In this study, an approach has been developed for differentiating between local and remote pollution over Taiwan, based on homogeneity perspective (variations of the standard deviation) of both AERONET measurements and NASA MERRA aerosol reanalysis (version 2, MERRA-2) over a 15-year period (2002 - 2017). The analysis of seasonal variations of the standard deviation of aerosol optical depth (AOD) measurements at six AERONET sites and MERRA AOD data in Taiwan showed that, in spring when remote aerosols dominate, the standard deviation is almost three times lower than that in autumn, when aerosols from local sources dominate. This finding was supported by MERRA AOD over the open ocean area: total AOD data were used to differentiate between local and remote pollution over both Taiwan and the open ocean area in the vicinity of Taiwan. Over Taiwan, MERRA total AOD showed a primary maximum in spring and a secondary one in autumn. Over the open ocean area, where there are no local sources of anthropogenic aerosols, MERRA total AOD showed only one maximum in spring and no maximum in autumn. This suggests that, in Taiwan, the maximum in autumn is attributed to local air pollution, while the pronounced maximum in spring is mainly caused by air pollution from continental Asia. The analyses of spatial distribution of 15-year monthly mean MERRA winds confirmed the above-mentioned results. Furthermore, similar to total AOD, MERRA sulfate AOD peaked in autumn over Taiwan, but not over the oceanic area: this indicates the contribution of local emissions of anthropogenic aerosols from the industrial sector. The standard deviation of MERRA sulfate AOD in spring is two-three times lower than the standard deviation in autumn: this is additional evidence that, in spring, sulfate aerosols from remote sources are predominant; while in autumn sulfate aerosols from local sources dominate.
Keywords: AERONET; Local pollution; MERRA aerosol reanalysis; MERRA-2; Remote pollution; Taiwan.