Atmospheric Sciences
                  Research Laboratory
                  Research Triangle Park NC 27711
United States
Environmental Protection
Agency
                  Research and Development
EPA/600/S3-86/028 June 1986
&EPA        Project  Summary
                  Further  Case  Studies  on the
                  Impact  of  Mesoscale
                  Convective  Systems  on
                  Regional  Ozone  and  Haze
                  Distributions
                   Walter A. Lyons and Rebecca H. Calby
                    This paper represents a continuation
                   of an earlier project to study the impact
                   of mesoscale convective precipitation
                   systems upon distributions of aerosol
                   and photochemical oxidant pollutants
                   in the planetary boundary layer (PBL).
                   In the original study, using data col-
                   lected during Persistent Elevated Pollu-
                   tion Episode/Northeast  Regional Oxi-
                   dant Study (PEPE/NEROS-80), analyses
                   of surface visibility and ozone data re-
                   vealed a dramatic  response in the
                   boundary layer pollutant patterns to
                   the passage of two very large convec-
                   tive storm systems:  a mesoscale con-
                   vective complex and an  intense squall
                   line. Regional visibilities, at times less
                   than 5 km, increased dramatically to as
                   high as 80 km over a multistate area. A
                   technique was developed to  estimate
                   the total amount of aerosol, presumed
                   to be primarily sulfate,  that  was dis-
                   placed from the PBL. Approximately
                   38 x 106 kg of sulfate was apparently
                   redistributed by the convective system.
                   The resultant clean air region was
                   termed a convective aerosol removal
                   event (CARE).
                    In this study, a search of the existing
                   PEPE/NEROS-80 data base  was ini-
                   tiated to discover additional CAREs. A
                   well-defined CARE was found in satel-
                   lite imagery off the Georgia coast on 14
                   August 1980.  Extensive  mesoanalyses
                   revealed that the clear air pocket em-
                   bedded within a large surrounding
                   PEPE originated from a mesohigh sys-
tem formed by a cluster of thunder-
storms over northern Florida and Geor-
gia the previous day. While the increase
in regional visibility and the volumetric
depletion of sulfate aerosol was of the
same magnitude as  in the mesoscale
convective complex squall line case,
the total mass of displaced aerosol was
7.7 x 10s kg, fully 50 times less. The re-
sponsible air mass thundershowers, far
smaller than the systems first studied,
clearly can result in less widespread but
still significant impacts on PBL pollu-
tion.
  In addition, the phenomenon of the
pseudo-CARE was discovered. "Clear
spots" embedded within  large quasi-
homogeneous PEPEs were found not to
be associated with specific thunder-
storms. In one case,  large-scale subsi-
dence over the Great Lakes associated
with lake breezes was hypothesized to
have induced CARE-like  clear zones
which drifted as far south as the Ohio
River. These observations have impor-
tant implications for mesoscale deposi-
tion modeling and for the interpreta-
tion of field program aerosol
measurements.
  This Project Summary was devel-
oped by EPA's Atmospheric Sciences
Research Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering infor-
mation at back).

-------
Introduction
  In the early  1970s,  meterological
satellites began providing graphic im-
ages of the long-range transport of
aerosol pollutants.  High resolution
Landsat imagery tracked  discrete
plumes from power  plants and  steel
mills for over 200 km.  There was specu-
lation that large haze areas seen over
Europe from polar orbiting  satellites
were in fact synoptic-scale air pollution
events. By 1976, the visible sensors of
geostationary satellite systems pro-
vided circumstantial evidence that large
areas of haze over the  eastern United
States were linked to  synoptic-scale
episodes of sulfate pollution. This phe-
nomenon, eventually  termed a persis-
tent elevated pollution episode (PEPE),
has  since been extensively studied.
Given the frequent recurrence of highly
polluted boundary layers extending
over multi-state areas, questions  soon
arose as to the fate of these pollutants.
Transport through the atmosphere
results not only from  horizontal advec-
tion, but from vertical motions due to
mesoscale convective processes. Initial
attention was paid to  non-precipitating
cumulus congestus clouds. Speculation
as  to  the impact of deep three-
dimensional cumulonimbus-scale verti-
cal  convective motions soon arose
along with the hypothesis that the im-
pact of large thunderstorm systems
should be clearly visible in the haze air
masses when viewed by geostationary
satellite systems. As  part of the PEPE/
NEROS-80, an active  effort  was
launched to  use the GOES satellites to
detect  mesoscale convective system
(MCS) thunderstorm impact  upon
boundary layer  aerosol  patterns.
  The authors previously presented an
extensive case study  that indicated
massive displacement of sulfate aerosol
and ozone from the boundary layer over
a mid-Atlantic multi-state area due to
the passage of a mesoscale convective
complex and a  related  squall  line sys-
tem. A large clear area, subsequently
termed a convective aerosol  removal
event (CARE), was detected both in the
GOES visible satellite imagery and by
the analysis of surface visibility reports.
It is hypothesized that thunderstorms
are an important sink for sulfate  aero-
sols and other pollutants, not only due
to wet removal to the  surface, but more
importantly, because  of the injection of
massive amounts of pollutants into the
middle and upper troposphere as a re-
sult of thunderstorm convective up-
drafts. Strong downdrafts replace the
near-surface air with air from mid-
tropospheric levels.  The  perturbations
induced  in the antecedent polluted
boundary layer may last for one or more
days until horizontal mixing processes
diffuse the CARE out of existence.
  These  preliminary results raised a
number of intriguing questions. First,
are CAREs formed  by smaller storms,
i.e., the more typical air mass thunder-
showers which dominate the southern
and eastern United  States during the
warm season? Second, were the events
of 2 August 1980 a unique occurrence,
not likely to be repeated with enough
frequency to invalidate  current trans-
port and  deposition modeling efforts?
Finally, are all such turbidity inhomo-
geneities detected in GOES satellite im-
agery the result of thunderstorms or are
there other mechanisms that could
come into play? These questions are ad-
dressed in the report.

