Air Quality Models: Application

From a sign at EPA's Office of Air Quality Planning and Standards: "The Modeler's Creed: When the data and model disagree, the data are wrong."

It is important to understand where AQMs stand in the "larger scheme" of things. The graphic at right (click to see full-sized, this is a large image!) shows in flowchart form the air pollution system, which includes the science and the public policy/legislative components. Notice that the modeling part is only one component of the overall air quality analysis picture. The specific purpose of air quality modeling is to determine the best control strategy by which air quality can be improved in some geographical area. If there is anything missing from this chart, it is any direct mention of the economic implications of control measures, but that is perhaps incoporated under "Legislation". You should recognize that there is considerable politics between the boxes "Control strategy options" and "Control measures". Evidence of that is the recent Supreme Court case regarding new air quality standards for atmospheric pollutants.

A complete understanding of the air pollution system diagram above is essential for success in any part of air quality work. The entire OS411 series has been designed to focus on specific parts of this diagram, especially the large areas of meteorology and tropospheric chemistry.

The graphic at right (click to see full-sized) shows a general scheme for air quality models. In this graphic, meteorological and emissions data, combined with the users control strategies, all combine to provide input to the air quality model, resulting in some type of dataset as a result.

In many air quality models, the meteorological data is approximated, collected in the field, or both, while the emissions inventories are often predicted using an emissions model. Emissions models typically use the principle of mass balance, and assume that emissions from a particular source for a specific pollutant in a specified time frame are equal to the product of the activity of the source in the unit activity. For a more detailed discussion of emissions inventories, the reader is directed to the Emissions section of the reading "Conceptual Plan for Air Quality and Meteorological Modeling".

Most air quality models require expertise in several disciplines for satisfactory usage. One of the goals of this entire package of materials is to help the student develop some degree of competence in each of these areas so that he or she might be able to understand specific air quality models. Most of the topics below are discussed at some stage in the materials. The optional case study reading, and, more importantly, most air quality models, require the analyst to be able to deal with the majority of the areas listed below:

• Meteorology and atmospheric physics
• Atmospheric chemistry
• Emissions inventories
• Computer Science (and computational science), which includes numerical analysis
• Regulatory Issues and Processes

The fundamental goal of most air quality models is that of regulatory compliance, in most cases the target chemical being tropospheric ozone. The US Environmental Protection Agency (EPA) requires communities to meet the air quality specifications as outlined by the National Ambient Air Quality Standards, or NAAQS. These standards specify the maximum concentration of a particular species that can occur in a specific period of time. For ozone, the NAAQS standard has been 0.12 parts per million (ppm) averaged over a one-hour period.

EPA has recently changed the one-hour primary NAAQS for ozone with a new eight-hour, 0.08-parts-per-million (ppm) primary standard. The 0.08-ppm standard would be met when the "three-year average of the third-highest daily maximum eight-hour average ozone concentration is less than or equal to 0.08 ppm." This decision was based upon a conclusion that such a standard would not only protect sensitive populations, but "provide a more stable basis upon which the States can design and implement their ozone control programs" as well. EPA seeks comments on other concentration-based forms "within the range of the second to the fifth highest daily maximum eight-hour average ozone concentrations."

The two examples below show how the EPA measures compliance in a community. Field monitors measure the amount of ozone over an eight-hour period daily (typically these measurements are from 8 am to 4 pm) for a period of three years. The reading for the particular area is calculated by taking the average of the third highest reading for each of the three years.

Example 1. Ambient monitoring site attaining the primary O₃ standard.

The primary standard is met at this monitoring site because the three-year average of the annual third-highest daily maximum eight-hour average O₃ concentrations (i.e., 0.078 ppm) is less than or equal to 0.08 ppm. The data completeness requirement is also met because the average percent of days with valid monitoring is greater than 90 percent, and no single year has less than 75 percent data completeness.

Example 2. Ambient monitoring site failing to meet the primary O₃ standard.

The primary standard is not met at this monitoring site because the three-year average of the third-highest daily maximum eight-hour average O₃ concentrations (i.e., 0.099 ppm) is greater than 0.08 ppm. Note that the O₃ concentration data for 1994 is used in these computations, even though the data capture is less than 75 percent, because the third-highest daily maximum 8-hour average concentration for that year is greater than 0.08 ppm.

What the air quality model looks to do is to incorporate all, and hopefully the best, of what is known about the science of atmospheric processes in order to allow the user to test a wide variety of "what-if" scenarios and the impact of a variety of "forecasts" based on a number of independent variables. Unfortunately, a great deal is still unknown about the science of how species interact in the atmosphere. Harvey Jeffries, an atmospheric chemist/modeler at UNC-Chapel Hill, makes the point that "methane is the only commonly present hydrocarbon in urban air whose atmospheric chemistry is so well understood that all of its reactions and rate constants and those of its reaction products can be written down- and methane is excluded from (regulatory) control!". He further elaborates that for the majority of chemical species in the atmosphere, the modeler can only make "reasonable" guesses as to the mechanisms and reaction rates of the 20 most frequently observed hydrocarbons in urban areas. Given this level of uncertainty, Dr. Jeffries raises the question as to the usefulness of air quality models:

Atmospheric chemistry models that have the potential to be useful in air quality policy decisions (then) are obviously not completely based on science but instead require that the model builder create model species, create their reactions, and the rate constants for these reactions. While it is true that this creative process is often based on similarity hypothesis with known species, this is closer to an art than it is to a science. It certainly is not done the same way by different model builders. Most reaction mechanism builders believe that it is essential to have some observational evidence to guide their creative choices; this is usually smog chamber data, which unfortunately only presents the macro transformations in systems (e.g., the time series data for oxides of nitrogen, initial hydrocarbons, and ozone); these macro transformations are the result of a sum of processes, some known and some unknown. The art lies in unraveling the extent to which the known and unknown chemical processes contribute to the macro transformations. Further, the number of smog chamber experiments that are available and are capable of assisting the mechanism developer in creating the model reactions among model species is rather limited.

Pressed with the need to make a forecast that has some basis in facts, both the cooperating scientist and the policy maker have been forced "to act as if" the model, based in part on science and in part on the modeler's creativity, would give accurate forecasts becuase it did give seemingly accurate predictions in a few test cases. It is obvious that such models can not fulfill the scientist's goal of explaining the world in terms of existing scientific knowledge because many of the entities represented in the model do not exist in the real world.

"This model is as well-formulated as any model I know, and it uses inputs and assumptions more likely than any other set, and therefore, until additional information becomes available that will change the model's formulation or the inputs, it makes sense to act as if this model's forecasts are accurate."

The complete document for the source above, "Science and Policy Interaction", is available as an Adobe Acrobat "pdf" file, and is recommended for a thorough understanding of the relationship between scientific "reality" and policy decision-making realities.

Another interesting article, "Air Quality Modeling: A Brave New World" is also available on the Web.