United States
                   Environmental Protection
                   Agency
 Atmospheric Sciences
 Research Laboratory
 Research Triangle Park NC 27711
                   Research and Development
 EPA/600/S3-87/026  Dec. 1987
SERft         Project  Summary

                   Contributions  of  the  Fluid
                   Modeling  Facility  to  EPA's
                   Complex Terrain  Model
                   Development Program
                   William H. Snyder
                     The contributions of the EPA Fluid
                   Modeling Facility (FMF) to the Complex
                   Terrain Model Development Program
                   (CTMDP) are described. These contri-
                   butions included a wide  range of
                   laboratory studies and a limited amount
                   of numerical modeling of flow and
                   diffusion in neutral and stably stratified
                   conditions in complex terrain. The goal
                   of the CTMDP is the developmem of
                   a dispersion model valid in complex
                   terrain, with emphasis on plume impac-
                   tion on  nearby hills  during nighttime
                   stable conditions. Work at  the FMF
                   prior to the inception of the program
                   divided the basic  framework for the
                   model—the dividing-streamline con-
                   cept—and the focal point around which
                   the field  program  was designed.
                   Throughout the course of the CTMDP,
                   the FMF interacted vigorously with the
                   model developers by providing support
                   in various ways. Early work provided
                   direct support as an aid to planning the
                   details and strategies of the field
                   experiments and testing the limits of
                   applicability of the dividing-streamline
                   concept. Later work included exercises
                   of ' 'filling in the gaps" in the field data,
                   furthering the understanding  of the
                   physical  mechanisms important to
                   plume impaction in complex terrain and
                   in stably stratified flows in general,
                   testing various modeling assumptions,
                   providing data for ' 'calibration"  of
                   various  modeling parameters, and
                   testing the ability of the laboratory
                   models to simulate  full-scale condi-
                   tions.  Simultaneously, the  FMF
responded to the needs of the regula-
tory arm of EPA. the Office of Air
Quality  Planning  and  Standards
(OAQPS), by providing guidance con-
cerning expected terrain effects and by
conducting demonstration  studies.
Finally, several supplemental studies
were  conducted, broadening  and
expanding upon the specific  requests
of the model developers and the
OAQPS.
  This Project Summary was devel-
oped by EPA's Atmospheric Sciences
Research Laboratory, Research Trian-
gle Park. NC. to announce key findings
of the research project that is  fully
documented in a separate report of the
same title (see Project Report ordering
information at back).

Introduction
  In the late 1970's the Office of Air
Quality Planning and Standards (OAQPS)
of the Environmental Protection Agency
(EPA) identified a crucial need to develop
an improved mathematical model that
dealt with plume impaction from  large
sources located in mountainous terrain
under stable flow conditions. A workshop
was convened in 1979 to focus on
complex terrain modeling problems and
to develop recommendations to EPA with
respect to the design of a program of
experiments and  model development
efforts. Subsequently, the EPA outlined
a plan  to achieve the objective through
an integrated program of model devel-
opment, fluid modeling experiments and
field studies of plume-terrain interac-

-------
tions on hills of progressively increasing
size  and  complexity.  This multi-year,
multi-faceted program is known as the
Complex Terrain Model  Development
Program (CTMDP).
  The  Fluid  Modeling Facility (FMF)
interacted vigorously with various groups
participating in the CTMDP, and provided
direct support and  guidance in  many
different ways. The FMF research pro-
gram has ranged from the  development
of broad guidelines and physical  con-
cepts to specific  site studies and regu-
latory applications. The FMF has provided
laboratory data to "fill in  the gaps" in
the field data and tested the validity of
convenient modeling assumptions.
  The complete report summarizes the
contributions, both direct and indirect, of
the FMF to the CTMDP. The discussion
provides a historical perspective and a
comprehensive list of FMF's accomplish-
ments  with respect  to furthering the
physical understanding   of flow and
diffusion in complex  terrain. Over 65
publications have been generated  from
work  conducted within  the FMF  on
complex-terrain research. Only a few of
the most  important  publications are
highlighted herein.


