AIR POLLUTION DYNAMICS AND MODELING Lixin FU and Yang CHEN Department of Environmental Sciences and Engineering, Tsinghua University, Beijing, P. R. China
Keywords: Air pollution, meteorology, atmospheric diffusion, dispersion, removal, deposition, air quality, dispersion modeling, turbulence, mixing layer, convection, transport, plume, models, Eulerian, Lagrangian, Gaussian, emission, buoyancy, weather, stability, stack effluents, tracer, transformation, roughness, sink, deposition, radiation, removal, inversion, fog, photochemical smog, smoke, land-use, stagnant, topography, dilution, abatement Contents 1. Introduction 2. Source characteristics 3. Air Pollution Meteorology 4. Atmospheric Removal Processes 5. Atmospheric Diffusion 6. Air Pollution Modeling 7. Pollution Accidents and Meteorological Control Related Chapters Glossary Bibliography Biographical Sketches
Summary Air pollution dynamics has a nature of complexity. Firstly, it is influenced by the characteristics of the emission source, especially the elevation where a pollutant is discharged. Then the meteorological conditions are discussed since they are determinant for the dispersal of the air pollutant. Some important meteorological factors are described in detail in this chapter. The air pollutant can be removed by dry and wet deposition processes in the atmosphere, and this has an important impact on the fate of the pollutant. Although a dynamic process can be described using a physical theory, it is not always possible to solve such mathematical equations. A general Eulerian and Lagrangian approach is introduced in this regard and the Gaussian plume equation is extensively illustrated, as it is the most commonly used one. A simple description is then given for constructing the necessary components for a comprehensive air quality model, and a brief concept of an air pollution episode complemented with meteorological control is also presented. Finally, the authors give their comments on future developments in air pollution dynamics and modeling.
1. Introduction Personal experience tells humans that air pollution constantly changes with meteorological conditions. Air laden with visible pollution can be transformed to clear, blue skies within hours by a sudden change in the weather, such as the passage of a weather front. It is obvious, therefore, that air pollution is a dynamic problem.
In practice, how air pollutants are transported to a specific location is always important. Most industrial effluents are discharged vertically into the air through a stack or duct. After leaving the discharge point, the contaminated gas stream (the plume) expands and gradually mixes with the ambient air. Horizontal air flow tends to bend the discharge plume downwind. At some point, the effluent plume levels off. While the effluent plume is rising, bending, and moving horizontally, the gaseous effluents are being diluted by the surrounding ambient air. As the contaminated gases are diluted by ever larger volumes of ambient air, they eventually reach the ground through dispersion. The mere presence of sources of emissions does not necessarily constitute air pollution. Every air pollution problem has three requisites: 1. There must be an emission of the pollutant or its precursor into the free atmosphere. 2. The emitted pollutant must be confined to a restricted volume of air. 3. The polluted air must interfere with the physical, mental, or social well being of people. Very often urban air pollution problems are aggravated by meteorological and topographical factors that concentrate pollutants in the city and inhibit quick dispersion and dilution processes. As complex as the phenomenon may be, it can be easily depicted by means of a simplified systems analysis diagram. Figure 1 represents such a system approach. Air pollution dynamics refers to the various processes operating during a pollutant’s lifetime in the atmosphere, from the emission source to a receptor.
Figure 1. Air pollution system 2. Source characteristics An important factor affecting ground level concentrations is the rise of the plume above the discharge point, and its subsequent transport. The higher the plume rises initially, the greater distance there is for diluting the contaminated gases as they expand and mix downward to the ground. The plume rise is determined by both the upward inertia of the discharge gas stream and by its buoyancy. The vertical inertia is related to the exit gas velocity and mass, whilst the plume’s buoyancy is related to the exit gas density relative to the surrounding air density. Increasing the exit velocity or the exit gas temperature will generally increase the plume rise, resulting in lower ground level concentrations. The physical stack height plus the plume rise, is called the effective stack height. When the pollutant plume rises significantly before leveling out, the calculation of ground level plume concentrations should use the effective stack height instead of physical stack height. The effective stack height can be estimated using a number of equations such as the Holland equation and the Davidson-Bryant equation. No allowance is made in these equations for conditions of atmospheric stability. Modifications are needed since the rise of the plume above the stack under unstable conditions is about 10% higher than calculated and under stable conditions about 10% lower. Also, the effect of water droplets cools the plume and causes it to lose buoyancy. Given a specific discharge height and a specific set of plume dilution conditions, the ground level concentration is proportional to the amount of contaminant materials discharged from the stack outlet for a specific period of time. Thus, when all other conditions are constant, an increase in the pollutant discharge rate will cause a proportional increase in the ground level
concentrations. This is the basic principle of reducing pollutant emissions in order to achieve air quality improvement.
