United States
                   Environmental Protection
                   Agency
Atmospheric Sciences
Research Laboratory
Research Triangle Park NC 27711
                   Research and Development
EPA/600/S3-88/011  May ^ 988
&ERA         Project Summary
                   Parametric Methodologies  of
                   Cloud  Vertical Transport for
                   Acid Deposition  Models
                   F. M. Vukovich and R. C. Haws
                     A  CUmulus VENTing (CUVENT)
                    cloud module has been develbped that
                    calculates the vertical flux of  mass
                    between the boundary layer and the
                    cloud layer by an ensemble of nonpre-
                    cipitating cumulus clouds. This model
                    has been designed to be integrated into
                    the Regional Acid Deposition  Model
                    (RADM) to establish the effect of cloud
                    venting in that model. In the first phase
                    of this project, using data obtained
                    during the VENTEX field piogram. a
                    parameterization scheme was  devel-
                    oped for the cloud model so that it may
                    be incorporated  directly  into the
                    RADM. This parameterization scheme
                    uses basic  meteorological data to
                    predict the convective cloud amount at
                    cloud base and cloud distribution
                    parameters.  In the second j phase, a
                    number of improvements and changes
                    were made to the CUVENT algorithm
                    so that it may interface with the RADM
                    by being integrated into the Regional
                    Scavenging  Module  (RSM).  The
                    changes included the incorporation of
                    sidewall detrainment and the develop-
                    ment of limiting conditions for the
                    existence of nonprecipitating cumulus
                    clouds relative to certain seasons, of
                    a realistic cloud liquid water profile, and
                    of a simplified model to estimate cloud
                    base cloud amount. In order to meet
                    the requirements of the RSM, CUVENT
                    provided information  about a  single
                    cloud, the ' 'processor cloud," that
                    represented the ensemble of nonpre-
                    cipitating cumulus clouds. In the final
                    phase, a "table look-up" version of
                    CUVENT was developed  in order to
                    significantly reduce the amount of the
computer execution time. The new and
improved version of CUVENT that was
developed in  the second  phase was
used to establish the tables of parame-
ters needed for the processor cloud.
The tables were  developed for five
characteristic atmospheres and ten
cloud amounts.
  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
  The primary purpose of this research
project was to develop a  cloud  model
(CUVENT) that calculated the vertical flux
of mass out of the planetary boundary
layer into the cloud layer due  to an
ensemble of  nonprecipitating, subgrid
scale, air mass  convective clouds.
CUVENT, by design, is a submodule to
the  Regional  Acid Deposition Model
(RADM), which is being developed by the
National  Center  for Atmospheric
Research/State University of New York,
and will serve to establish the effects of
cloud venting for the RADM.

Approach
  CUVENT was established in  three
separate phases. These are summarized
below:

a.  Phase 1:
  The first phase consisted of two major
steps:

-------
 1.  the selection and modification of a
    model to calculate the vertical trans-
    port of mass due to an ensemble of
    nonprecipitating, subgrid scale, air
    mass convective clouds; and

 2.  development of  a closure procedure
    for that cloud model that is consist-
    ent with the RADM.

  The cloud model that was selected for
CUVENT determines acceptable cloud
classes that are defined by entrainment
rates which are functions of atmospheric
state. This model assigns the  entrain-
ment rate to a particular cloud size, which
is represented by cloud depth and by a
fractional  cloud amount.  The  greatest
entrainment  rate is assigned to those
clouds having the least depth,  and the
smallest entrainment  rate, to clouds
having the greatest  depth. The percent
of area covered by  each cloud class is
also determined. G iven these parameters
and the primary forcing functions, heat
and  moisture convergence, the model
computes  the vertical flux of mass due
to the  ensemble of  nonprecipitating
cumulus clouds.
  This model,  called  CUVENTI,  was
originally developed  to be applied using
observed data.  Radiosonde data were
used to define the  moist static energy
distribution, and satellite data were used
to determine  the   cloud  parameters.
However, RADM operates in a predictive
mode and is self-consistent. Incorpora-
tion of observed  data will produce
unacceptable errors. Since  CUVENTI is
to be  integrated into the RADM, it also
must be self-consistent in the sense that
it can only use data  provided to it by the
RADM. Therefore, a  procedure had to be
developed to adapt  the cloud model to
the predictive  mode;  i.e., the closure
procedure, RADM provides data through
which the moist static energy distribution
can be calculated. Therefore, the basic
aspects of the closure procedure con-
sisted of three major parts:

 1.  A test for moist convergence;

 2.  The estimation  of the  cloud base
    convective cloud amount;

 3.  The development of the convective
    cloud amount as a function of cloud
    classes.

The test to  determine whether moist
convection existed at a particular time
consisted of two parts:
1.  A test to determine whether the
    forcing function, the surface  heat
    flux, was consistent with convection;
    and
2.  A test to determine whether mois-
    ture conditions were consistent with
    moist convection.