Background
  The major facilities of the FMF consist
of a meteorological wind tunnel and  a
stratified water  channel/towing tank.
The wind tunnel has a  test  section 3.7m
wide, 2.1m high, and  18.3m long, and
a speed range of  0.5 to  10m/s. The
towing tank is 2.4m wide, 1.2m deep,
and 25m  long.  It  is  density-stratified
using layered mixtures of salt water.
  Research work conducted at the FMF
prior to the inception of the CTMDP had
a strong influence  on  the  directions to
be taken in the field work and  on the
type of model (i.e., physical concepts) to
be developed. The stratified towing tank
was commissioned in  1976 and rather
fundamental studies were begun imme-
diately on the structure of stably stratified
flow  over  idealized three-dimensional
hills and on diffusion from a point source
within  a   stably  stratified  field  of
turbulence.
  The first published reports on this work
described  the flow structure observed
over a bell-shaped hill under neutral and
stably stratified conditions.  Earlier theo-
retical  work, model experiments, and
observations all indicated that, when the
stratification  is strong enough, the air
flows in approximately horizontal planes
around the topography. Up to that time.
however,  there had  been  little firm
laboratory or field data as to how strong
the stratification must be for any given
streamline starting below the hill top to
pass round the side rather than over the
top of the hill.
  Hunt and Snyder (1980) showed evi-
dence for a dividing streamline of height
HB such that streamlines below H, would
impinge on the hill surface and follow
the surface around the sides, whereas
streamlines above H, would go over the
top. They suggested the simple forumula



