Modeling Studies of the Transport and Transformation of Pollutants in the Lower Troposphere

Open Access
- Author:
- Stein, Ariel Fabian
- Graduate Program:
- Meteorology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- October 12, 2001
- Committee Members:
- Brian Dempsey, Committee Member
Gregory S Jenkins, Committee Member
Dennis Lamb, Committee Chair/Co-Chair
William Henry Brune, Committee Member - Keywords:
- sulfate
ozone
aerosols
photochemical modeling - Abstract:
- The utilization of mathematical models to represent the fates and transformations of atmospheric pollutants constitutes a fundamental practice that has contributed to the conceptual understanding of a variety of phenomena. These models provide the necesary framework for incorporating diverse atmospheric processes into a coherent system for studying their interactions. Consequently, we use several increasingly complex simulation tools to comprehend different aspects of pollution phenomena, such as the dispersion of non-reacting pollutants emitted from a smokestack, the forecast of tropospheric ozone formation over a regional scale, and the sensitivity of sulfate aerosol to changes in nitrogen oxides and hydrocarbon source strengths. Starting with the simplest approach for simulating the dispersion of a chemically non-reactive contaminant, we explore the comparability of Gaussian-type model predictions to atmospheric measurements. This work investigates the differences between time-averaged observations and ensemble mean concentrations as predicted by Gaussian models in laboratory and atmospheric boundary layer (ABL) flows. For a given averaging time it is shown that this difference is smaller in laboratory flows than in the ABL under the same stability and statistical conditions. Furthermore, with data from the Willis-Deardorff convection tank experiments, it is shown that the values of the normalized differences between observations and model-predicted concentrations in the ABL exceed 50 % for an averaging time of the order of 1 hour. These findings give a clear indication of the need for development of more accurate and sophisticated modeling tools to depict the atmospheric dispersion of non-reactive pollutants. In view of the fact that Gaussian models show high discrepancies between modeled and observed concentrations and that this approach is unable to simulate the formation of secondary pollutants, it is necessary to increase the level of complexity to model phenomena such as photochemical smog. Thus, a three-dimensional model, the Hybrid Single-Particle Lagrangian Integrated Trajectories model with a generalized non-linear Chemistry Module (HY-SPLIT CheM), has been utilized to forecast summertime ozone concentrations over the northeastern United States. The ability of HY-SPLIT CheM to simulate ozone mixing ratios has been evaluated by comparing calculated summertime ozone mixing ratios against measured values for the month of July, 1999. Generally, a fair agreement is observed at most stations, especially taking into account the large number of assumptions used to construct the model. HY-SPLIT CheM is the first operational implementation of the particle-in-grid approach applied to air quality modeling. This model constitutes a feasible tool due to its simplicity and low cost of implementation. The study of the formation of aerosol particles, specifically sulfate (SO42-) aerosols, involves a complicated coupling among gas-phase chemical reactions (as in the formation of ozone), aqueous-phase photochemical, and meteorological processes within the simulation framework. The formation of SO42- is chemically linked to primary emissions of sulfur dioxide (SO2) via atmospheric oxidants and therefore also to the emissions of nitrogen oxides (NOx) and volatile organic compounds (VOC). These compounds represent the chemical precursors of ozone, which in turn constitutes the main source of SO2 oxidants. The response of SO42- production to controls in NOx and VOC emissions depends in part on the resulting changes in oxidant levels and the competition that naturally exists between the gas- and aqueous-phase pathways for SO2 oxidation. We therefore propose the use of a combination of concentrations of nitric acid, particulate nitrate, hydrogen peroxide, and sulfate as a non-dimensional indicator of the effectiveness of VOC or NOx controls in decreasing SO42- abundance. The concentrations of these indicator species were calculated from a series of photochemical model simulations with varying rates of NOx and VOC emissions using a state-of-the-art three-dimensional Eulerian model (MODELS-3) that covers the northeastern United States. It is shown that sulfate concentrations are likely to decrease as VOC emissions are reduced when the non-dimensional indicator is less than about 1.5. Under these same conditions a reduction in NOx emissions would, however tend to increase the SO42- levels. On the other hand, a higher value of the indicator identifies a regime in which reductions in NOx emissions are more effective for reducing sulfate than VOC emissions are. In addition, a description of the sulfate-formation pathways, along with a theoretical analysis of the transition between NOx- and VOC-sensitive regimes, provides a strong rationale for the use of the sulfate-sensitivity indicator.