3. Air Pollution Meteorology While pollutant stack emission parameters such as gas velocity, temperature and molecular weight are important, atmospheric dispersion of a pollutant is primarily dependent on meteorological conditions. Therefore, the air quality of a region is greatly influenced by the local meteorology. Weather parameters such as ambient temperature, wind speed, cloud cover, solar radiation, and inclement conditions (rain, snow, hail, etc.) can determine the atmospheric dynamics and therefore impact the severity of air pollution problems. Meteorology is the study of the various dynamic processes of the atmosphere. Meteorological scales of motion can be divided into three categories as follows: 1. Macroscale. Phenomena occurring on scales of thousands of kilometers, such as semipermanent high and low pressure areas that reside over the oceans and continent. (The term synoptic is commonly used to denote macroscale.) 2. Mesoscale. Atmospheric motions occurring on scales of hundreds of kilometers, such as land-sea breezes, mountain-valley winds, and migratory high and low pressure fronts. 3. Microscale. Phenomena occurring on scales of the order of 1 km, such as the meandering and dispersion of a chimney plume and the complicated flow regime in the wake of a large building. Meteorological phenomena are important for the study of air pollution. Each of these scales of motion plays a role in air pollution, although over different periods of time. For example, micrometeorological effects take place over scales of the order of minutes to hours, whereas mesoscale phenomena influence transport and dispersal of pollutants over hours to days. Finally, synoptic scales of motion have characteristic times of days to weeks. The term “longrange transport” commonly refers to transport on the synoptic scale, which is important for global and regional issues such as the greenhouse effect and acid rain. The importance of wind direction and speed on the dispersal of pollutants is always important, and the variation of these two parameters with time of day and season of the year is an even greater factor. This variation is relatively constant at a given location and is to be carefully considered when a plant is located and the degree of control necessary for emissions of air pollutants is determined. The pressure differences and the rotation of the earth combine to produce a localized wind rose. The wind rose of a region refers to its characteristic wind patterns with respect to wind speed and wind direction on an annual base. The wind speed also changes with height and it increases with increased elevation; this is called the wind shear. Wind speed is also a function of topography and urbanization. Mountains, hills, trees, buildings and other obstructions can divert wind patterns, increase atmospheric turbulence, influence general atmospheric stability, and, thereby, affect air pollution dispersion. Regarding urban air pollution, the region of the atmosphere governing transport and dispersion is the so-called planetary boundary layer, roughly the lowest 500m. The planetary boundary layer represents the extent of influence of the earth’s surface on the wind field in the atmosphere. Within the planetary boundary layer, winds are determined by the prevailing high-level air flows and the roughness of the surface.
3.1. Atmospheric Stability and Mixing Layer Within the lowest 500m, the atmospheric temperature profile (the variation of temperature with elevation) has an important effect on turbulence and pollutant dispersal. For a mixed air, the dry adiabatic (the condition of no heat exchanging with outside air) lapse rate is 5.4 per 1000 ft., that is, the air temperature decreases 5.4 for each 1000 feet increase in elevation. Conditions under which vertical displacements are not affected by buoyancy forces is called neutral stability. Generally the lapse rate of the atmosphere is greater or less than adiabatic. A delicate equilibrium like this is rarely found in the atmosphere. A negative lapse rate or inversion causes very stable air, it is unfavorable for the dispersion of air pollution. There are two general types of inversions: radiation and subsidence. The radiation inversion usually takes place at night, as the earth surface cools by radiation at a faster rate than the air, causing the air near the ground to be cooler and denser than the air above. Radiation inversions are very common, taking place during the night for more than two-thirds of the year in some locations. The fog layers in valleys observed during the morning is visible evidence of radiation inversions. Subsidence inversions are caused by the sinking motion of atmosphere in high pressure areas and they generally have decreasing humidity above the inversion base. They are more influenced by large scale meteorology. The height of subsidence inversions varies from the surface to about 5000 feet, depending on weather conditions. When an atmospheric temperature inversion is present, the atmosphere is considered to be stable and very little mixing or turbulence occurs as the denser air is near the ground. Therefore pollutants in the air do not get dispersed and diluted easily, causing severe air pollution problems if there is a mass of emissions. On the other hand, if the temperature decreases faster with height than the dry adiabatic lapse rate, air parcels at any height are unstable; that is, if they are displaced either upward or downward, they will continue their movement in the direction in which they were displaced. This is referred to as an unstable condition. In such an unstable air, there is substantial vertical mixing of the air and air pollutants are transported vertically and dispersed rapidly. During these conditions, the warm air near the ground (which may contain air pollutants) will rise and be replaced by cooler, clean air from the cold upper layers, creating good mixing and dilution. It is common that different types of lapse rates may exist at the same time in actual cases. Representatives of these are shown in Figure 2. Typical conditions contributing to intense air pollution episodes have been depicted in curves A and B. In both cases, vertical mixing of polluted air near the ground with the atmosphere above is hindered by the stable inversion layer. Only when the slope of the lapse rate is reversed will the air become turbulent and vertical mixing be resumed, as shown in curve C.