as the criterion to determine whether a
plume embedded in the flow approaching
the hill would impact on the surface or
surmount  the top, for 0
-------
parcel at some height upstream possess
sufficient kinetic energy to overcome the
potential energy required  to  lift itself
through  the  density  gradient from  its
upstream elevation to the hill top?" The
left-hand side may be interpreted as the
kinetic energy of the parcel far upstream
at elevation Hs, and the right-hand side
as the potential energy  gained by  the
parcel in being lifted from the dividing-
streamline height Ha to  the hill  top h
through the density gradient dp/dz. This
integral formula was  persumably appli-
cable to a fluid with any shape of stable
density profile and, presumably, with any
shape of approach-flow velocity profile.
In practice, it must be solved iteratively,
because the  unknown Ha is the  lower
limit  of  integration;  the formula  can
easily be reduced to the simpler formulae
(1) and  (2) by using  the  boundary con-
ditions applicable to those special cases.
The third study thus  attempted to verify
this  integral formula under  density
profiles similar to those expected at CCB.
A typical nighttime  temperature profile
in the Snake River  Basin (site of CCB)
was found to consist of a strong, surface-
based inversion of depth 50 to 100m and
a weaker inversion  above extending to
several hill heights. Hence, the stratified
towing  tank  was filled  with  a  strong
density gradient near the surface  and a
weaker  gradient above. A vertical rake
of 3 tubes  was positioned well upwind
of the hill (a model of CCB), and neutrally
buoyant dye was emitted from each tube.
For each tow, a particular stack height
(center tube) was chosen and the general
formula  was integrated numerically
using the  measured  density profile to
predict the towing speed required such
that the  center  streamer would rise to
the elevation of the saddle point of CCB,
i.e.,  the  minimum height of the draw
between the two peaks.  If the formula
were correct, then, the lower streamer
should go around the side of the hill, the
upper streamer  should go over the top,
and  the center  one  should split. The
height of the break-point between  the
two gradients was then reduced and the
process repeated. In all,  12 tows were
made, varying the height of the break-
point or  the  dividing-streamline height
(release height) each time.
  Figure 1  shows a  side view  of  the
impinging streamers during a typical tow,
i.e., the  upper streamer  going through
the draw, the lower streamer going round
the side, and the middle one splitting.
The density profiles were integrated in
accordance with Equation (3) to find the
 Figure 1.    Oblique view of impinging streamers on CCB. Middle dye streamer is released
            on the dividing-streamline height; others at ±1 cm (±6 m full scale).
dividing-streamline heights as functions
of the towing  speed. The  agreement
between the predictions and  observa-
tions was excellent. The results of this
set of experiments provided  confidence
in the validity  of the general  integral
formula for predicting the height of the
dividing streamline for a wide range of
shapes of stable density profiles.
  Subsequent to the field  study, one
particular hour  of the field data at CCB
was selected for simulation in the towing
tank. That hour was 0500 to 0600, 24
October 1980 (Case 206), which may be
characterized as very stable, i.e.,  light
winds  and strong  stable temperature
gradients. Measurements made during
the towing-tank experiments  included
ground-level  concentrations  under var-
ious stabilities and  wind  directions,
vertical distributions of concentration at
selected  points, plume distributions in
the absence of  the hill, and visual
observations  of plume characteristics
and trajectories.
  This series of tows showed that the
surf ace-concentration distributions were
extremely sensitive  to changes in wind
direction. For example, Figure 2 shows
that the distribution  shifted from  the
north side of the hill  to the south side
with a shift of only 5° in wind direction.
Comparisons of individual distributions
with field results showed much larger
maximum surface concentrations and
much  narrower  distributions in the
model results. To account for the large
variability in the winds measured during
the hour, a matrix of 18 tows (three wind
directions x six wind  speeds) was con-
ducted,  and  the concentration patterns
were superimposed. The resultant super-
imposed model concentrations compared
very favorably with field measurements.
The largest model concentrations were
within a factor of two of the highest field
values, and 70% of the model concen-
trations were within a factor of two of
the observed field values.
  Numerous  other  studies  were  con-
ducted to test the validity and limits of
applicability of the  dividing-streamline
concept  for  example, examining  the
effects of shear in the  approach-flow
velocity profile, of the crosswind aspect
ratio of the hill, of the hill slope, and the
effect of the wind angle on a long ridge.
These results were published separately
as parts  of papers on studies done for
a variety of different purposes, but the
specific aspects dealing with the validity
and  applicability  of  the  dividing-
streamline concept  were extracted and
published collectively by Snyder et al
(1985).
  In  response to  a request from  the
model developers, a series of measure-
ments was made of plume characteris-
tics in  flat terrain  and  over a  three-
dimensional hill immersed in the simu-
lated neutral atmosphere boundary layer
of  the  meteorological  wind tunnel.
Effluent  was released at a  number of
elevations, upwind distances, and posi-
tions laterally offset from the centerplane
determined by the wind direction and the
center of the hill. Sufficient concentra-
tion measurements were made to enable
the construction of plume cross sections
at  the downwind  position  of the hill
center and, in a few cases, at the upwind
base of the hill. These data were analyzed
to provide the desired  information  on
horizontal and vertical plume deflections
and deformations  effected by the  hill.
One of the more dramatic examples is
shown  in  Figure 3.  In  this case,  the
source was on the centerplane at ground
level, 6 hill heights upwind  of the hill
center (the skirt of  the hill extended to

-------
                                                        600 M
            600 M
            Scale
                                                            Source
Figure 2.
     200 M


