Minimum value for lateral turbulence (aka, minimum ov)
Overview of Issue
The lateral turbulence (or the standard deviation of lateral velocity to the average wind direction),
commonly referred to as ov, is the amount of fluctuation in the wind speed in the direction
perpendicular to the mean wind and represents the turbulent flow across a plume in the boundary
layer. The lateral dimension of a plume is directly correlated to the value of ov, with larger oy values
resulting in larger plumes and lower concentrations. Thus, the observed or estimated value of ov has a
direct impact on model predicted concentrations. The formulations to calculate ov result in values of ov
of zero when wind speeds approach zero (i.e., low wind conditions). However, field data suggests that
the minimum values of ov do not approach zero. As a result, a minimum value for ov has been
implemented in the AERMOD dispersion model (and other dispersion models) to adhere with the
observed field data.
In other words, estimates of ovin low wind conditions are often too small (approaching zero), requiring a
minimum value to be set. During these low wind conditions, plume volumes are inherently small,
generally resulting in higher concentrations. Increasing the minimum ov will result in lowering the
maximum concentrations estimated for surface releases. For elevated releases, specifically when terrain
considerations are important, increasing the minimum ov may increase or decrease concentrations,
depending on conditions. It is expected to likely lower most concentrations, though this may not
correspond to the highest modeled concentrations.
Current Implementation in AERMOD
The calculation of ov in AERMOD is described in section 4.1.6 of the AERMOD Formulation and
Evaluation Document. The total lateral turbulence is the sum of the mechanical (ovm) and convective
(ovc) portions:
Ovt ®vc. "I" ^vm
The mechanical turbulence at the surface is a function of the surface friction velocity (u*):
Gym = 3.6 * ul
and varies linearly from the surface up to the top of the mechanically mixed layer. The convective
turbulence within the mixed layer is constant and is a function of the convective velocity scale (w*):
-------
Summary of Current Literature or Research
The lateral turbulence parameter, ov, has been discussed in a number of journal articles over the past
few years; however, there is not a specific emphasis on demonstrating what value should be selected as
the minimum ov.
Hannah et al., 1985
The more current literature suggests that this paper is the basis for the default selection of the minimum
ov in AERMOD. Figure 3 of this paper shows observation data collected aboard a research vessel
operated off the California coast from four different research cruises. The ov values for this analysis were
calculated as the product of the mean wind (U) and the standard deviation of the wind direction (oe).
The figure shows the range of ov values, with an apparent lower limit of 0.175 m/s and a mean value of
0.5 m/s.
Luhar, 2009
This work examines the various experimental and analytical methods employed to determine ov and ou,
or the longitudinal turbulence. There is particular emphasis on the methodologies under low-wind
conditions, when the accuracy of existing methods is more sensitive to the method selection. The work
identifies two equations typically used to determine ov from field data:
ov = U * sin 
-------
minimum ow value of 0.02 m/s). The paper proposes that at low wind speeds, these two differences
alone could result in AERMOD having concentrations 12.5 times higher than SCICHEM.
The paper also presents a model evaluation between SCICHEM, the base version of AERMOD 12345, and
one version of AERMOD that applies the adjusted u* approach, a minimum ov of 0.3 m/s, and
modification to the application of the pancake plume for the Oak Ride and Idaho Falls field study
databases. The results suggest improved model performance for the beta AERMOD options, though it is
unclear which of the options have the greatest impact on performance. SCICHEM performance is similar
to the beta AERMOD performance.
Hoinaski et al., 2017
This work examines the estimates of ov and resulting estimates of oy in AERMOD based on two field
studies in USA's Round Hill II (Cramer, 1957) and Germany's Uttenweiller (Bachlin, 2002) experiment
databases. The work emphasizes the effect of the averaging time for the calculation of the
meteorological model inputs and concentrations. The work does not directly address the minimum o
value, but demonstrates the sensitivity of modeled results (and modeled over-predictions) to the
estimation of the ov by also running AERMOD with on-site values of ov. The work suggests that the
Lagrangian time scale might also need examination, particularly for longer travel times. It should be
noted that the data from these field studies range from 30-s averaging times up to 10-min, well below
the standard time step for AERMOD (1-hour).
Considerations for Updates in Model System
As outlined above, the ov value has a very direct impact on plume size and modeled concentrations. It
may seem like adjusting the minimum sigma v is straightforward way to address modeled over-
predictions for low wind conditions for surface releases. However, the findings in these papers show
clearly that ov values can be lower than the current default of 0.2 m/s. Additionally, adjusting the
minimum ov values may "fix" some over-predictions for surface releases, but may negatively affect
modeled predictions for elevated releases. The reviewed literature points to several methods to
determine the ov value that should be considered rather than simply adjusting the minimum ov value.
The data sets analyzed are also fairly limited, which suggest more data sets should be identified or made
available to investigate the issue.
3
-------
References
Bachlin, W., Ruhling, A., and Lohmeyer, A. (2002). Bereitstellung von validierungs-daten fur
geruchsausbreitungs-modelle-naturmessungen. Baden-Wurttemberg, Germany.
Cramer, H. E., and F. A. Record. (1957). Field studies of atmospheric diffusion and the structure of
turbulence. Am Ind Hyg Assoc Q. Taylor & Francis, 18:126-131.
Hanna, S.R. and B. Chowdhury. (2014). Minimum turbulence assumptions and u* and L estimation for
dispersion models during low-wind stable conditions, Journal of the Air & Waste Management
Association, 64:3, 309-321, DOI: 10.1080/10962247.2013.872709.
Hanna, S.R., L.L. Schulman, R.J. Paine, J. Pleim, and M. Baer. (1985). Development and Evaluation of the
Offshore and Coastal Diffusion Model. J. Air Poll. Control Assoc., 16, 1039-1047.
Hoinaski, L., Franco, D. and Henrique de Melo Lisboa. (2017). An analysis of error propagation in
AERMOD lateral dispersion using Round Hill II and Uttenweiller experiments in reduced averaging
times. Environmental Technology, 38:5, 639-651, DOI: 10.1080/09593330.2016.1205672.
Luhar, A.K., Estimating Variances of Horizontal Wind Fluctuations in Stable Conditions, Boundary-Layer
Meteorol (2010) 135: 301. https://doi.org/10.1007/sl0546-010-9480-5.
Sagendorf, J.F., and C.R. Dickson. (1974). Diffusion under low windspeed, inversion conditions. NOAA
Technical Memorandum. ERL ARL-52, Air Resources Laboratory, Idaho Falls, 89 pp.
4
-------