Re
pr
inted
from
J
ouRNAL
OF
APPL
I
ED
METEOROLOGY,
Vo
l. 20,
No.
11,
Nove
m
ber
1
98
1
American Meteorological Soci
ety
Printed
in U. S. A.
Convective Downmixing
of
Plumes in a Coastal Environment
GR
EGO
RY
J . M c R
AE,
FR
E
DRI
CK H.
SH
A
IR
and J o
HN
H . S E
IN
FEL
D
1312
JOURNAL
OF
APPLIED
METEOROLOGY
VOLUME
20
Convective Downmixing
of
Plumes in a Coastal Environment
GREGORY
J. McRA
E,
FREDRJCK
H.
SHAIR
1
and
JOHN
H.
SEINFELD
1
Environmen
tal Quality
Laboratory.
Ca
lif
ornia Ins
titut
e
of
Technology.
Pasadena
91/25
(Manuscript received 30 April
1981
,
in
final form 2 August 1981)
ABSTRACT
This
paper
describes the r
es
ults
of
an
atmospheric
tracer
study in which
su
lfur hexafluoride
(SF
8
)
was
used
to
investigate
the
transport and dispersion
of
efflue
nt
from a
power
plant located in a coastal
environment.
The
field study
demon
s
tr
ated
th
at material emitted into an elevat
ed
st
ab
le l
ayer
at
night
can
be tran
spo
rted out
over
the
ocean,
fumigated
to
the surface,
and
then
be
returned
at
ground
level
by
the
sea
breeze
on
the next
day.
At night when cool
stab
le
air
from
the
land
encounters
the
warmer
ocean
convective
mixing
erodes
the
stab
le layer forming
an
internal
boundary
layer. When the growing
boundary
layer
encounters
an elevated plume
the
pollutant material, eotrained
at
the
top
of
the
mixed
layer.
can
be
rapidly transported
in
- 20 min
to
the surface. Various expressions for
the
characteristic
downmixing time
(X
= Z
1
1w.)
are
developed utilizing
the
gradient Richardson
number,
the Monin-
Obukhov
length
and
turbulence intensities. Calculations using
these
expressions a
nd
the
field
data
are
compared
with simil
ar
stu
die
s
of
co
nvective mixing
over
the land.
1.
Introduction
A
major
influence
on
pollutant
dispersion and
transport
in
coastal
environments
is
the
presence
of
land/sea
breeze
circulation
systems.
Unfortunately
the
characterization
of
turbulent
transport
is compli-
cated
by
the
presence
of
flow reversals
and
differing
atmospheric
stabilities. Since many large
sources
are
located
in
shoreline
environments,
it is
important
to
understand
the
mixing
characteristics
within
the
boundary
layer. A field
experiment
designed
to
de
-
termine
the
fate
of
pollutants
emitted
into the off-
shore
flow
associated
with a land/sea
breeze
circula-
tion
system,
was
carried
out
by
Shair
e
ta/
. (1981).
In
that
study
it
was
found
that
tracer
material
emitted
into an
elevated
stable
layer
at
night could
be
transported
out
over
the
ocean,
fumigated to
the
surface,
and
then
be
returned
at
ground level
by
the
sea
breeze
on
the
next
day.
The
objectives
of
this
work
are
to
examine
the
vertical
transport
processes
responsible
for this rapid downmixing
and
to
char-
acterize
the
mixing rates within the internal boundary
layer
formed
when
cool
air
from
the
land is
advected
out
over
a
warm
ocean
surface.
2. Description
of
field experiment
Because
of
the
complexity
of
atmospheric
flows,
the
only direct
way
to
relate
the
emissions from a
particular
source
to
observed
concentrations
is
to
tag the
source
exhaust
gases
so
they
can
be uniquely
identified.
Over
the
last
few years a variety
of
1
D
epa
rtment
of
Chemical Engineering.
0021-8952/8
1/11
1312-13$07.25
«:>
1982 American Meteorological
Society
atmospheric
tracer
s,
including sulfur hexafluoride
(SF
6
),
fluorescent particle
s,
halocarbons
and
deu-
terated
meth
a
ne
, h
ave
been
used
in
transport
and
diffusion studies. Sulfur hexafluoride
was
u
se
d in
this experiment
because
it is gaseous, physiologically
inert
, chemically s
table
, and easily
detected
using
electron-capture
gas
chromatography
(Simmonds
et
at.,
1972). Drivas
and
Shair
(1974),
Lamb
et at.
(l978a,b)
and
Dietz
and
Cote
(1973)
have
success-
fully
demonstrated
the
utility
of
SF
6
as
a
tracer
in
large-scale field studies.
Current
analysis
technique
s
have
achieved
detection
limits
of
2 X
I0
-
13
part
s
SF
6
per
part
of
air.
