Dispersion Modeling of Inert Particulate Matter in the El Paso, TX- Cd. Juarez, MX Region

The El Paso, TX-Cd. Juarez, MX region is subject to the emission of inert particulate matter (PM) into the atmosphere, from a variety of sources. The impact of these emissions has been studied extensively in for regulatory compliance in the area of health effects, air quality and visibility. Little work has been done to study the fate and transport of the particulate matter within the region. The Environmental Physics Group at The University of Texas at El Paso has recently applied the SARMAP Air Quality Model (SAQM) to model the dispersion of inert particulate matter in the region. The meteorological data for the SAQM was created with the Penn State/NCAR meteorological modeling system, version 5 (MM5). The SAQM was used to simulate three common occurrences for large particulate emission and concentration. The first was times of heavy traffic volume at the international bridges which cause large numbers of cars to sit, with engines running, for extended periods of time. The second was moderate to high wind events that cause large amounts of coarse particulate matter to become entrained in the atmosphere and transported into and around the region. The third is a temperature inversion which traps the particulate matter at the surface during morning rush hour. The initial conditions for particulate matter, for the two cases involving mobile emissions, were derived from the 1999 version 3 national emissions inventory (NEI) mobile, on-road data from the EPA. Output from the MM5 was used to as the meteorological driver for the SAQM. The MM5 was initialized with data from the NCAR reanalysis project. Meteorological data collected in the region bye the Texas Commission on Environmental Quality (TCEQ) and the EPA was used for Four Dimensional Data Assimilation. The MM5 was nudged with gridded, surface and observational data. Statistical analysis was done on the MM5 for the variables, wind speed, wind direction, temperature and mixing ratio. The statistics performed included RMSE, RMSEs, RMSEu and index of agreement. MM5 output with low RMSE and high index of agreement was used to drive the SAQM. The MM5 grid domains were 39x39 at 36km, 47x47 at 12km, 55x55 at 4 km and 40x40 at 1.3km. The SAQM was applied on to the 1.3km domain. For the case of emission at the international bridges, the bridges' latitude and longitude were translated to grid cell locations. The NEI data derived for those locations were set as emission rates for those cells. The SAQM was run for a 24hr period starting at twelve pm local time with the emissions ending after morning rush hour. The same conditions were done for the inversion time period with the addition of emissions for major roadways and arterial feeders. No data is available for concentrations of entrained particulate matter during wind events. Thus, the entrainment episodes were simulated with varying initial concentrations along the boundary of the domain. The emission rates were varied for each simulation to give both a very intense episode, and a moderate episode lasting for 12 hrs with the SAQM simulation ending after 24 hrs. Analysis for all the simulations was done to show the spatial and temporal evolution of the PM. Temporal comparisons were done between EPA PM2.5 to show identify similarities in the evolution of the SAQM with observation.