摘要：Highlights•Kaohsiung Refinery shutdown impacts pollutant concentrations (vary by pollutants).•It is better to estimate the formation of O3based on the PMF-MLR analysis.•The background sources significantly change O3concentrations.•Long-range transport had a significant impaction to O3.AbstractIt is always difficult to compare, let alone estimate, the difference of air pollutant concentrations before and after closure of a major source because the pollutants cannot be traced or predicted after entering the ambient. Indeed, we are not aware of any studies specifically related to the air pollutants impacted by a winding-down source. In this work, we applied nine years (2010–2018) online measurement of air pollutants (including PM10, PM2.5, NO2, SO2, O3and VOCs) to investigate (i) the temporal behavior of air pollutants before and after closure of an oil refinery park by using pair-wise statistics and correlations between wind speed and direction, and (ii) the source impacts on O3concentrations using PMF coupled with multiple linear regression (MLR) analysis (PMF-MLR). Example applications are presented at two monitoring sites (A and B) close to the Kaohsiung Oil Refinery (KOR), located in the southern industrial city of Taiwan. The results show that the KOR shutdown changed air pollutant concentrations to a certain extent in these study areas. We also conclude that, instead of using propylene-equivalent and ozone formation potential (OFP) concentrations, it is better to estimate the formation of O3based on PMF-MLR analysis as developed in this study. The PMF analysis has identified various VOCs sources at both sites including solvent usage, petrochemical industrial sources, industrial emissions, vehicle-related sources, vegetation emissions and aged air-masses. Also, the MLR model shows that both the background sources and petrochemical industrial sources may significantly change O3concentrations.
关键词：KeywordsenWinding-down source analysisUrban air qualityBivariate polar plotsSource apportionmentMutiple linear regression
