United States
                     Environmental Protection
                     Agency
Environmental Sciences
Research Laboratory
Research Triangle Park NC 27711
                     Research and Development
EPA-600/S3-84-057  May 1984
&EPA          Project  Summary
                     Development of the
                     MESOPUFF II  Dispersion  Model

                     J. S. Scire, F. W. Lurmann, A. Bass, and S. R. Hanna
                       The development of the MESOPUFF
                     II regional-scale air quality simulation
                     model is described. MESOPUFF II is a
                     Lagrangian   variable-trajectory  puff-
                     superposition model  that  has been
                     designed  to  treat  transport,
                     transformation, diffusion, and removal
                     processes of pollutants  emitted from
                     multiple point and/or area sources at
                     transport distances beyond the range of
                     conventional  straight-line   Gaussian
                     model (i.e., beyond ~ 1O-50  km).
                       The major features of this model and
                     enhancements  over its  predecessor,
                     MESOscale PUFF  (MESOPUFF),
                     include  the use  of hourly surface
                     meteorological data,  twice-daily
                     rawinsonde  data,  and  hourly
                     precipitation data; separate wind fields
                     to represent flows within and  above the
                     boundary  layer;  parameterization of
                     vertical  dispersion in terms  of micro-
                     meteorological  turbulence   variables;
                     transformation of sulfur dioxide (SO2)
                     to  sulfate (SO<)  and nitrogen oxides
                     (NOx) to nitrate (NO3~): a  resistance
                     model for  dry deposition;  time- and
                     space-varying  wet  removal;  and a
                     computationally efficient puff-
                     sampling function.  The  scientific and
                     operational bases of the  methods used
                     in  the  model  are discussed.  The
                     resultsfrom several model algorithms
                     also are compared against experimental
                     data.

                        This Project Summary was developed
                     by EPA's   Environmental  Sciences
                     Research  Laboratory, Research
                     Triangle 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
  The regional and long-range transport
and transormation of sulfur oxides and
nitrogen oxides emitted from major point
sources  have  been  of considerable
concern.There is a need for easily usable,
cost-efficient air-quality models that can
realistically treat the various physical
processes important on these scales The
MESOscale  PUFF (MESOPUFF)  model
has been extensively modified to revise
and more realistically treat the transport,
vertical  dispersion,   chemical
transformation, and dry and wet removal
processes. The new model, designated
MESOPUFF  II,  is one element  of an
integrated  modeling  package  that
includes components for preprocessing
of meteorological data (READ56, MESO-
PUFF II) and for postprocessing of pre-
dicted concentrations (MESOFILE II).

Major Model Features
  MESOPUFF II uses a  puff-superposi-
tion  approach to represent continuous
plumes. The  pollutant material in each
puff is transported independently of that
in other puffs and is also subjected to
dispersion, chemical transformation, and
removal processes. Some of the general
features of the MESOPUFF II modeling
system are as follows:

  (1) Hourly surface meteorological data,
     twice-daily rawinsonde data, and
     hourly precipitation data are read
     from magnetic tapes.

  (2) Wind fields to represent the mean
     flow  in  the boundary layer and
     above   the  boundary  layer  are
     constructed, although several
     options are given.

-------
  (3) Boundary-layer structure is treated
     in  terms  of  micrometeorological
     parameters that includethesurface
     friction  velocity,  mixing height,
     convective  velocity  scale,  and
     Monin-Obhukov length.

  (4) Space- and time-varying chemical
     transformation can  be  performed
     simultaneously for up to five pollu-
     tant species, including sulfur diox-
     ide (SO;,),  sulfate (S0,:), nitrogen
     oxides (NOx=NCH NO?), nitric acid
     (HNO3), and nitrate (N03 ).

  (5) Dry deposition is prescribed by a
     resistance model  in  the surface-
     depletion mode,  and the source-
     depletion method is optional.

  (6) Space-  and   time-varying  wet
     removal is parameterized according
     to precipitation rate and scavenging
     coefficients.

  The  meteorological   data inputs
required by the preprocessors consist of
the routine twice-daily upper air sound-
ings,  hourly  surface   meteorological
observations,  and  hourly  precipitation
measurements reported by the National
Weather Service. The preprocessor pro-
grams have been designed to read the
standard-formatted meteorological data
tapes available from the National Climatic
Center in Asheville, North Carolina.
  Wind fields for MESOPUFF II are con-
structed from the  hourly surface wind
observations,  as well as from the  twice-
daily rawinsonde wind profile data.  The
surface station network data allow better
temporal and spatial resolution than do
the  upper air  sounding  data,  which
involve much larger distances and  less
frequent measurements. The wind fields
are  constructed at  two  user-specified
levels--a  lower level  representing  the
mean boundary-layer flow and an upper
level  representing   flow  above  the
boundary layer.
  Boundary-layer structure is parameter-
ized in  terms  of  micrometeorological
variables  computed  from the surface
station  data  and  information  about
surface  characteristics  (land  use,  or
roughness lengths) provided by the user
for each grid point. The surface friction
velocity, u*. the convective velocity scale,
w*,  the Monin-Obukhov  length,  L, and
the  boundary-layer  height,  Zi,   are
computed.
  Chemical transformation rate express-
ions were developed from the results of
photochemical model simulations over  a
wide range of environmental conditions.
The rate expressions include gas-phase
N0« oxidation, gas- and aqueous-phase
components of SOj oxidation, and the
chemical equilibrium of the nitric acid,
ammonia,  and ammonia nitrate system.
The parameterized transformation rates
depend  on solar radiation, background
ozone concentration,  and atmospheric
stability. The SOj oxidation rate is empir-
ically increased at high relative humidity
to account for aqueous-phase  reactions.
In the case of  NOx, the transformation
rate also depends on the NOx concentra-
tion.
  The spatial  and temporal variations of
dry deposition are treated by a resistance
model. The pollutant flux is proportional
to the inverse of a sum of resistances to
pollutant transfer   through  the
atmosphere   to  the  surface.  The
resistances depend on the characteristics
of the  pollutant  and  the underlying
surface  and   atmospheric conditions.
MESOPUFF II contains options for the
commonly used source-depletion method
or for more realistic surface  depletion,
where pollutants are removed only from a
surface  layer  in the three-layer mode.
  Precipitation  scavenging can be the
dominant  pollutant removal mechanism
during precipitation periods. MESOPUFF
II contains a scavenging ratio formulation
for wet  removal. The  scavenging ratio
depends on the type and rate of precipita-
tion  (derived  from  hourly precipitation
measurements, if available) and the char-
acteristics of the particular pollutant.
  In   addition,  improvements  were
made in the method of summing the con-
tributions  of individual puffs to the total
concentration at a receptor location. The
model uses an integrated form  of the
puff-sampling function  that eliminates
the problem of insufficient puff overlap
commonly encountered with puff-super-
position models. This development allows
continuous plumes to be more accurately
simulated   with fewer puffs,  thereby
saving computational time and reducing
computer  storage requirements.

