* than the source dis tribution (S-j). These three terms being not, as previously stated, al ways equal. In this circumstance, elements concentrated in the sources are not accurately reproduced.