Concentration distributions measured during individual tows of CCB with
h = 0.31 andH0/h = 0.38; wind direction:	117°.	122°.
N
      2 i-
      7  -
                                                               05
Figure 3.   Plume cross sections measured in presence (	) and in absence (	) of
           axisymmetric CCB model at x = 0 (hill center). H,/h = 0, xs/h = -6. y,/h = 0.
5h). Plume cross sections measured at
the position of the center of the hill, both
in the presence and in the absence of
the hill  are shown. The hill effected a
91% increase in the lateral plume width.
In this  case, the  maximum surface
concentration (at the same downwind
distance) was decreased  by a factor of
2 but, of course, the area of coverage
by large concentrations  was  greatly
increased. Detailed data  reports were
provided to the model developers imme-
diately,  and the  results were  published
by Snyder and Lawson (1986).
  As  a  result  of  these  and  similar
measurements, refinements were made
to CTDM. Specifically, the calculation
procedures were modified to  utilize  the
strain inferred  or  measured over  the
crests of  two-   and three-dimensional
hills in the wind tunnel, i.e., theT-factors
in the model were adjusted in accordance
with  wind-tunnel data. Substantial
improvements in the CTDM predictions
of terrain amplification   factors were
obtained, as described by Strimaitis and
Snyder (1986).
  One of the important overall goals in
this effort was to ascertain what circum-
stances lead to the largest ground-level
concentrations(glc's), i.e., are larger glc's
expected when the plume from  an
upwind  source  impinges on the hill or
when the source is downwind of that hill
such  that the  plume is  caught in a
recirculation region and downwashed to
the surface? Which are likely to  lead to
larger glc's,  two-dimensional or three-
dimensional hills?  Stable  conditions or
neutral  conditions?  In  each of  these
circumstances,  what orders  of  magni-
tude of  surface  concentrations  may be
expected?
  A simple method used to intercompare
effects of terrain on the maximum  glc
and to determine worst-case conditions
is through the the terrain amplification
factor, A, which is defined as the ratio
of the maximum ground-level concentra-
tion occurring  in the presence  of  the
terrain feature, Xm«, to the maximum that
would occur from  the  same  source
located  in flat terrain, x™, i.e.,  X*=Xmx/
X™. This  definition  is useful only  for
elevated sources, of course, because for
ground-level  sources, the maximum
surface concentration  occurs   at  the
source itself.
  A wide range of neutral-flow wind-
tunnel studies was conducted at the FMF
on diffusion from sources  located in  the
vicinities of two- and three-dimensional
hills. Table 1 lists approximate values of

-------
Table 1.    Summary of Terrain Amplification Factors for Sources in the Vicinity of Hills in
           Neutral Flow
Source
Location
    Hill Type
Downwind
Downwind
Upwind

Upwind

Top
Two -Dimensional
                                 Three-Dimensional
Three-Dimensional
Two-Dimensional
Two-Dimensional
                                                                        10-15
                                        5-6
                                                                         2-4
                      1-3
                                       0.5-1
maximum terrain  amplification  factors
that were found in the various situations.
From the standpoint of a  fixed stack
height,  the worst location for  a  source
appears to be just downwind of a two-
dimensional  ridge.  Downwind  sources
generally result  in larger glc's because
of the excess turbulence generated by
the  hills  and because  the  effluent is
generally emitted into a low speed region
where the streamlines  are  descending
toward  the surface.  Maximum A's  are
considerably  larger  than those  down-
wind of three-dimensional hills. Also, the
sizes of the recirculating cavity regions
downwind of three-dimensional hills are
generally  much smaller than  those
downwind of two-dimensional  ridges.
With  regard to upwind sources,  terrain
amplification factors are larger for three-
dimensional hills because, in such flows,
streamlines can  impinge on  the surface
and/or approach the surface more
closely than in two-dimensional flows.
  The maximum terrain amplification
factors as listed in Table 1 are useful only
for scoping a particular problem or for
finding the worst possible situation. They
do not provide practical estimates for use
by, say, an air pollution meteorologist in
determining the  maximum glc  resulting
from  a particular  power  plant  or  for
determining the best location for that
plant. For that purpose, the concept of
a "window" of excess concentrations is
more useful. For  any given plant location
(say, upwind  of a hill), there  is a  limited
range of stack heights  A/s for which a
significant amplification of  the glc will
occur. (For sake of argument, we will
here define significant as a factor of 2.)
This amplification can occur only if  the
position of the maximum glc lies on or
near the hill  surface. For small H,,  Xmx
will occur upwind of the hill and thus
be little influenced by the hill, so that
A (=x™/Xm*) will approach  unity. If Ha
is too large (for example, H£>h, the  hill
height), Xm* will  lie well beyond the  hill
        and A will again approach unity. In either
        case, there is little  amplification.  A
        "window" of intermediate stack heights
        and locations exists, however, where A's
        will  significantly  exceed unity.  These
        "windows" of critical Ha  values have
        been measured by Lawson and Snyder
        (1985)  for two typical hill shapes  that
        might be found in the real world,  one
        axisymmetric,   the   other   two-
        dimensional. The  results are shown  in
        Figure 4. The 1.4-window, for example,
        extends to about  14/7  upstream,  10/7
        downstream, and as high as 1.8/7 in the
        vertical for the axisymmetric hill. For the
        two-dimensional hill,  this 1.4-window
        extends about  8/7 upstream, 15/7  down-
        stream,  and  as  high as  2.2/7 in  the
        vertical.
          Such  contour  maps as provided  in
Figure 4  can be  very  useful  for the
practioner. Once an acceptable terrain
amplification factor (or "excess concen-
tration") is decided upon, it is a simple
matter to  trace the window  on the
contour map to determine the area (plant
location  and/or  stack height) to be
avoided. Conversely, from such maps,
the likely  maximum  glc for a potential
site and stack height can be estimated.
The use of terrain  amplification factors
simplifies  the application of these  data
to full-scale  situations. The expected
maximum  glc in flat terrain is calculated
(from  mathematical models or standard
curves), then the concentration in the
presence of the hill is simply the  product
of this quantity and the TAF.

