Spatiotemporal variations and characterization of the chronic cancer risk associated with benzene exposure

Abstract

A spatiotemporal analysis of benzene was performed in east of the USA and in a representative station in Baltimore County, in order to assess its trend over a 25-year time span between 1993 and 2018. A novel time series analysis technique known as TBATS (an ensemble of Trigonometric seasonal models, Box-Cox transformation, ARMA error plus Trend and Seasonal components) was applied for the first time on an air contaminant. The results demonstrated an annual seasonality and a continuously declining trend in this respect. The success of Reformulated Gasoline Program (RFG), initiated in 1995, was obviously detected in time series data since the daily benzene concentrations reduced to one-sixth of its original level in 1995. In this regard, the respective values of mean absolute scaled error (MASE) were 0.35 and 0.45 for training and test series. Given the observed concentrations of benzene, the hot spot areas in east of the US were identified by spatial analysis, as well. A chronic cancer risk was followed along the study area, by both a deterministic and probabilistic risk assessment (PRA) techniques. It was indicated that children are at higher risk than that of adults. The range of estimated risk values for PRA was higher and varied between 6.45 × 10−6 and 1.68 × 10−4 for adults and between 8.13 × 10−6 and 8.29 × 10−4 for children. According to the findings of PRA, and referring to the threshold level of 1 × 10−4, only 1.2% of the adults and 28.77% of the children were categorized in an immediate risk group.