Figure 2. Typical atmospheric soundings The mixing depth is defined as the height to which the warm air rises and mixes with the cooler air until it meets its equal in temperature. It is the upper level for air pollution dispersion, and a function of the seasons, the ground temperature, sunshine and other meteorological factors. On a clear summer day with sunshine, the mixing depth may reach several thousand meters, whereas in winter with lower ground temperatures and less sunshine, it may be only one or two hundred meters. The mixing layer is a very important concept in the formulation and application of mathematical models of air pollution dispersion. Although the horizontal wind component is fairly well known in normal weather forecasting, the vertical component is not as well understood and often varies in direction and strength with elevation. Since vertical mixing is generally depicted by the term mixing depth as a function of atmospheric stability, great difficulties may be created when the air is stably stratified but nevertheless has a significant vertical wind component, as can occur, for example, at a warm front. The concept of the mixing layer is mostly useful when there is convective motion causing the lowest layers to be so sufficiently stirred that the gases are uniformly distributed within them, and there is a strong inversion at the top of these layers preventing any mixing with the upper layers. Convective mixing may be created by two mechanisms: buoyant convection due to the heating of air at the bottom, as when the ground is warmed by sunshine; or mechanical mixing, which is caused by the wind traversing rough ground with buildings, trees, and other obstacles. Urban protuberant profiles make many eddies within the boundary atmosphere, and
these tend to produce a more uniform distribution of whatever the air contains. Such uniformity may be quickly achieved if the mixing layer is shallow enough for the air parcels to traverse its depth in a few minutes, while there is a well-marked inversion at the top of the layer. The lapse rate of the air is dry adiabatic in such a well-mixed layer. 3.2 Large-scale Transport The role of long-range atmospheric transport in the transfer of man-made and natural substances around the globe has been more and more clearly revealed. The nuclear accident at Chernobyl highlighted the fact that one nation’s atmospheric emissions may profoundly affect its neighbors, especially for issues like acid deposition, greenhouse effect, and stratosphere ozone depletion. The movement of large masses of air driven by differences in temperature and pressure and the rotation of the earth is the kernel of macro-meteorological events. Basic movement of air is caused by the heating of air masses at the equator and cooling at the poles, with a resultant north-south circulation. The rotation of the earth from west to east tends to drag the atmosphere along with the surface so that the winds in the northern hemisphere tend to curve toward the right and those in the southern hemisphere toward the left. In order to describe and quantify the large-scale transport of materials through the atmosphere, better understanding of macro-meteorology is necessary, especially in cases of continent-to-continent and continentto-ocean transport. This can be achieved more rapidly by using a combination of techniques. Evidences show that all studies on this scale will benefit from the combination of field and modeling components that are coordinated and interactive. Reliance on a single tool is limited by the capability and uncertainty. The physical conditions of the atmosphere at any time are important in weather forecasting. These include measurements of atmospheric pressure; temperature; humidity; wind speed and direction; amount, type, and height of clouds; precipitation; visibility; and other special conditions in the atmosphere. They are usually shown on a weather map based on observations at many stations throughout a country, continent or the globe. After these data have been plotted, they are subject to analysis by drawing isobars, i.e., lines of constant pressure, normally at 3 milli-bar intervals. The boundary lines (fronts) between different air masses are drawn and the type of front indicated. Shaded areas reveal incidence and type of precipitation; arrows indicate direction of the wind. Isotherms are usually drawn for 0 oF and 32 oF as well. These maps are also useful tools for air pollution prediction, since even a cursory examination will reveal whether atmospheric conditions in a given area are likely to be favorable or unfavorable for dispersion of pollution. For example, when a cold front moves into a warmer area the air next to the ground will become heated and result in an unstable atmospheric condition (pollutants will disperse readily). But an air mass approaching a colder area will become cooled near the ground, and a stable atmosphere will persist in the surface layers. The wind velocities shown on weather map are those of the so-called gradient wind, which is caused by the rotation of the earth. The winds are generally along the isobars and at an altitude (2000 ft and above) high enough to be unaffected by surface friction. Flow is clockwise around high-pressure areas and counterclockwise around low-pressure areas. Near the ground, surface friction affects the speed and direction of the gradient wind so that winds generally blow between 20 and 30 degree across the isobars toward the low-pressure center and at half the speed of the gradient wind. In addition to daily weather maps, monthly maps of average weather conditions are also helpful particularly in determining areas and periods of stable high-pressure atmospheric conditions. High-pressure conditions (anticyclones) are
