Model Description

The Oklahoma Dispersion Model is a simple tool that has been developed to aid in assessing the atmosphere's ability to disperse gases and small particulates that are released near the ground. In addition to dispersion conditions, it is also important to know the transport direction of such material (e.g., drift). The focus of the Model is the evaluation of dispersion conditions with respect to downwind concentrations at distances in the 1/4 mile to 2 mile range (although greater distances would also apply). With knowledge of current and future (predicted) conditions for these two items, one can better assess appropriate times that would minimize downwind pollutant concentrations resulting from the near-surface release of gases and particulates, whether they result from a pesticide application, a prescribed burn, or land application of animal waste.

With respect to dispersion conditions, the Oklahoma Dispersion Model is used in conjunction with weather conditions reported by the Oklahoma Mesonet to produce a map of current dispersion conditions across Oklahoma. Another part of the Model is used in conjunction with the latest 60-hour NGM MOS forecasts to produce maps of dispersion conditions at 3-hour intervals into the future. With respect to transport direction, Mesonet data are used to create a map of current wind conditions (as well as temperature and relative humidity), while the MOS forecasts are used to create similar maps at 3-hour intervals into the future. In addition, table output based on the NGM forecasts showing future dispersion and transport conditions is available for specific sites.

Dispersion Conditions

For short distances (less than 10 km or 6 miles), the dispersion of gas and small particulates is typically modeled by the Gaussian distribution, where the horizontal width of the pollutant plume has a "normal" distribution with standard deviation "sigma-y" and the vertical depth of the plume has a similar distribution with standard deviation "sigma-z". These standard deviations increase with distance from the emission source and are a function of the meteorological conditions.

At a given downwind distance, the concentration resulting from a surface-based emission source is proportional to 1 / [(wind speed)(sigma-y)(sigma-z)]. Thus, the greater the wind speed or the sigma values, the lower the concentration.

A "dispersion condition" rating scheme has been developed with respect to the downwind concentrations in the 1/4 mile to 2 mile range. The Model calculates a specific downwind concentration and, based on the algorithm, assigns one of six dispersion categories. Graphics and text output are also created. The following scheme is utilized:

	Dispersion Conditions	Code	Color of Maps
	Excellent	EX	Dark Green
	Good	G	Green
	Moderately Good	MG	Light Green
	Moderately Poor	MP	Beige
	Poor	P	Orange
	Very Poor	VP	Red

With respect to downwind sensitive areas, these categories are such that at a given distance (e.g., 1 mile) from the emission source, the concentration at the plume centerline (horizontal and vertical center) is smallest under "Excellent" conditions and greatest under "Very Poor" conditions. This scheme is thus conservative in the sense that the dispersion categories "protect" nearby sensitive areas in the path of the center of the plume; the wind may not carry the plume to a particular area, but if it does, these would be the resulting pollutant concentrations at that site. These categories are also generally applicable for concentration averages over a time scale of 15 minutes to one hour.

The Oklahoma Dispersion Model utilizes currently recognized EPA recommended algorithms (USEPA, 1987) for the calculation of Pasquill-Gifford stability categories A-F (Turner, 1969). Sigma-y and sigma-z calculations utilize the Briggs (1973) equations.

The dispersion categories that result are best applicable for flat uniform terrain and no precipitation. Under light wind conditions, especially on clear nights, gases and particulates tend to "drain" gravitationally downslope and dispersion conditions may be worse than what the Model suggests. With variable terrain and vegetation (e.g., forested hilly terrain), results may also be different. During periods of precipitation, dispersion may be enhanced, resulting in better dispersion conditions than the model suggests. In general, however, the Oklahoma Dispersion Model provides useful information for the wide range of situations to be encountered throughout Oklahoma. It is not designed as a site-specific model taking local topography and vegetation into account.

A. Current Dispersion Conditions

For current conditions, the Model utilizes the latest weather conditions from the Oklahoma Mesonet. In particular, 15-minute averages of the following are utilized: 10-m wind speed; standard deviation of 10-m wind direction; solar radiation; and the vertical temperature gradient between 9 and 1.5 m.

1. Daytime Conditions

For daytime conditions with 10-m wind speed greater than 1 m/s (the threshold value of the 10-m Mesonet wind sensor), the Model uses the average of two methods to calculate Pasquill stability class. The first method utilizes solar radiation (the "SRDT" method) in conjunction with wind speed. The second method utilizes the standard deviation of wind direction (the "sigma-A" method) to get initial estimates for stability class; these are then adjusted to final values based on the wind speed value.

For 10-m wind speed less than or equal to 1 m/s, the Model uses only the solar radiation method (SRDT method) to calculate stability class.

2. Nighttime Conditions

For nighttime conditions with 10-m wind speed greater than 1 m/s, the Model uses the standard deviation of wind direction (sigma-A method) to get initial estimates for Pasquill stability class; these are then adjusted to final values based on the wind speed. If the wind speed is less than 2 m/s, the sigma-y value is based on Pasquill class D (neutral stability) to mimic documented plume meander under such conditions.

For 10-m wind speed less than or equal to 1 m/s, the Model uses class D to calculate sigma-y and utilizes the Mesonet tower vertical temperature gradient to assign a stability class (E or F) for the calculation of sigma-z.

B. Future Conditions

Because the NGM MOS forecasts only predict certain variables, we are limited in the methods we can use to predict dispersion conditions. However, these forecasts do include wind speed, cloud cover, and ceiling height, so that the Turner (1964) algorithms can be used to calculate a stability class (SC =1 to 7). We assign Turner classes 6 and 7 to Pasquill class F. As with current conditions, if the wind speed is less than 2 m/s and it's nighttime, we utilize class D to calculate sigma-y to mimic plume meander.

Transport Direction

A. Current Conditions

A map showing current wind speed and direction, as well as temperature and relative humidity, is available using the most recently observed conditions from the Oklahoma Mesonet. In addition, maps going back 6 hours in 15-minute intervals are available in the Related Links Section for the Dispersion Model. (These maps are useful to see if any wind shifts are moving toward your area).

B. Future Conditions

The NGM MOS forecast weather maps employ the same station plot as does the Mesonet map of current conditions. Like the dispersion condition maps based on the MOS forecasts, they are available in 3-hour increments through the duration of the forecast period.

REFERENCES

Briggs, G. A., 1973. Diffusion Estimation for Small Emissions, ATDL Contribution File No. 79, Atmospheric Turbulence and Diffusion Laboratory.

Turner, D. B., 1964. A diffusion model for an urban area. J. Appl. Meteorol., 3:83-91.

Turner, D. B, 1969. Workbook on Atmospheric Dispersion Estimates, Public Health Service, Publ. 999-AP-26, 84 pp.

USEPA, 1987. On-Site Meteorological Program Guidance for Regulatory Modeling Applications. EPA-450/4-87-013, U.S. Environmental Protection Agency, Research Triangle Park, NC.