ABSTRACT: Air pollution is a major environmental concern in the El Paso-Juárez region. According to the Ministry of Environment and Natural Resources (Secretaría de Medio Ambiente y Recursos Naturales, SEMARNAT) the city of Juárez is one of the city's in México with the highest atmospheric levels of pollution because of its accelerated and unplanned urban growth. One air pollutant of concern is nitrogen dioxide (NO2) due to its detrimental health effects that have been associated with airway inflammation in healthy people and increased respiratory symptoms in people with asthma. Land use regression modeling is a GIS based approach that seeks to predict pollution concentrations at a given site based on surrounding land use, traffic characteristics, and other geographic variables in a multivariate regression model. This type of model has been a practical and effective method to predict intraurban variation in nitrogen dioxide in several places in North America. It will be useful to create a similar model at the El Paso-Juárez borderland region to assess nitrogen dioxide exposures. This research evaluates the strength and association of different land regression variables into predicting NO2 concentrations. Monitoring for NO 2 levels was conducted at 27 locations in the city of Juárez that included 22 schools and 5 homes in a prior study from December 2002-September 2003. Main point sources for NO2 were identified in the El Paso-Juárez region and include: international ports of entry, cement plants, electric engine factories, and petroleum refineries. Distance to main point sources as well as traffic volume and traffic density on major streets near the monitoring locations were calculated utilizing ArcGIS 9.3.1 and used as predictor variables. Significant pearson correlations with NO2 concentrations were found with the following predictive variables: distance to a cement plant (DIST_CP), traffic density within the 1000 meter buffer zone (TD_1000), distance to an oil refinery (DIST_OR), distance to electric engine factories (DIST_EE), and distance to a major street with the second highest traffic volume (DIST_2 nd ). Most of the significant correlations found were consistent with the findings in previous studies conducted in the El Paso-Ciudad Juárez area by Gonzales, et al. (2005) and Smith, et al. (2006). A model built through a stepwise multivariate regression analysis revealed that the three main variables for NO2 variations include distance to a cement plant (DIST_CP), distance to a major street with the second highest traffic volume (DIST_2 nd ), and distance to an oil refinery (DIST_OR), predicting 59 percent of the variation in NO2 concentration. A bootstrapping analysis of 1,000 iterations evaluated and verified the robustness of the model. Recommendations for future analysis on this pollutant include a closer look into the effect of distance to a cement plant in the variation of NO2 concentrations, and perhaps the inclusion of other variables that could increase the predictability of the model, such as elevation above sea level.
ETD Collection for University of Texas, El Paso.