Conclusions
  The primary purpose of this project
was to determine if a CARE could  be
associated with small air mass thunder-
storm systems. This situation contrasts
the extremely large CARE generated by
a massive mesoscale convective com-
plex and  squall  line  on 2 August 1980.
On 13 August 1980, a cluster of small
local  thunderstorms developed  over
northern  Florida and southern Georgia.
Detailed  mesoanalyses  showed that
several smaller mesohighs  eventually
congealed into one larger system,
which ultimately covered an area of
172,000 km2. By the next morning,  an
area  of  improved  visibility moved
across eastern  Georgia  and extreme
southern South Carolina. Visibility at
the affected stations improved  to  an
average of 18 km, for a total increase of
7.6 km. The estimated mean volumetric
decrease of sulfate aerosol  concentra-
tion within this CARE was on the order
of 15 (xg/rn3. These changes were of the
same order as those associated with the
MCC and squall line of 2 August 1980,
indicating little significant difference in
the basic efficiency of the process. How-
ever, the total mass of sulfate estimated
to have been displaced from the CARE
was only 7.7 x 105 kg, or  approximately
1/50 of that in the first case study. Thus,
evidence is presented that smaller thun-
derstorm clusters can produce CAREs
of similar intensity, but in smaller geo-
graphical areas, and consequently, with
lower sulfate mass (and presumably ox-
idant) displacement from the polluted
boundary layer.
  Since, with some exceptions, the up-
drafts of convective precipitation sys-
tems originate  in and  draw much of
their mass flux from the polluted plane-
tary boundary layer, this leads to specu-
lation as to the ultimate fate of the dis-
placed aerosol.  (A simple schematic
model is presented in the report.) Some
of the aerosol is, of course, removed to
the surface by wash-out and wet depo-
sition mechanisms. The  radially out-
ward spreading mesohigh, acting as a
plow, undercuts the polluted boundary
layer air, physically uplifting some of
the material in homogeneous layers to
altitudes as high  as 3000 m AGL Given
the intense vertical transport within a
storm's convective tower, large quanti-
ties of boundary layer pollutants are
transported to great heights. Some of
the material is detrained from the cumu-
lonimbus tower into the middle tropo-
sphere. It is suspected that most of the
material is eventually evaporated from
the anvil cloud debris near the  base of
the tropopause. Some material  may be
actually injected into the stratosphere
by overshooting turrets associated with
the more intense thunderstorms. Since
cumulonimbi are noted for their rain-
producing capability, some of the pollu-
tants must certainly become involved in
the water-phase chemistry and con-
tribute  to acidic precipitation. What re-
mains to be quantitatively evaluated is
the percentage of the original boundary
layer pollutants vented into the anvil re-
gion of the cloud versus those that
rained  out to  the surface. Since the
more intense supercell  thunderstorms
are known  to have comparatively low
precipitation efficiencies, these are pos-
tulated to be the  most efficient venters.
The probable role that  thunderstorms
play as sinks of  boundary layer pollu-
tion, not to the surface, but to the upper
troposphere, may need  to be seriously
considered as components of regional
deposition models.
  One  totally unexpected result of the
project was the discovery  of the
pseudo-CARE. Areas of lower turbidity
or "holes" within  the haze PEPE
air mass,  originally thought to  be
thunderstorm-induced CAREs, were un-
able to be related to specific mesoscale
precipitation systems. During a two-day
period  in June, 1975, there was intense
subsidence associated with lake
breezes over the Great Lakes. These in
turn deformed the polluted boundary
layer and  appear to have significantly
depressed the top of the aerosol layer.
Upon cessation of the lake breeze, these

-------
"aerosol holes" drifted with the north-
easterly flow into the Ohio Valley, main-
taining themselves for 18 to 24 hours.
Other apparent pseudo-CAREs were
discovered, including one long, narrow
swath of reduced cumulus convection
and apparent increased turbidity  that
was associated with a synoptic-scale
mid-tropospheric subsidence event.