always responsible for extreme cases of local air pollution. In the central part of the highpressure area, winds are generally light and the atmosphere quite stable. When a high-pressure air becomes stagnant in one locality for several days, conditions for pollution build-up are extremely favorable. 3.3. Mesoscale Transport and Micrometeorology In addition to large-scale effects of air movement, those of a mesoscale nature must not be overlooked. For example, local circulation and temperature variations in valleys and on the slopes of hills and mountains are very important from an air pollution viewpoint. When the general circulation imposes moderate to strong winds, valleys that are oriented at an acute angle to the wind direction channel the wind. The valley effectively peels off part of the wind and forces it to follow the direction of the valley floor. This happens in some mountainous areas. Similarly, tall buildings (with resultant eddies) and other type of ground cover greatly affect airflow. In addition, the differences in heating and cooling rates between the valley floor and sides can bring about variation in the air density and pressure, influencing local airflow. In daytime, the heated walls will cause the valley air to be warmed. It will become more buoyant and flow up the valley. At night the cooling process will cause the wind to flow down the valley, and this local flow of cold air tends to suppress the upward motion of emitted pollutants. Another example of mesoscale airflow is the sea-land breeze. Under a stagnating anticyclone, a strong local circulation pattern may occur across the shoreline of large water bodies. The difference in the air temperature over land and water caused by differences in the heating and cooling, causes a pressure gradient and accompany in air flow. On clear days the land surface heats faster and rises to a higher temperature than the water surface. As a result of the higher pressure of the cold air over the water, the air flows from the water toward the land (a sea or lake breeze). At night the process is reversed. Radiational cooling of the land results in lower temperatures over the land surface than over the water, creating a flow of air from the land toward the water (a land breeze). The effect of the sea breeze on stability is to impose a surface-based inversion on the temperature profile. But this is a changing condition as the air moves from the water over the warm ground, it becomes heated from below. Thus, for stack plumes originating near the shoreline, the stable lapse rate causes a fanning plume close to the stack. As the air moves inland the lapse condition grows to the height of the stack. At some point inland, a fumigation (confined to the lower atmosphere) plume results. These sea-land breezes are generally light and are overshadowed when large scale air movements are obvious. In addition to air flows originated from mesoscale phenomena, the air motion in the lowest layers of the atmosphere is exceptionally important for local air pollution problems. Such air motion, taking place adjacent to a solid boundary of variable temperature and roughness, is virtually always turbulent. This atmospheric turbulence is responsible for the transport of heat, water vapor, and pollutants from the surface to the atmosphere as a whole. The objective of micrometeorology is to understand the basic phenomena that influence atmospheric turbulence. Great achievements have been made and several semi-empirical theories are established. 3.4. Vertical Transport During the process of pollutant transport in the atmosphere, mechanisms that vertically redistribute substances in the atmosphere, particularly between the planetary boundary layer and the free troposphere, have a significant influence on the efficiency of transport and
dispersion. Thus the elevation at which a pollutant is emitted into the atmosphere or is carried by vertical atmospheric motion may substantially influence its eventual range of transport. The change in air temperature with change in pressure, or with altitude, is an important factor in the vertical transport and dilution of air pollutants. When the atmospheric lapse rate is greater than the dry adiabatic one, the atmosphere is in unstable condition and thermal vertical transport occurs. A second mechanism of vertical exchange is the entrainment of free tropospheric air into the turbulent boundary layer at the interface between the two regions. This may happen when air in the lower free troposphere undergoes radiative cooling or cooling from the evaporation of recently entrained cloudy air. Such entrainment reduces the depth increase of the boundary layer. Frontal uplifting is a third important vertical-exchange mechanism. Advancing fronts displace air both horizontally and vertically. Clouds and precipitation associated with fronts are caused by the uplifting, and the boundary layer air is lifted into the free troposphere by this process. The traveling mid-latitude and tropical disturbances make this mechanism most effective. A particular case is the so-called “heat island” effect, as often seen in modern cities. The heat island results from a mass of material, either natural or anthropogenic, that absorbs and reradiates heat at a greater rate than the surrounding area. This causes moderate to strong vertical convection currents above the heat island. Such effect is superimposed on the prevailing meteorological conditions but is nullified by strong winds. Typical examples of places that have a heat island are large industrial complexes and small to large cities. Due to heat island effects, urban atmosphere is less stable than the atmosphere over the surrounding countryside. This is mostly apparent in cold winters when considerably more fuel gets consumed for heating. It has been reported that some European cities average as much as 5 oC above the surrounding countryside. Depending upon the location and elevation of the pollutant sources, heat island effects can be either favorable or unfavorable for the dispersion of air pollution. For ground level sources such as automobiles, the unstable air resulted from the heat island effect will allow a greater air volume for dilution of the pollutants. On the other hand, plumes from tall stacks that would be carried out over the countryside without increasing ground level pollutant concentrations under stable conditions, will be mixed by the heat island instability to the near ground level, resulting in higher concentrations. In conclusion, the transport of pollutant is enhanced if vertical-exchange processes transfer these substances from levels where wind speeds are low and removal processes are active into regions of high winds, vertical stability, and inefficient removal. This means a pollutant can be carried further before it reaches the ground, resulting in a diluted ground level concentration. Such vertical transfers are most often caused by convection and frontal lifting, carrying pollutants to above the surface boundary layer into the free troposphere.