MESOPUFF II Modeling System
  The MESOPUFF II modeling package is
schematically illustrated in Figure 1. The
two  meteorological  preprocessor  rou-
tines  are   READ56 and MESOPAC  II.
READ56 processes the rawinsonde data,
and  MESOPAC II reads the output file
created  by READ56 and the  standard-
formatted  hourly surface meteorological
data and   precipitation  data.  A single
output file is  produced that includes all
the time- and  space-interpolated fields of
meteorological  variables  required  by
MESOPUFF II.
  All  source,  receptor, and  program-
control information is input by formatted-
card  images.  The control  parameter
inputs determine which options are used
in  the computations and what type of
output is produced.
  The  model was  run for  a  two-day
period,  to  evaluate  the  SO;  to  SOj"
transformation  mechanism  and to
qualitatively demonstrate the behavior of
some other model algorithms. The modeled
period was taken from the August 1 978
Tennessee  Plume  Study,  which   was
conducted near the Cumberland power
plant  in  northwestern  Tennessee.  The
two-day period (August 22-23) included
chemical  measurements   by   aircraft
traverses through the plume at distances
of  18 km to 160 km and represented 2 to
10 hours   of  travel.   The  predicted
transformation  rates   were  generally
close  to the observed rates derived from
the pollutant measurements, especially
for the drier, sunny period on August 22.
During this  period, SO3 oxidation  was
probably dominated  by  gas-phase
reactions, whereas on  the  23rd, when
plume-cloud interactions were observed,
the   transformation  rates  were
underpredicted (values  were two-thirds
the observed rates). The larger observed
values  are  attributed  to  greater
contributions  by  aqueous-phase
reactions.

Conclusions
  The  scientific  and  operational  ap-
proaches to modification of the model are
completely described in the report. The
results of the model runs, including a
comparison  of observed  and predicted
transformation  rates,  the  qualitative
behavior  of  plume   growth, plume
fumigation, and dry deposition (surface
depletion)  are  also   presented. A
companion report entitled "User's Guide
to  the MESOPUFF II Model and related
Processor Programs"  provides a  brief
technical description of  the methods and
a complete set of user instructions.

-------
                                                       I   READ56 Control
                                                       I  Parameter Inputs
                                                   'Twice-Daily\
                                                   Ra winsonde
                                                     Data Files
                                                     (TDF5600
                                                      F or mat I *
                                                                     READ56 Upper Air
                                                                   Preprocessor Program
           fME SOP AC II
           I  Control Parameter
           I       Inputs
        Hourly
    Meteorological
      Data Files
      (CD 144
        Format!
                                          Formatted
                                         Twice Daily
                                         Rawinsonde
                                       V Data Files
                                                                                                 No
           f  MESOPUFFII
           I  Control Parameter
           I        Inputs
                                                                              (Optional)
                                      MESOPAC II Meteorological
                                         Preprocessor Program
                                           Hourly
                                           Ozone
                                       Measurements
                 Hourly
             Meteorological
                Variables
                                    MESOPUFFII Dispersion Model
      f   MESOFILEII
      I Control Parameter
      I       Inputs
                           Concentration
                               Tables
   Predicted
Concentrations
                     MESOFILEII
                Postprocessor Program
          Direct Access
          Concentration
            Files lor
            Plotting
Concentration
    Tables
Figure 1.     MESOPUFF II modeling package.

-------
      J. S. Scire, F. W. Lurmann. A. Bass, and S. R. Hanna are with Environmental
         Research and Technology, Inc., Concord. MA O1742.
      James M. Godo witch is the EPA Project Officer (see below).
      The complete report, entitled "Development  of the MESOPUFF II Dispersion
         Model," (Order No. PB 84-183 753; Cost: $11.50. subject to change) will be
         available only from:
              National Technical Information Service
              5285 Port Hoy a I Road
              Springfield, VA 22161
              Telephone: 7O3-487-4650
      The EPA Project Officer can be contacted at:
              Environmental Sciences Research Laboratory
              U. S. Environmental Protection Agency
              Research Triangle Park. NC 27711
                                           U.S. GOVERNMENT PRINTING OFFICE: 1984 — 759-O15/7715
United States
Environmental Protection
Agency
Center (or Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300

-------