  A fundamental part of this test cen-
tered around the relative location of the
entrainment zone and the lifting conden-
sation (LCL) zone.
  If the  atmosphere supports moist
convection based on tests discussed in
the above paragraph, then the convective
cloud amount was determined. A statis-
tical model was developed  which  esti-
mates the convective  cloud amount at
cloud  base using basic  meteorological
parameters. The data that were used to
develop  the statistical  model were
obtained  during the National Acid Pre-
cipitation Assessment Program (NAPAP)
venting experiment (VENTEX) that took
place  near Lexington,  Kentucky, in the
summer  of  1984.  Surface, upper air,
sodar, laser, and aircraft  measurements
were  made  to  characterize the lower
atmosphere during  current cumulus
convection  and to characterize  the
cumulus  clouds in terms of the cloud
amount, the cloud depth, and the cloud
width. The measurement program was
a joint effort by Battelle Pacific Northwest
Laboratories, Argonne National Labora-
tory, and the Research Triangle Institute.
  The statistical model  for the cloud
amount consisted of terms that charac-
terized the major forcing function for air
mass convective clouds: the surface heat
flux and the moisture parameters in the
boundary layer. The major  parameters
were the dew-point depression at the
surface and at the top of the boundary
layer, the surface heat flux and the time
rate of change of the surface heat flux,
the height of the lifting condensation
level,  and the height of  the top of the
boundary layer. One hundred and  five
separate comparisons were  used to
construct the model; the resulting corre-
lation for these comparisons was 0.81.
  As  previously discussed, the cloud
classes were defined by an entrainment
rate, as well as the cloud depth. In order
to establish the cloud amount for a given
cloud class, it was necessary to assume
that the cloud depths are continuous over
the spectrum of cloud classes, that the
cloud  depth  is  a function of the cloud
width for nonprecipitating convective
clouds. Utilization of these assumptions
permitted development of a mathemat-
ical expression that established the cloua
amount as a function of the cloud depth.
The  system of  equations that defined
thermodynamics and  dynamics of the
cloud processes and defined the parame-
ters necessary to solve the cloud model
produced  a self-consistent cloud  flux
model  (CUVENT) which defined  the
vertical mass flux due to an ensemble
of nonprecipitating air mass convective
clouds. The establishment of CUVENT in
this matter made it consistent with the
predictive mode of the  RADM.

b.  Phase 2:
  In the second phase of this research
project, a number of improvements and
changes were made to CUVENT. These
included:

1.  the  incorporation  of  sidewall
    detrainment;
2.  the development  of  a  test  which
    limits conditions for the  existence of
    an ensemble of nonprecipitating air
    mass clouds to certain seasons;
3.  the development of a more realistic
    cloud  liquid  water   profile  for
    CUVENT; and

4.  the development  of a simplified
    model to estimate cloud amount at
    cloud base.

  In CUVENTI, the first generation  ver-
sion of CUVENT, detrainment took place
at the cloud top  only.  Under  these
conditions, the maximum predicted
vertical velocity in the cloud was found
near cloud top and the vertical velocity
went rapidly to zero at cloud top. Most
observations of the  vertical  velocity
distribution in  cumulus clouds  have
indicated  that  the  maximum vertical
velocity is generally found midlevel in the
cloud. Such vertical velocity distribution
can be obtained by incorporating sidewall
detrainment in the cloud model.
  Cloud  liquid water  is  needed in
CUVENT in order to calculate the updraft
moist static energy, and it  is essential
in the solution for each cloud  class. In
this  model, the vertical distribution of
cloud liquid water was provided as a data
set that was derived from the literature.
  The initial statistical  model which was
used to estimate  the  convective cloud
amount at cloud base for CUVENTI used
two terms which might have  complicated
the  integration of  CUVENT into  the
RADM. One of these terms was the time
rate of change  of surface heat flux. In
order to calculate  this  term,  it  was

