The primary air quality model in use today in the U.S. is the AERMOD dispersion model.  Extensive testing has been performed to validate AERMOD.  Most of the testing focuses on point sources, such as stack emissions. How accurately does AERMOD estimate air quality impacts for area sources?

Similar to point source treatments, AERMOD can be expected to produce relatively unbiased distributions of exposure that will occur at various locations.   It cannot be expected to match measured data in time and space. What are the area source treatment limits in AERMOD (and predecessor models such as ISCST3)?

As with most analyses, it depends.  Regarding area sources, the question needs to be focused on the energy exchange of the area source compared with the basis for the AERMOD dispersion rates. The dispersion rates in AERMOD are based on Project Prairie Grass (PPG).  The PPG study conducted in the summer of 1956 in O’Neil, Nebraska, is likely the most extensive test of atmospheric dilution rates ever conducted.

The question is: how well does AERMOD perform for area sources that have a high heat capacity and thermal conductivity? Examples include the application of agricultural fumigants on moist fields and wastewater lagoons. The most limiting air quality impacts from surface-based area sources are during nocturnal conditions with highly stable atmospheric conditions.  The PPG study included the evaluation of atmospheric conditions during nocturnal periods. Still, there are two significant limitations:  (1) soil conditions during the summer of 1956 in O’Neil, Nebraska, were at or near the soil wilting point, and (2) except four tracer release periods early in the PPG study, periods with wind speeds less than 2 m/sec were not included during nocturnal periods.

The gradient Richardson Number is a standard indicator of atmospheric stability:

  [Eq. 1]

In other words, the denominator of the Richardson number represents the shear production of mechanical turbulence that acts to produce atmospheric dilution. In contrast, the numerator (N) represents the suppression of turbulence (if the temperature profile is positive/stable) or the enhancement of turbulence (if the temperature profile is hostile/unstable).

We have found through research on approximately 50 studies of agricultural fumigants that the numerator of the gradient Richardson number trends towards isothermal conditions, i.e., there are rarely significant inversion conditions during nighttime periods that would be treated as stable conditions based on AERMOD.  This can produce neutral atmospheric conditions for area sources with high heat capacity and high thermal conductivity.  If the winds are light enough, e.g., < 2 m/sec, insufficient mechanical turbulence can be generated to promote neutral atmospheric conditions, and stable conditions can still occur.  In other words, atmospheric stability over such area sources tends to neutral atmospheric conditions unless wind speeds are very low.

Our team was the Principal Investigator for EPA’s first major urban-scale air toxics study to model air quality.  This was the Philadelphia Integrated Environmental Management (IEMP) study of Philadelphia in the early 1980s.  We inventoried emissions for a large number of chemicals.  

Furthermore, these data are modeled and compared to a 10-station air monitoring network we established.  The modeled and measured data additionally matched reasonably well.  Our team had a question: what statistical tests should we do to demonstrate model performance? 

Our team contacted John Irwin, a senior modeler at the EPA. He responded to us—before the internet—and encouraged us to conduct any tests we desired, emphasizing that those tests were just for reference. Moreover, he stressed that allowing the measured data to “teach” us how to model effectively is far more critical. His insights continue to guide our approach to air quality modeling even 40 years later.

Models are complex.  Furthermore, many equations are used, and modeling involves sophisticated methodology.  A modeler can have an excellent grasp of all these concepts and calculations.  So, the data knows more.  It was there then and can provide a roadmap for refining the analysis.