| Abstract: | A nuclear microprobe with high spatial resolution and high analyti- cal sensitivity was applied to analyze atmospheric aerosol at five monitoring sites in Shanghai city. Meantime, a new pattern recognition technique, which used the micro PIXE spectrum of a single aerosol particle as its fingerprint, was developed to identify the origin of the particle. The results showed that the major contributors to the at- mosphere pollution were soil dust (31.6%), building dust (30.8%), and the next were vehicle exhaust (13.7%), metallurgic industry excrements (5.6%), oil combustion (5%) and coal combustion (2.3%). Besides these, about 10% of the particles could not be identified. Based on the cluster analysis of these particles, they could be divided into eight groups. By inference, they might belong to some sub-pollution sources from soil dust, building dust and metallurgic industry excrements. Moreover, some new pollution sources from tyres and chemical plants were also revealed. |