4. Atmospheric Removal Processes 4.1 Wet Deposition Air pollutants are finally removed from the atmosphere by one of two mechanisms: wet deposition and dry deposition. The absorption of pollutants into droplets followed by droplet removal by precipitation is referred to as wet deposition. A number of different terms are used more or less synonymously with wet deposition, such as precipitation scavenging, wet removal, washout, and rainout. These terms refer to the removal of pollutants from the atmosphere by various
types of precipitation, rain, snow, etc. Washout is sometimes used specifically to refer to the process of in-cloud scavenging and rainout to refer to below-cloud scavenging by falling rain, snow, or hail. Droplets in clouds or falling raindrops are often termed ‘hydrometeors’. 4.2. Dry Deposition Dry deposition refers to the transfer of pollutants in the atmosphere, both gaseous and particulate, to the earth’s surface, including soil, water, and vegetation, where it is removed. There are three distinct steps constituting the dry deposition process. The first step involves the transport of the pollutant through the surface layer to the immediate vicinity of the earth’s surface. This step is controlled by turbulent diffusion in the atmospheric surface layer, and is viewed as the aerodynamic component of the transfer. The second step consists of the diffusion of the pollutant through the laminar sublayer just adjacent to the surface to the ultimate absorbing substrate. This step is normally called the surface component of the transfer. Although the laminar sublayer is typically so thin as only to the order of 10 -1 to 10-2 cm, the diffusion through this layer can be critically important in the overall rate of deposition, if this is the control step of the whole process. The solubility or absorptivity of species at the surface determines how much of the species that diffuses through the laminar sublayer actually is removed, and this final process, called the transfer component, constitutes the third step. Some non-reactive gaseous species such as argon or helium are almost not removed at all by dry deposition, since they are not absorbed even if they have diffused to the surface. Because the mechanisms of transport in the process of dry deposition are complex, it is very difficult to represent the processes at their most fundamental level of detail. Since the rate of removal of a species by dry deposition is of the greatest interest, it can be described in terms of dry deposition velocity vd, to simplify the whole process. The deposition velocity is best viewed as a proportionality constant between vertical flux and concentration, and it is to be determined empirically. The dry deposition process can be simulated conceptually with electrical or heat flow through a series of resistances. The three steps of transfer process are represented by three resistances, the aerodynamic resistance, the surface layer resistance and the transfer resistance, taking place from the atmosphere to the surface. These resistances are denoted ra, rs, and rt, having units of sec cm-1. Their relationship with vd is:
(1)
5. Atmospheric Diffusion 5.1. Atmospheric Diffusion Theories Study of the atmospheric aspects of air pollution is aimed at being able to describe mathematically the spatial and temporal distribution of contaminants released into the atmosphere. It is common to refer to the behavior of gases and particles in turbulent flow as turbulent “diffusion” or atmospheric “diffusion,” although the processes responsible for the observed spreading or dispersion in turbulence are not the same as those acting in ordinary molecular diffusion. A more precise term would be atmospheric dispersion, but to conform to common terminology atmospheric diffusion is used here. Because of the inherently random
character of atmospheric motions, one can never predict with certainty the distribution of concentration of marked particles emitted from a source. Although the basic equations describing turbulent diffusion are available, atmospheric flow and turbulence are so complicated that there is no single mathematical model which can be used as a practical means of computing atmospheric concentrations over all ranges of conditions. There are two basic ways to describe turbulent diffusion the Eulerian approach and the Lagrangian approach. The mathematical relationships based on the two approaches are dissimilar, but the ultimate pollutant concentrations can be related. Each of the two models is a valid description of turbulent diffusion; the choice of which approach to adopt in a given situation will be dependent on the specific features of the problem. The behavior of pollutant in the Eulerian approach is described relative to a fixed coordinate system. This type of description is the common way of treating heat and mass transfer phenomena. It intends to formulate the concentration statistics in term of the statistical properties of the Eulerian fluid velocities, which were measured at fixed points in the fluid. A formulation of this type is very useful not only because the Eulerian statistics are directly measurable (as derived from continuous time recordings of the wind velocities by a fixed network of anemometers), but also because the mathematical expressions are directly applicable to situations in which chemical reactions are an important component. But a big drawback is that the Eulerian approaches lead to a serious mathematical obstacle known as the closure problem, for which no generally valid solution has yet been obtained. The second approach is the Lagrangian method in which concentration changes are described relative to a mass of moving air. It attempts to describe the concentration statistics by means of the statistical properties of the displacements of groups of pollutants released in the air. The mathematics of this approach is more tractable than that of the Eulerian approach, and no closure problem is encountered. But the applicability of the resulting equations is limited because of the difficulty of accurately determining the required pollutant statistics. Moreover, if nonlinear atmospheric chemical reactions are involved, the equations are not directly applicable. In summary, it can be seen that both approaches have certain inherent difficulties, making it impossible to achieve an exact solution for the mean concentration of pollutants in turbulent flow. For more practical purposes, several approximate theories have been used for computation of mean concentrations of pollutants in turbulence. The K-theory is most commonly used to solve atmospheric diffusion equations, and the statistical theory is of significant practical value, based on the behavior statistics of individual particles in a stable, homogeneous turbulence field. 5.2. Gaussian Plume Equation If the determination of the concentration of reactive pollutants is to be obtained (involving the physics of the turbulent flow and mixing processes and the chemical interaction of reacting gaseous species), a numerical calculation approach must be applied. Such numerical analysis is very time-consuming even if a modern high-performance workstation is available. Solutions of this type for just a single set of meteorological conditions require the expenditure of a large computational effort. Even further, effort and cost are needed for the prediction of long-term averages of compound concentration. For most practical situations, the Gaussian plume model is a simple method to apply, and can give useful results provided its limitations are properly observed.
