Clearer Skies Over China: Reconciling Air Quality, Climate, and Economic Goals

7. Atmospheric Modeling of Pollutant Concentrations

7 Atmospheric Modeling of Pollutant Concentrations Yuxuan Wang 7.1 Introduction Pollutant concentrations near the surface of the earth are of central concern in protecting public health, agricultural productivity, and ecosystems. The mapping of surface concentrations from estimates of emissions like those developed in chapter 6 is seldom straightforward, however, as the atmosphere is a highly complex system in terms of both its physical transport mechanisms (or dynamics) and its chemistry. Chemical species emitted at or near the surface will mix into the surrounding air and will be carried away by winds and other forms of atmospheric transport such as convection. Species emitted in their final pollutant form—that is, those transported but not chemically transformed—are referred to as primary pollutants. Other emitted species undergo chemical transformations in the air, leading to the formation of secondary pollutants such as ozone (O3, a gas) and sulfate, nitrate, and other aerosols (a scientific term for solid particles or liquid droplets suspended in gas). Numerical models that incorporate our current understanding of the physical and chemical processes controlling atmospheric composition are designed to predict the relationships between emissions and the resulting pollution distributions in the atmosphere. In this chapter we discuss the results of a state-of-the-art numerical model of atmospheric dynamics and chemistry quantifying the effects on air quality over China of the policy scenarios described in previous chapters. The atmospheric modeling approach adopted here is a major advance over the simplified one used in the preceding book of the current research program, Clearing the Air: The Health and Economic Damages of Air Pollution in China (Ho and Nielsen 2007). In section 7.2, we first provide a brief overview of numerical model types. Some general but fundamental differences between the modeling approaches adopted in 264 Chapter 7 this chapter and in Clearing the Air will be discussed. A deeper description of the model employed in this study is then presented in section 7.3. Section 7.4 presents the model results, focusing on the effects of the different policy scenarios on air quality. A summary and discussion are given in section 7.5. 7.2 Overview of Numerical Atmospheric Models Four general types of processes determine the variation in space and time of the concentrations of chemicals in the atmosphere: emissions, transport, chemistry, and deposition. The numerical models that predict the relationship from emissions to pollution concentrations need to simulate all these processes. As inputs, they require emission data (including location and amount) and meteorological information (such as wind, pressure, solar radiation, and rainfall). Model outputs are distributions of pollutant concentrations in the atmosphere. The spatial and temporal resolutions of the outputs are determined by the resolutions of the input data and by computational constraints. In this section, we provide a general description intended for nontechnical readers of the two types of models used most commonly in air quality studies: atmospheric dispersion models and Eulerian gridded models. 7.2.1 Atmospheric Dispersion Models Dispersion models are designed to simulate the transport and evolution of a single pollutant plume emitted from a point source or other sources that can be approximated as a point source. A typical example of a point source is the smokestack of an industrial plant. Given atmospheric wind and stability, the dispersion model calculates the rate at which the plume is transported downwind of the source, considering also the diffusion of the plume that results from atmospheric turbulence. As the mass of the pollutant plume is conserved in the atmosphere, the diffusion process that enlarges the volume of the plume results in decreased concentrations with time in three dimensions of the atmosphere, that is, in downwind, crosswind, and vertical directions. Dispersion models use mathematical algorithms that simplify dispersion and dilution phenomena to predict the evolution of the plume in the atmosphere. Some advanced dispersion models may include pollutant deposition at the surface by assuming a certain loss frequency in response to terrain topography and surface roughness. Simple chemical processes may also be included during the dispersion process through a characteristic chemical decay time scale that is a function of temperature, pressure, humidity, and other atmospheric conditions. Atmospheric Modeling of Pollutant Concentrations 265 Dispersion models are used mainly to simulate the evolution of a point-based plume. Although some dispersion models can have multiple plumes (“puffs”) from a source or multiple source locations, the dispersion processes of these plumes/ sources are independently calculated, and...