A Descriptive Analysis of Temporal Patterns of Air Pollution in Atlanta, GA and an Assessment of Measurement Error in Air Pollution Monitoring Networks in Atlanta, GA
Wade, Katherine Signs
MetadataShow full item record
This research is intended to serve as an in-depth analysis of air pollution patterns and monitoring networks in the Atlanta area. A ten year database of carbon monoxide (CO), sulfur dioxide (SO2), nitrogen oxides (NOx), ozone (O3), and particulate matter (PM2.5 and PM10) measurements at 17 monitoring stations across the Atlanta area was developed for use in this research. Temporal profiles of air pollutants are analyzed and described. Several factors are identified that impact these profiles, including changes in emissions, meteorology, and photochemistry. Most sites exhibited decreasing annual average concentrations during the study period, with the exception of O3 and NOx, both of which initially increased and then decreased. CO, NOx, and SO2 all have the lowest concentrations in the summer months, while O3 and PM2.5 are highest in the summer months. CO, NOx, and SO2 are also slightly lower on the weekends. CO and NOx have peak daily concentrations at rush hour, while O3 and SO2 peak in the afternoon hours. Instrument error was evaluated through audit and calibration data and collocated data. Collocated data is assumed to be a more accurate representation of instrument error; the percent error calculated using collocated data is much higher than that calculated using audit data. Percent errors were similar for all pollutants using audit and calibration data (2-4%) and were similar for all concentration ranges. Percent errors using collocated data were several times larger. Semivariogram plots are developed to quantify spatial variation of air pollutants. These plots can be interpreted to give the fraction of temporal variation in a pollutant that is actually due to spatial variation. As expected, primary pollutants have higher spatial variation than secondary pollutants. Population weighted averages of the semivariogram function are developed to give a level of uncertainty for a pollutant across the study area. Pollution rose plots are developed to qualitatively examine local sources that are impacting the monitoring sites used in this research. Point sources are easily identified in SO2 plots, as are mobile sources in CO and NOx plots. Pollution roses are also corrected for time of day and season to eliminate false sources.