Under certain idealized conditions, the mean concentration of a species emitted from a point source has approximately a Gaussian distribution. This serves as the basis for a large class of simplified atmospheric dispersion formulas in common use, although they are most valid only in the case of a stationary, homogeneous turbulence field. Gaussian-based formulas are widely used and sufficiently accurate for most practical applications. The basis of these formulas is the widely known Gaussian plume equation, an expression for the mean concentration of a species emitted from a continuous, elevated point source. The Gaussian plume equation is based on the approximation that the concentration downwind of a point source in the atmospheric boundary layer follows Gaussian distribution but with unequal dispersion coefficients in the horizontal ( ) and vertical ( directions. Thus the downwind concentration at a location of x, y, and z reads:
)
(2) Where Qp is the emission intensity of the pollution source (in grams per second), and u is the average wind speed. The Gaussian plume method has been widely used as the computation basis in a number of computer programs, which are available for selected special applications. 5.3. Atmospheric Diffusion Equation Although the Gaussian equations have been widely used for air pollution dispersion calculations, the inability to account changes in wind speed with altitude and nonlinear chemical reactions limits the situations in which they may be used. The atmospheric diffusion equation provides a more general approach to atmospheric diffusion calculations than the Gaussian models, based on the K-theory. It can be proved that the Gaussian model is a special case of the diffusion equation when the wind speed is uniform and the eddy diffusivities are constant. The atmospheric diffusion equation in the absence of chemical reaction is
(3) The key issue in the use of Eq.(3) is to choose the functional forms of the wind speeds, u, v, and w, and the eddy diffusivities, Kx, Ky, and Kz, for the particular case of interest. In view of the hypothetical nature of the closure relationship and the other approximations involved, this equation must be considered as an approximate, semi-empirical expression. The validity of the atmospheric dispersion equation is to be judged by its ability to achieve predictions that agree either with observations under a variety of conditions, or with solutions to a series of problems obtained by another method of known veracity. The atmospheric diffusion equation is to be solved numerically.
6. Air Pollution Modeling 6.1. Dispersion Modeling A dispersion model refers to a mathematical description of the atmospheric transport and dispersion process quantified in terms of source and meteorological parameters for a given time. The outputs of the dispersion models include the desired estimates of concentration of a particular pollutant for a specific location and time period. Such a model is first subject to verification, which is carried out by comparing actual measurements of concentrations of the particular atmospheric pollutant with the calculated results from the model by means of statistical techniques. Normally meteorological parameters required for use of dispersion models include wind direction, wind speed, and atmospheric stability. In some cases, the lapse rate and vertical mixing height may also be desirable. Most models will require source data about the physical stack height, the diameter of the stack at the emission discharge point, the exit gas temperature and velocity, as well as the mass rate of emission of pollutants.
