July 31, 1989



MEMORANDUM

SUBJECT:  Request for Support in Evaluating Meteorological Data
Bases

FROM:     William G. Laxton, Director
          Technical Support Division (MD-14)

TO:       William B. Hathaway, Director
          Air, Pesticides and Toxics Division, Region VI (6T)


     This memorandum responds to your questions regarding the
identification and treatment of low mixing height situations. 
The response reflects discussions that have taken place between
our staffs.

     First, in regard to the attachment to your request, we agree
that there appear to be some situations in the data base where
mixing heights are low and perhaps should be questioned as being
realistic.  However, the impact of these hours of suspiciously
low mixing heights on design concentrations depends on whether
one is applying the EPA models or the Texas models.  For the EPA
models, plumes that are above the mixing height are treated as
having no effect on ground level concentrations.  However, the
Texas models appear to contain a plume penetration factor which
may result in plumes having some contribution to ground level
concentrations in such situations.  Also, the EPA models assume
unlimited mixing for stable hours, whereas the Texas models use
the hourly mixing heights with a penetration factor of 1. 
Because of these differences, we believe that the significance of
low mixing height values in your Region, when using EPA
regulatory models, may not be as great as indicated by the
examples you provided.

     In regard to your request to make modeling runs for several
stack scenarios using the 1982 Houston meteorological data base,
we do not have the resources for the remainder of FY-89 to
perform these analyses.  Perhaps the Region or the State could
make these runs.  Given the discussion above, we are not sure how
one would interpret the results of this modeling when using the
Texas criteria for the EPA models.

     In regard to your question about whether very low mixing
heights actually occur in nature, we believe that it is possible
to experience near zero mixing heights (both urban and rural). 
However, the likelihood of such very low mixing heights varies
greatly with season, climatology, and the size of the urban area. 
In many cases this is a transitional phenomena of short duration. 
There are other situations where low mixing heights have been
shown to be a significant factor.  For example, we are aware from
some PM10 studies that in interior valleys of the West in the
winter season, extremely low mixing heights are a reality and are
an important phenomena contributing to high PM10 levels associated
with ground level and fireplace emissions.  We do not know
whether such low mixing heights could realistically occur in
major urban areas along the Gulf Coast.  In regard to information
contained in the literature on the phenomena, we are not aware of
any specific studies on the subject.  Your staff may wish to
discuss the matter with John Irwin who is working on a study that
compares estimated and observed mixing heights.

      In regard to your question about whether a policy change is
needed, because of the considerations discussed above, we are not
inclined to change our present policy of "case-by-case"
determination.  Also, we are  unaware of other Regions who have
not been able to operate under this policy.  Our staffs have
discussed this aspect and we believe it might be possible for you
to develop a Regional policy or perhaps a policy that applies to
the Gulf Coast climatology.  If your staff decides to develop
such a policy, the Model Clearinghouse would be willing to review
it from the standpoint of technical defensibility.  We are
sympathetic to your desire to have more generic criteria to
identify unacceptable mixing heights, in the interest of
consistency.

     Finally, we believe that it is equally important to define
how periods of suspiciously low mixing heights will be treated in
modeling, rather than to just identify such periods.  We had been
under the impression that Texas treated such periods as missing
data for which no concentration estimates were made.  However, it
is now our understanding that Texas does indeed interpolate
between valid data points or otherwise substitute more realistic
values for the period.  This is in accordance with our policy,
and we support this approach; however, we urge that the
procedures used be adequately documented.

     If you have any questions, please contact me.

cc:  D. deRoeck (MD-15)
     D. Grano (MD-15)
     J. Irwin (MD-80)