Page 375 - Fundamentals of Air Pollution 3E

330 20. Air Pollution Modeling and Prediction

over the terrain feature. Because the dispersion is calculated from fluctua-
tion statistics, the meteorological measurements to provide data input for
the model are quite stringent, requiring the use of tall instrumented towers.
Evaluations both of the model (28) and of a screening technique derived
from the model (29) indicate that the model does a better job of estimating
concentrations than previous complex terrain models.

IV. MODELING POLLUTANT TRANSFORMATIONS

A. Individual Plumes

An understanding of the transformation of SO 2 and NO X into other con-
stituents no longer measurable as SO 2 and NO^ is needed to explain mass
balance changes from one plume cross section to another. This loss of the
primary pollutant SO 2 has been described as being exponential, and rates
up to 1% per hour have been measured (30). The secondary pollutants
generated by transformation are primarily sulfates and nitrates.
The horizontal dispersion of a plume has been modeled by the use of
expanding cells well mixed vertically, with the chemistry calculated for
each cell (31). The resulting simulation of transformation of NO to NO 2 in
a power plant plume by infusion of atmospheric ozone is a peaked distribu-
tion of NO 2 that resembles a plume of the primary pollutants, SO 2 and NO.
The ozone distribution shows depletion across the plume, with maximum
depletion in the center at 20 min travel time from the source, but relatively
uniform ozone concentrations back to initial levels at travel distances 1 h
from the source.

6. Urban-Scale Transformations
Approaches used to model ozone formation include box, gradient trans-
fer, and trajectory methods. Another method, the particle-in-cell method,
advects centers of mass (that have a specific mass assigned) with an effective
velocity that includes both transport and dispersion over each time step.
Chemistry is calculated using the total mass within each grid cell at the
end of each time step. This method has the advantage of avoiding both
the numerical diffusion of some gradient transfer methods and the distor-
tion due to wind shear of some trajectory methods.
It is not feasible to model the reaction of each hydrocarbon species with
oxides of nitrogen. Therefore, hydrocarbon species with similar reactivities
are lumped together, e.g., into four groups of reactive hydrocarbons: ole-
fins, paraffins, aldehydes, and aromatics (32).
In addition to possible errors due to the steps in the kinetic mechanisms,
there may be errors in the rate constants due to the smog chamber data