Models are usually classified into either short-term or climatological categories. Short-term models are generally used under the following circumstances: 1) to estimate ambient concentrations where it is difficult to sample, or from an unreal pollution source; 2) to estimate the required emergency source reductions during periods of air stagnation under air pollution episode alert conditions, and 3) to estimate the most probable locations of high, short-term, ground-level concentrations as part of a site selection evaluation, for the localization of air monitoring programs. Basically the climatological model is a mathematical description of the meteorological transport and dispersion processes of an area on which are superimposed rates of emissions of pollutants from various sources. These models are used to estimate mean concentrations over a long period of time or to estimate average concentrations at particular times of the day for some period such as each season. If the model is proved to be sufficiently sound and input data are accurate enough, the concentration of any pollutant at any point in the area at any time is obtainable from the model. The establishment of a suitable model for a locality would help the air pollution control agency to make a great many decisions in a scientific way, such as: how much of an increase or decrease in emissions from a given location would affect downwind atmospheric concentrations; where to locate samplers to measure the greatest effect of a given source; how to relate emission regulations to air quality standards; where to locate new pollution sources; how much will a new zoning ordinance affect air quality; and so on. The description of the dispersion itself as affected by stability, wind direction and speed, source height, buildings, and topography is the most important step in model development. Many models make use of a Gaussian distribution which varies between elevated and groundlevel sources, and which choose most suitable plume dispersion parameters for various stability types. Of course, no model can provide good results with poor input data; this may constitute a significant difficulty with urban pollution modeling, as great uncertainties do exist in pollutant emissions in urban areas, particularly the variations with time. As more complex models are being developed, much more time is needed for tedious, repetitive calculations in order to validate these models. Meanwhile, the performance of computers used for model calculations has been improved so much that more complicated modeling problems can be handled easily. Even for stack height determination studies, many consulting and design organizations have developed computer codes for sorting out meteorological data and calculating concentrations for a variety of assumed combination of conditions. In some cases, the nature of local meteorology and available meteorological data or the complexity of atmospheric dispersion may be such that more knowledge of atmospheric flow and dispersion is necessary. This is usually achieved by a tracer approach. Solid particles or gases of unique composition are released at a stack or other source at a known rate and for a known period of time. The air at downwind points is then collected and measured by a filter or other suitable equipment to quantify the concentration of the unique material. Analysis of the material collected at a sufficient number of stations allows one to determine diffusion patterns of the stack effluent under existing weather conditions. Finely divided fluorescent particles, like zinc and cadmium sulfides, have been found to be suitable for this purpose. In one test of this type, concentrations of more than 200 particles per cubic meter were found 107 miles from the discharge point. When plume movement may be affected by nearby
structures or topography, wind tunnel experiments often provide information of value. This has been used to determine the effects of adjacent building shape on the behavior of stack effluents in power plants, and the roadside building configuration on the dispersal of motor vehicle exhaust in an urban street. 6.2. Air Quality Models The physical models, originating as prototypes of practical atmospheric dispersion models, are intended to simulate the atmospheric processes affecting pollutants by means of a representation of the actual air pollution problem. It is a simplified air quality model for research purposes only. A physical model sometimes employed to study the dispersion of pollutants consists of a small-scale replica of the urban area or a portion thereof in a wind tunnel. The problems associated with properly duplicating the actual atmospheric scales of turbulent motion make physical models of this variety of limited usefulness. While useful for isolating certain elements of atmospheric behavior and invaluable for studying certain critical details, physical models cannot serve the needs of ambient air quality models capable of relating emissions to air quality under a variety of meteorological and source emission conditions over an urban area. Actual air quality models, sometimes named mathematical models, can broadly be classified under two types: (1) models based on statistical analysis of past air monitoring data, and (2) models based on the fundamental description of atmospheric transport and chemical processes. The emissions of pollutant can be related to atmospheric concentrations throughout an ambient air quality model. Such a link between emission changes from source control measures and the resultant changes in atmospheric pollutant concentration is very helpful for decision-making. Considerations of emission patterns, meteorology, chemical transformations, and removal processes are usually involved in air quality models (see Figure 3).
Figure 3. Components of a comprehensive air quality model In planning for the abatement and control of air pollution in an urban area, many difficult questions and complex issues will be encountered. Most of they can be best addressed through the use of a comprehensive urban air quality model, in some cases there are no alternative means for examining the critical issues.