-------
required that the surface heat flux from
the previous hour is saved in memory
at each RADM grid point. Furthermore,
the precision and accuracy needed for
the term that involved the difference in
height between the top of the entrain-
ment zone and the bottom of the lifting
condensation zone was questionable
using  RAOM meteorological  data.  In
order to  remove the  complication  of
requiring a memory array for the surface
heat flux and in dealing with the ques-
tionable accuracy of the term involving
the height difference, it was decided that
both these terms should be removed from
the model and a new model be developed
whose terms are physically  consistent
with fair weather cumulus, but are more
practical and have no need for storage
capabilities. Such a model was developed
and the terms in the model  were the
dewpoint  depression at the top of the
boundary layer, the surface heat flux, the
height of the lifting condensation level,
and the free convection scaling velocity.
This model provided a correlation coef-
ficient with observations of 0.79.
  The air mass cumulus clouds which
are accounted for by  CUVENT have  a
definite warm  season preference. This
preference was used to limit utilization
of CUVENT. In order to develop this warm
season  preference and  to  determine
restriction limits,  a frequency analysis
was  established  which  matched  the
existence  of nonprecipitating  cumulus
clouds with the surface dry bulb temper-
ature and the surface dewpoint temper-
ature.  Based on  this  analysis, it was
decided to use the 80 percentile from the
cumulative  probability  distribution  to
define the restriction limit for  cumulus
clouds. The 80 percentile value provided
a surface temperature of 10 degrees and
a surface  dewpoint temperature of 0°C
(i.e., the restriction limit states that there
is an 80 percent probability that nonpre-
cipitating cumulus clouds will exist if the
surface temperature isgreaterthan 10°C
and the surface dew point is greater than
0°C).
  CUVENTI was designed to estimate the
vertical  mass flux by  nonprecipitating
clouds. This configuration for CUVENTI
was developed for its integration with the
RADM.  Subsequently,  CUVENT  was
assigned  to be  a subroutine to the
Regional Scavenging Model  (RSM).  In
order to meet the  requirements of the
RSM, CUVENT had to  provide informa-
tion about a single cloud that could be
used to represent the  ensemble  of
subgrid  scale nonprecipitating clouds.
This single cloud was hereafter referred
to as  the "processor cloud." CUVENTI
was modified to develop the information
about the processor cloud that is required
by the RSM. That information  included
the updraft vertical velocity distribution,
the updraft temperature distribution, the
updraft specific humidity distribution, the
cloud  amount  distribution,  the  cloud
liquid water distribution  in the updraft,
the cloud liquid water  distribution  in
areas  outside  the updraft, and the
precent area covered by updraft.
  After all changes and improvements
were  made, the new  version of  the
CUVENT  algorithm was  called  CUVEN-
TIA. CUVENTIA  underwent various sen-
sitivity analyses to examine the behavior
of algorithm as the vertical distribution
of temperature and dewpoint (moist
static  energy) change from one atmos-
pheric state to another. Various temper-
ature  and dewpoint profiles were used
and results were presented in terms of
the vertical velocity distribution in  the
processor cloud.

c.   Phase 3:
  In the  final  phase of this research
project, a "Table Look-Up"  version  of
CUVENT, CUVENTIIA, was  developed in
order to reduce  significantly the compu-
ter execution  time of the  algorithm.
CUVENTJftJwas used to  develop the
tables of  significant parameters for the
processor cloud. The tables were devel-
oped for five characteristic atmospheric
categories and ten cloud amounts. The
distinction between atmospheric cate-
gories was based on the lapse  rate and
moisture  in the cloud  layer. Fifteen
atmospheric states that were character-
ized by their temperature and dewpoint
profiles were examined. The reduction
from fifteen to five categories was based
on the change in cloud parameters for
the processor cloud across the  range of
atmospheric states defined by the cate-
gories. These within-category variations
were small for the five chosen categories.
However, compromises had to be made
because there was a restriction on the
acceptable storage requirement that
CUVENT could demand from the RADM
system. These two factors limited the
number of atmospheric categories to five.

-------
     F. M. Vukovich and R. C. Haws are with Research Triangle Institute, Research
       Triangle Park, NC 27709.
     Jason K. S. Ching is the EPA Project Officer (see below).
     The complete report, entitled Parametric Methodologies  of Cloud Vertical
       Transport for Acid Deposition Models," (Order No. PB88-191 374/AS; Cost:
       $32.95, 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 Project Officer can be contacted at:
             Atmospheric Sciences Research Laboratory
             U.S. Environmental Protection Agency
             Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
                                                                                                     •'-c ;--.cp; - n   A\ f
                                                                                                         • I ~U  x b
Official Business
Penalty for Private Use $300

EPA/600/S3-88/011
   0000329
                                    ASfHCt
                                                                                    GOVERNMENT PRINTING OFFICE: 1988-548-013/87

-------