Authors: Karl F Chastko*, University of Toronto Mississuaga, Matthew D Adams, University of Toronto Mississuaga
Topics: Spatial Analysis & Modeling, Geography and Urban Health, Temporal GIS
Keywords: GIS, Air Pollution, Land-Use, Error-Analysis, Modeling
Session Type: Paper
Start / End Time: 10:00 AM / 11:40 AM
Room: Grand Ballroom B, Astor, 2nd Floor
Presentation File: No File Uploaded
Air pollution represents a significant health risk due to an array of respiratory complications and chronic diseases, which can result from prolonged exposure to air borne pollutants. Exposure models for air pollution are typically created by interpolating long-term averages obtained from stationary and mobile monitoring devices in an area of interest. When models incorporate mobile air pollution data there is often significant error associated with the data due to the spatially and temporally discontinuous nature of the data. Researchers attempt to minimize the error by applying temporal adjustment factors; however, minimal evidence supports the effects on the estimate uncertainty of these adjustment factors. Temporal adjustment factors from derived from literature, as well as custom adjustment factors, were applied to a diverse set of data to investigate the robustness and validity of these adjustments to predict PM2.5 concentrations under varied conditions. Simulated mobile air monitoring collection campaigns were generated by sampling from the original datasets. This allowed for accurate estimation of model errors and also allowed for easy identification of spatial trends in the data. The results of this study showed that attributes of the data such as; the data distribution, land-use characteristics of the sample area and sampling period, all influenced the performance of a particular adjustment factor. This information was then used to develop a frame work for selecting an appreciate adjustment factor for a mobile air pollution modeling.