7. Pollution Accidents and Meteorological Control When air pollutants are held over a city for an extended period, either because of a strong inversion or because of the city’s stagnant local meteorology, the buildup of pollution can develop situations known as “smog”. There are two distinct forms of smog: that typically
associated with London, which gave the phenomenon its name (smoke plus fog), and the more recent type associated with Los Angeles, which is entirely different in origin and is generally termed photochemical smog. The classic London-type smog arises from high concentrations of particles and sulfur dioxide (originated from coal burning) in the atmosphere when the humidity is high, while the most severe photochemical smog is associated with the interactions between nitrogen oxides and hydrocarbons (from exhaust of motor vehicles) in the presence of strong sunlight. Making use of meteorological data can play a significant role in the air resource management program. It is well known that meteorological conditions vary tremendously form one day to another, form one season to another, and from one location to another. There are times and places when almost any amount of pollutants can be dispersed and diluted readily, while at other times and in other places, a minor emissions can be a problem. Humans must adapt their activities to the variations of weather conditions, so that it will be most favorable to good ventilation and adequate dispersion of air pollutants. Theoretically all major pollution sources not subject to emission control could be located in those areas where atmospheric dispersion processes were most favorable, or could be so managed that emissions were nil or at a minimum during times of adverse conditions. These measures are commonly known as landuse control and meteorological control, respectively. Land-use control entails consideration of air pollution factors in zoning decisions in a region. Meteorological control of pollutant emissions is simple in concept: maximum emissions are allowed when good atmospheric dispersion is expected, and they are minimized when the air is stagnant. Such controls have been applied where emissions could be interrupted or postponed without unbearable economic penalty. For example, it is common practice in many areas to restrict open burning of trash and other wastes to those days or hours when atmospheric conditions are favorable to rapid dispersion. Open burning is never allowed at night in some localities because of the great likelihood of stagnant air. Other localities disallow certain activities with high emissions when the wind speed exceeds a certain value; these activities could include plowing, harrowing, land clearing, building demolition, etc. There are some cases where meteorological controls are rarely applicable, such as chemical plants, pulp mills, etc., where processes are continuous in nature. Shutting-down operation in these places may cause even more pollution and heavy economic penalty. In a few particular cases, courts have ordered cessation or curtailment of certain emissions under specified wind and weather conditions.
8. Remarks Air pollution dynamics and modeling have undergone a period of intensive studies in the last 30 to 40 years. A lot of methodologies have been developed and applied in many actual situations and resulted in an abundance of practical experience. While theory problems have almost been resolved up to now except for those inherent to meteorology and atmospheric turbulence, practical difficulties still exist in various applications of these theories. There are a number of air quality models available worldwide, based either on a numerical approach when chemical reactions are involved or on Gaussian equations if non-active species is treated, but they all have their own shortcomings. In most cases the emission and dispersion of particulates are not dealt with, and knowledge of atmospheric aerosol behavior and its interactions with gaseous compounds are not fully established. The complex configurations of urban surface cannot be sufficiently accounted for in numerical urban air pollution models at resolutions high enough for the needs of human exposure study. When statistical approaches
are adopted to develop micro-scale dispersion models, they are generally not of a universal nature. The inadequacies of current air quality models are apparent when their predictions are compared with observations. Due to the turbulent and constantly changing nature of the urban atmosphere, satisfactory accuracy is difficult to achieve, and will require a significant effort in the future.
Related Chapters Click Here To View The Related Chapters Glossary Dry adiabatic Mixing depth Temperature profile
: the condition of no heat exchanging with outside air : the height to which the warm air rises and mixes with the cooler air until it meets its equal in temperature : the variation of temperature with elevation
Bibliography Elsom D.M. (1992). Atmospheric Pollution-A Global Problem. Blackwell, USA. [The book presents an overview of worldwide regulations on motor vehicle emissions]. Hamilton R.S. and Harrison R.M. (1991). Highway pollution, 339pp. Elsevier Science Publishers, Amsterdam, Netherlands. [The book presents a series of articles on the dispersion, environmental impacts, and control strategies of motor vehicle emission]. Heinsohn R. J. and Kabel R.L. (1999). Sources and Control of Air Pollution. Prentice Hall, Upper Saddle river, New Jersey, USA. [This discusses major air pollutant source characteristics]. Knap A.H. (1988). The Long-Range Atmospheric Transport of Natural and Contaminant Substances. Kluwer Acadamic Publishers, Dordrecht, Netherlands. [The book described the characteristics of long distance air pollution transport]. Lindberg S.E., Page A L. and Norton S.A. (1990). Acidic Precipitation, 222pp. Springer-Verlag Newyork Inc., USA. [This describes dry and wet deposition dynamics of acid species]. Scorer R.S. (1990). Meteorology of Air Pollution. Ellis Horwood, Chichester, England. [This covers the major meteorological conditions that are important for air pollution dispersion]. Seinfeld J.H. and Pandis S.N. (1998). Atmospheric Chemistry and Physics. Wiley, New York, USA. [This book gives a systematic description of atmospheric diffusion physics and dynamics]. Biographical Sketches Mr. Lixin Fu is an associate professor in the Department of Environmental Sciences and Engineering, Tsinghua University, Beijing, China. He was born in May 1966 in Hunan Province of China. He received his Ph.D. degree from Tsinghua University in 1998. Mr. Yang Chen is a research student working for a master degree in the same department. He was born in June 1974 in Zhejiang Province in China. To cite this chapter Lixin FU, Yang CHEN, (2005), AIR POLLUTION DYNAMICS AND MODELING, in Point Sources of Pollution: Local Effects and It's Control, [Ed. Qian Yi], in Encyclopedia of Life Support Systems (EOLSS), Developed under the Auspices of the UNESCO, Eolss Publishers, Oxford ,UK, [http://www.eolss.net] [Retrieved August 31, 2007]