From
a practical
point
of
view
both
the release
technique
s
and
sampling
protocols
are
well established
and
reliable.
Each
experiment
was
carried
out
by
injecting
the
tracer
gas into the
number
4 stack
of
the
Southern
California Edison
El
Segundo
power
plant
located
on
the
s
hore
of
Sant
a Monica
Bay
(Fig. 1).
This
particular
chimney
is
61
m high
and
4.3
min
diam-
eter.
The
tracer
was
released
at
a
time
when
the
flow,
at
the effective s
tack
height,
was
offshore.
Before
each
experiment
an initial
estimate
of
the
plume
rise
was
determined using Briggs' formulas (Briggs,
1969; 1975) for neutral
conditions.
For
the
particular
load conditions (0.57
of
capacity),
an
exhaust
gas
temperature
of
365 K and a gas flow
rate
of
230 m
3
s-
1
,
the
plume rise
was
estimated
to
be
250 m.
This
information,
together
with
the
vertical wind distribu-
tion obtained from pibal releases,
was
used
to
es-
tablish
the
time
to
initate
the
tracer
injection
so
that
the
material
was
released into the offshore flow.
After the
experiment
a more detailed calculation,
~ANTA
MONICA
BAY
KEY
PLUME CENTERLINE
.....
.•..
.
AT
4 A.M
FROM
SURFACE
WI
NOS
----
~!A
·
:~:o~OM
AT
3A.M
.
FROM
300"'
WINDS
SCALE 1:
250,000
9 ;
!0
"
..
...
0 5
jQ
18
ZOKiOI'fttten
(os
41-Q
~~s
&
'<lsq,
=t
0 I I • • 1
2300
0 I 2
=t
'""
"
""
(SF
6
).
PPT 0 I
...
I - I
2002
3 4 5
1
00
Q l
..........--
....
\r
I
.f!
I
............
5 6
i1
7 8
200
100
0
9
9
10
II
=k
I:
I I
II
12 13
14
TIME, P.O.T.
FIG
.
I.
Sulfur hexafluoride (SF
8
)
measurements made on board
RJV
Acania
22
July
1977
coordinated with ship course and possible plume trajectories
derived from surface and elevated wind measurements:
(e
),
release site; (e) onshore monitoring s
it
es.
z
0
<
"'
l:
"'
"'
"
"'
00
:::
(')
;lj
>
[TJ
en
::t
>
;lj
>
z
0
en
[TJ
z
"Tl
[TJ
r
0
VJ
...
VJ
SANTA
MONICA
BAY
KEY
PLUME
CENTERUNE
AT
4 A.M.
FROM
SURFACE
WINOS
FROM
100
M WINOS
FROM
300M
WI
NOS
<os
41-'Q
~~s
0 !
10
15
20
KIIOfiliettrs
s
~sq;
400~
200
L
----------~------------·~~----~~~-;
0 I 2
:=
0
'""
2<,
"~
1
5
•lt .
~.
PPT 0
2
3 4 5
400~
200~-~1
05
6 7 8
•ool_
2
:~
~
•
1~1
::~
' ' '
0
12 13
14
II
TIME,
P.O.T.
FIG.
2.
Sulfur hexafluoride
(SF
6
)
measurements made on board R/V Acania
24
July
1977
coor
dinated with ship course and possible plume trajectories
derived from surface and elevated wind measurements: (e ) release site; (e) onshore monitoring sites.
-
\;)
-
.$»
0
c::
:;d
z
>
r
0
"!j
>
.,
.,
r
....
ttl
0
a::
m
~
m
0
:;10
0
r
0
0
-<
~
r-
c::
3:
I'll
""
0
NOVEMBER
1981
McRAE
,
SHAIR
AND
SEINFELD
1315
accounting
for
the
actual
vertical
variations
in
wind
and
temperature
profiles,
was
carried
out
using
the
Schatzmann
(1979) integral
plume
rise
model,
using
meteorological
data
from
Schacher
et
a!. (1978).
During
the
first
test,
on
22
July
1977, 90 kg
of
SF
6
was
released
at
a
rate
of
5.0
g s-
1
from
0005-0500
(all
times
PDT).
During
the
second
test
245 kg
of
SF
6
were
released,
at
a
higher
rate
of
I3.6
g s-
1
,
from
2303
on
23
July
1977 until 0400
on
24
July.
The
amount
and
release
rates
for
each
experiment
were
selected
so
that
there
was
sufficient
material
to
distinguish
the
source
from
the
background
at
the
maximum
sampling
distance.
