Page 371 - Fundamentals of Air Pollution 3E
P. 371

326               20. Air Pollution Modeling and Prediction

       level, such as automobile pollutants in urban areas, and to continuous or
       instantaneous ground-level area sources. It is not appropriate for elevated
       point source diffusion until the plume has grown larger than the space
       scale. Numerical rather than analytical solutions of Eq. (20-18) are used.
         Errors in advection may completely overshadow diffusion. The amplifi-
       cation of random errors with each succeeding step causes numerical instabil-
       ity (or distortion). Higher-order differencing techniques are used to avoid
       this instability, but they may result in sharp gradients, which may cause
       negative concentrations to appear in the computations. Many of the numeri-
       cal instability (distortion) problems can be overcome with a second-moment
        scheme (9) which advects the moments of the distributions instead of the
       pollutants alone. Six numerical techniques were investigated (10), including
        the second-moment scheme; three were found that limited numerical distor-
        tion: the second-moment, the cubic spline, and the chapeau function.
          In the application of gradient transfer methods, horizontal diffusion is
        frequently ignored, but the variation in vertical diffusivity must be approxi-
        mated (11-14).


        D. Trajectory Models
          In its most common form, a trajectory model moves a vertical column,
        with a square cross section intersecting the ground, at the mean wind
        speed, with pollutants added to the bottom of the column as they are
        generated by each location over which the column passes. Treatment of
        vertical dispersion varies among models, from those which assume immedi-
        ate vertical mixing throughout the column to those which assume vertical
        dispersion using a vertical coefficient K z with a suitable profile (15).
          Modeling a single parcel of air as it is being moved along allows the
        chemical reactions in the parcel to be modeled. A further advantage of
        trajectory models is that only one trajectory is required to estimate the
        concentration at a given endpoint. This minimizes calculation because con-
        centrations at only a limited number of points are required, such as at
        stations where air quality is routinely monitored. Since wind speed and
        direction at the top and the bottom of the column are different, the column
        is skewed from the vertical. However, for computational purposes, the
        column is usually assumed to remain vertical and to be moved at the wind
        speed and direction near the surface. This is acceptable for urban application
        in the daytime, when winds are relatively uniform throughout the lower
        atmosphere.
          Trajectory models of a different sort are used for long-range transport,
        because it is necessary to simulate transport throughout a diurnal cycle in
        which the considerable wind shear at night transports pollutants in different
        directions. Expanding Gaussian puffs can be used, with the expanded puff
        breaking, at the time of maximum vertical mixing, into a series of puffs
   366   367   368   369   370   371   372   373   374   375   376