Seasonality patterns and distinctive signature of latitude and population on ozone concentrations in Southern Ontario, Canada

Ozone is a serious health concern and the only major pollutant in Ontario that is increasing, despite major efforts to control atmospheric pollutants in North America. Ambient ozone levels are affected by local atmospheric conditions (temperature, humidity, radiation, precipitation) and the concentration of its main precursors, nitrogen oxides (NOx) and volatile organic compounds (VOCs). In the present study, our aim is to elucidate the day-to-day and seasonal variability of the influence of synoptic atmospheric descriptors on daily average ozone concentrations, as well as to establish a predictive framework of the likelihood of exceedance of harmful levels in Southern Ontario. Using 31 stations in Southern Ontario with daily ozone concentration data from 2007 to 2015, we developed parsimonious models using as predictors a suite of atmospheric variables, available from the National Centers for Environmental Prediction and the National Center for Atmospheric Research (NCEP/NCAR) reanalysis dataset, and terms that recreate the seasonality and weekend/weekday effects of the previous day ozone concentrations. In all sites, seasonal (among-month) variability was distinctly lower than within-month variability. Our analysis showed that the population size reduced the average ambient ozone concentrations, increased their variability, and decreased the predictive power of our models. Ambient ozone levels in higher latitude sites displayed stronger dependence on the antecedent conditions, whereas the human activities weakened the serial correlation in the ozone time series in urban areas. Seasonal changes in the magnitude and sign of the slopes of the atmospheric predictors were consistent within the spatial domain of our study: shortwave radiation consistently increased ambient ozone levels, temperature displayed a positive relationship during the summer but a negative one during the winter; humidity was characterized by a negative relationship with ozone; precipitation increased ozone during late summer and winter, with no consistent effect during the rest of the year; wind speed and direction also showed contrasting effects between summer and winter months. Seasonality strongly affected the probability of exceedance in low population sites but had a minor role in more densely populated areas, while the difference in ozone concentrations between weekend and weekdays increased in urban sites. Overall, our results suggest that population and seasonality of the preva