Summary
  The EPA Fluid Modeling Facility has
conducted a  wide  range of laboratory
studies and a limited amount of  numer-
ical modeling of flow and diffusion  in
association with the CTMDP  The  goal
of the CTMDP is the development  of a
dispersion model valid in complex ter-
rain, with emphasis on plume impaction
on nearby hills during nighttime stable
conditions. Work at the FMF prior to the
inception  of the program provided the
basic  framework  for the model—the
dividing-streamline  concept—and the
focal point around  which to design the
field program.
         •c;..
           -15
          -15
        Figure 4.
                      -10
Contours of constant terrain amplification factors over (a) axisymmetric hill and
(b) two-dimensional ridge. Note that vertical scale is exaggerated by a factor of 3.

-------
  Throughout the course of the CTMDP,
the FMF interacted vigorously with the
model developers by providing support in
various ways. Early work provided direct
support  in planning  the details and
strategies of the field experiments and
solidifying and  testing  the  limits  of
applicability of the  dividing-streamline
concept.  Later work included  excerises
of "filling in the gaps" in the field data,
furthering the  understanding  of the
typical mechanisms important to plume
impaction  in complex terrain  and  in
stably stratified  flows in general, and
testing  the ability of the laboratory
models to simulate full-scale field con-
ditions. And, as the needs arose, the FMF
tested  various modeling assumptions,
concepts, and hypotheses and provided
data for "calibration" of various param-
eters within the CTDM model.
  Simultaneously, the FMF responded to
the needs of the  regulatory arm  of EPA,
OAQPS, by providing guidance concern-
ing  expected  terrain effects and  by
providing a demonstration study—an
example  for industries to  follow  in
conducting good-engineering-practice
stack height determinations in complex
terrain. Also, a broad range of  supple-
mental studies was conducted, expand-
ing  and  enlarging  upon the specific
requests of the OAQPS  and the CTDM
model developers to provide information
of general use to the  scientific  and  air
pollution modeling  communities. Many
of the data sets generated in the course
of this program  have  been provided to
and  used by various groups (nationally
and internationally) in  the development,
testing and evaluation of complex terrain
dispersion models.
  The  most significant  contributions
included  (1) the  conceptual framework
for the  mathematical model  (i.e., the
division  of the flow-field  into two
regimes, a lower layer below the
dividing-streamline height which flows
in essentially horizontal planes around
the hill,  and an  upper layer  above the
dividing-streamline height  which is
treated as modified potential flow over
a cut-off hill) and the detailed experimen-
tal validation and establishment of limits
of applicability  of  these concepts,  (2)
verification of the integral formula for the
height of the dividing-streamline—this
allowed  computations of the dividing-
streamline height  under   arbitrary
approach-flow  conditions,  including
shear in  the approaching wind-speed
profile and nonlinear  temperature gra-
dients, (3) demonstration of the extreme
sensitivity of surface  concentration
patterns to wind direction under strongly
stratified conditions, (4)  measurements
of plume deflections and deformations
over hills in neutral flow—these permit-
ted adjustment of the T-factors in CTDM
and resulted in substantial  improve-
ments in the CTDM  predictions, and (5)
the introduction of the concept  of "win-
dows of excess concentration"  and
measurements of terrain amplification
factors—these  provided  simple  and
practical  methods  for  estimation  and
intercomparison of effects of terrain and
source locations on maximum  ground-
level concentrations that  may result from
sources placed in the vicinities of hills.
  Only the highlights  of the FMF con-
tributions to the CTMDP are contained
in the present summary. The complete
report provides much more detail and a
comprehensive list of  over 65  publica-
tions generated from the work conducted
at the FMF on complex-terrain research.