Recommendations
  After completion of two case studies,
we have found evidence that three
major classes of thunderstorms gener-
ate CAREs. The  mesoscale  convective
systems studied  were an  air mass
thunderstorm cluster, a mesoscale con-
vective complex, and a squall line. At
least one major type of thunderstorm
(the isolated supercell thunderstorm)
has not yet been studied. In any case,
additional studies for all types of
thunderstorms are  required for a better
understanding of the  relationship be-
tween the storm type, total rainfall, and
sulfate displacement from the polluted
boundary layer.
  While the use of surface meterologi-
cal data allows reasonable estimates to
be made of the total aerosol displaced
during a CARE, these data, in turn,  pro-
vide virtually no information about the
ultimate fate of these aerosols. A key
question is how much material is deliv-
ered to the surface by wet  deposition
compared to how much is injected into
the free troposphere? Field observa-
tions are necessary to support the vari-
ous hypotheses of numerical  cloud-
venting modelers. An intensive field
program is needed during which  sev-
eral different approaches should be em-
ployed. First, the mass budget for sul-
fate and other aerosols during the life
cycle study of an isolated intense thun-
derstorm should be determined. This
study  should attempt to determine the
partitioning of aerosol  mass to the sur-
face.  It should  also determine how
much  remains at mid-levels and at the
base of the  tropopause once it is in-
jected  into the  free troposphere,  and
ideally, if any is  injected into the strato-
sphere. Tracers other than sulfate can
be employed. Ozone, which is less solu-
ble than many gases and does not  par-
ticipate directly in the precipitation
process, should be measured. However,
it should be noted that extensive intru-
sions of stratospheric ozone  into a tall
cumulonimbus cloud circulation could
complicate the interpretation of the re-
sults. Tracers such  as the new genera-
tion of hydrocarbons could prove highly
useful. By "spiking" the updraft around
an isolated  and identifiable thunder-
storm and measuring its total final re-
distribution, insight would be gained
into the vertical transport and disper-
sion properties of a thunderstorm. The
upcoming National STORM Program
provides an ideal setting for such an ex-
periment.
  Additional efforts should be made to-
ward the collocation of meteorological
wet and dry deposition and air quality
monitoring sites for  both  routine and
special networks. It is extremely diffu-
cultto interpret the results of changes in
aerosol and  other pollutant concentra-
tions if corresponding on-site meteoro-
logical information is not available. Fur-
thermore, whenever  possible,
measurements, particularly of sulfate,
should be made on shorter than 24-hour
intervals. Mesoscale systems are inher-
ently  short-period phenomena.  Long
sampling periods eliminate the high fre-
quency fluctuations that are critical to
the understanding of the  mechanisms
and impacts of thunderstorm  aerosol
displacement. Interpretation of aerosol
patterns as  seen in satellite imagery,
provides a valuable overview of atmos-
pheric processes.  However, they  are
very difficult to interpret, as are on-site
measurements  of  aerosols and  other
pollutants, without the appropriate sup-
porting meteorological information. It
must be understood that the data collec-
tion and archival systems of the nation's
meteorological  services are extremely
crude and do not provide (at least cost-
effectively) most of the information re-
quired for the proper analysis of satel-
lite and pollutant measurements. High
resolution radar data, real-time  light-
ning stroke  data, and  high resolution
animated  GOES imagery  are,  in gen-
eral, not archived in convenient formats
as part of routine government opera-
tions. This must be done on a real-time
basis by using operational facilities cur-
rently available from both the govern-
ment  and  private sectors.  Attempts to
acquire such data after  the  fact  are
costly and will almost always produce
lower quality and less useful  informa-
tion  to support field program data
analyses.
  The  unexpected discovery of the
pseudo-CARE  phenomenon suggests
that the air quality community should
further interface with dynamic meteor-
ologists. There appear to be a number
of dry atmospheric processes that lead
to mesoscale subsidence  phenomena
that significantly alter boundary layer
pollution profiles. The intrusion of large
amounts of mid-tropospheric air into
the boundary layer, unless recognized
as such, could seriously bias the inter-
pretation of measurements made dur-
ing aerosol transport and transforma-
tion experiments.

-------
     Walter A. Lyons and Rebecca H Calby are  with MESOMET. Inc.. Chicago,  IL
      60601.
     Francis S. Binkowski is the EPA Project Officer (see below).
     The complete report, entitled "Further Case Studies on the Impact of Mesoscale
      Convective Systems on Regional Ozone and Haze Distributions," (Order No. PB
      86-194 453/AS; Cost: $11.95, 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
Official Business
Penalty for Private Use $300

EPA/600/S3-86/028
                0000329    PS
                                                AGENcy

-------