PLSR and ANN estimation models for PM10-bound heavy metals in Dunkerque (Northern France)

Abstract

The aim of this work is to develop statistical estimation models of some EU regulated heavy metal levels (Pb, Ni) and some non-regulated heavy metal levels (Mn, V and Cr) in the ambient air of the city of Dunkerque (Northern France) so that they might be used for air quality assessment as an alternative to experimental measurements, since these levels are relatively low compared to the EU limit/target values and other air quality guidelines. Three different approaches were considered: Partial Least Squares Regression (PLSR), Artificial Neural Networks (ANN) and Principal Component Analysis (PCA) coupled with ANN. External validation results evidence that PLSR and ANN-based statistical models for regulated metals and for Mn and V provide adequate mean values estimations while fulfill the EU uncertainty requirements.