References
Hunt, J. C. Ft., Puttock, J. S., and Snyder,
  W. H.  1979. Turbulent Diffusion from
  a Point Source in Stratified and Neutral
  Flows  Around a  Three-Dimensional
  Hill: Part I. Diffusion Equation Analysis.
  Atmos. Envir., 13:1227-1239.

Hunt, J. C. R. and Snyder, W. H.  1980.
  Experiments on Stably and Neutrally
  Stratifed Flow Over a Model Three-
  Dimensional  Hill. J. Fluid Mech.,
  96:671-704.
Lawson,  R.  E., Jr.,  and Snyder, W. H.
  1985. Stack Heights and Locations in
  Complex Terrain.  Preprints Vol: 7th
  Symp. Turb. Diff, Nov. 12-15, Boulder,
  CO, 223-226. Amer. Meteorol. Soc.,
  Boston, MA.

Snyder, W. H.,  Britter, R. E., and Hunt,
  J. C. R.  1980. A Fluid Modeling Study
  of  the  Flow Structure and  Plume
  Impingement on a Three-Dimensional
  Hill in Stably Stratified Flow. Proc. Fifth
  Int. Conf. on Wind  Engr. (J. E. Cermak,
  ed.),  1:319-329. Pergamon Press, NY,
  NY.

Snyder, W.  H.  and  Lawson,  R.  E.,  Jr.
  1986.  Laboratory Observations of
  Plume  Deformations  in Neutral Flow
  Over a  Three-Dimensional  Hill. Pre-
  print Vol: AMS 5th Jt. Conf. Appl. Air
  Poll. Meteorol. with  APCA, Nov.,
  Chapel Hill, NC. Amer. Meteorol. Soc.,
  Boston, MA.

Snyder, W. H., Thompson, R. S., Eskridge,
  R.  E., Lawson, R.  E., Jr., Castro, I. P.,
  Lee, J.  T., Hunt, J. C. R., and Ogawa,
  Y.  1985. The Structure of Strongly
  Stratified Flow Over Hills:  Dividing-
  Streamline Concept. J. Fluid Mech.,
  152:249-288.

Strimaitis, D. G. and Snyder, W. H. 1986.
  An Evaluation of the Complex Terrain
  Dispersion Model  Against Laboratory
  Observations: Neutral Flow  Over 2-D
  and 3-D Hills. Preprint Vol:  AMS 5th
  Jt. Conf. Appl. Air  Poll. Meteorol. with
  APCA,  Nov.,  Chapel Hill, NC,  Amer.
  Meteorol. Soc., Boston, MA.
   The EPA author William H.  Snyder (also the EPA Project Officer, see below)
     is with the Atmospheric Sciences Research Laboratory, Research Triangle
     Park, NC 27711.
   The complete report, entitled "Contributions of the Fluid Modeling Facility to
     EPA's Complex  Terrain Model Development Program," (Order No. PB 87-
     227 682/AS; Cost: $13.95, subject to change) will be available only from:
          National Technical Information Service
          5285 Port Royal Road
          Springfield, VA 22161
          Telephone: 703-487-4650
   The EPA Officer can be contacted at:
          Atmospheric Sciences Research Laboratory
           U.S. Environmental Protection Agency
           Research Triangle Park, NC 27711

-------
United States                    Center for Environmental Research
Environmental Protection            Information
Agency                        Cincinnati OH 45268
Official Business
Penalty for Private Use $300

EPA/600/S3-87/026
      0000329   PS
      U $  EHVIR  PROTECTION  AGENCY
       REGION  5 LIBRARf
       230  S OSARB0RK STREET
       CHICAGO             IL   60604

-------