GEOSTATISTICAL ANALYSIS AND VISUALIZATION OF HOURLY OZONE DATA

The recent recognition of the detrimental impacts of near-surface ozone concentrations has led to the establishment of regionwide monitoring networks. The size of the collected data sets, as well as the presence of pseudorandom patterns in measured ozone concentrations, has motivated the use of advanced techniques for data analysis and presentation. In this work, a series of geostatistical and visualization procedures are applied to the analysis of hourly ozone measurements collected from 29 stations in the southeastern United States. The exploratory data analysis and variography clearly confirm the diurnal pattern of ozone fluctuations, and suggest that during daytime hours, local factors are dominant over the ozone production process. In contrast; during evening hours, regional factors and small-scale uneven variations become more dominant in the ozone depletion process. Using the geostatistical technique of cross-validation, results are generated in order to evaluate the rural classification of investigated stations. The findings of these studies indicate that: (1) the non-rural status of near-urban or presumably downwind stations is an hour-specific condition; (2) prevailing wind directions on each day can significantly alter the rural status of presumably upwind sites; and (3) local topography of a site can be an important factor in influencing its rural status during different hours of the day. Geostatistical techniques are shown to be not only useful for optimal mapping and network design, but also can be coupled with visualization tools to generate animated pseudocolored sequential maps. These maps are an efficient means for analysing spatial and temporal patterns of ozone concentrations.