Point source influence on observed extreme pollution levels in a monitoring network

This paper presents a strategy to quantify the influence major point sources in a region have on extreme pollution values observed at each of the monitors in the network. We focus on the number of hours in a day the levels at a monitor exceed a specified health threshold. The number of daily exceedances are modeled using observation-driven negative binomial time series regression models, allowing for a zero-inflation component to characterize the probability of no exceedances in a particular day. The spatial nature of the problem is addressed through the use of a Gaussian plume model for atmospheric dispersion computed at locations of known emissions, creating covariates that impact exceedances. In order to isolate the influence of emitters at individual monitors, we fit separate regression models to the series of counts from each monitor. We apply a final model clustering step to group monitor series that exhibit similar behavior with respect to mean, variability, and common contributors to support policy decision making. The methodology is applied to eight benzene pollution series measured at air quality monitors around the Houston ship channel, a major industrial port.