If
the
total
amount
of
tracer
released
during
each
experiment
were
to
be
uniformly
distributed
throughout
a
volume
of
1600
km
2
x 300 m
(i.e.,
the
area
of
Santa
Monica
Bay
times
the
estimated
plume
rise
above
the
ocean
surface),
then
the
average
tracer
concentration
would
have
been
50
ppt
2
,
a
value
well
above
both
the
detection
limit
and
normal
background
levels.
Most
of
the
current
world
background
concentration
of
<
0.5
ppt
is
a
result
of
leak
ages
from
high-voltage
power
transformers
and
switch
ing
systems
where
SF
6
is
used
for
corona
discharge
suppression.
Hourly
averaged
air
samples
were
collected
con-
tinuously,
from
0500-
1700
during
each
of
the
test
days,
at
29
coastal
sites
located
from
Ventura
to
Corona
del
Mar
(Figs. 1
and
2).
This
was
to
observe
the
tracer
flux
across
the
coast
during
the
sea
breeze
on
the
day
following
the
nighttime
release
.
Subsequent
mass
balance
calculations
using
these
measurements
were
able
to
account
for
virtually
100%
of
the
material
released
during
both
experi-
ments
(Shair
et
at
.,
1981).
Samples
were
analyzed
using
the
methodology
described
in
Lamb
et a!.
(l978a,b).
In
addition,
grab
samples
were
coiJected
every
5 min
on
board
a
ship
traversing
Santa
Monica
Bay
and
analyzed
using
portable
electron-capture
gas
chromatographs
.
This
sampling
protocol
pro-
vided
rapid
feedback
on
the
tracer
concentrations
and
plume
position
during
each
experiment.
The
measurements
taken
on
board
the
ship
are
shown
in
Figs.
I
and
2.
Sampling
on
board
the
ship
was
started
I h
before
each
release
so
that
any
possible
background
levels
could
be
detected.
All
samples
were
collected
in 30
cm
3
plastic
syringes
and
were
analyzed
within
one
day
of
each
experiment.
At
the
coastal
monitoring
sites
battery-powered
sequential
samplers
were
used
to
determine
the
hourly
aver-
aged
SF
6
concentration
levels.
In
addition
auto-
mobile
sa
mpling
traverses
were
conducted
peri-
odically
along
coastal
highways
between
1000-1427
on
22
July
and
between
02?
'i-l540
on
24
July.
Grab
samp
les
were
collected
at
0.8-3
.2
km
int
ervals
along
the
coasta
l
highway
between
Redondo
Beach
and
Malibu.
The
results
from
the
shore
measure-
2
Parts
per
trillion.
ments
and
automobile
traverses
were
used
by
Shair
et a/.
(l98I)
to
calculate
the
flux
of
SF
6
across
the
coast.
The
tracer
experiments
were
carried
out
in
collaboration
with
investigators
from
the
Environ-
mental
Physics
Group
at
the
Naval
Po
stgraduate
School
in
Monterey,
California
.
The
research
vessel
Acania
was
used
as
a
platform
to
collect
meteoro-
logical
data
in
the
vicinity
of
Santa
Monica
Bay.
The
ship
was
equipped
with
a
complete
suite
of
meteoro-
logical
equipment
capable
of
multilevel
measure-
ments
(4.2,
7.0
and
22.5 m
above
the
ocean)
of
mean
and
fluctuating
quantities.
Since
complete
details
of
the
instrumentation
can
be
found
in
Houlihan
et
a/.
(1978)
and
Schacher
et
a/
. (1978),
the
ma-
terial will
not
be
repeated
here.
For
the
particular
st
udy
of
the
mixing
rates
over
the
ocean,
measure-
ments
were
made
of
sea
surface
temperature
T
8
,
air
temperature
Ta,
humidity/dew
point
Td,
true
wind
speed
u,
direction
e,
and
temperature
inversion
height
Z
1
•
The
wind
direction
(}is
particularly
useful
since
it
can
be
used
to
differentiate
local
(land
and
sea
breeze)
circulations.
Both
the
wind
speed
and
direction
have
been
corrected
to
account
for
the
ship
motion.
In
addition,
during
the
period
19-26
July,
I4
radiosondes
were
released
to
examine
the
vertical
temperature
structure.
During
each
tracer
experiment
pibals
were
rele
ase
d
each
hour
at
a
site
close
to
the
release
point
so
that
the
horizontal
winds
as
a
function
of
elevation
could
be
determined
. Ob-
servations
made
at
the 100
and
300m
levels
were
used
to
calculate
plume
trajectories
from
the
release
point.
Some
of
these
results
are
superimposed
on
Figs.
1
and
2.
The
complete
data
sets
describing
the
meteorological
conditions
are
contained
in
the
reports
by
Schacher
et
a!. (1978, 1980).
For
con-
venience
a
summary
of
key
information
from
these
sources,
together
with
the
calculated
virtual
heat
flux Q
0
,
is
presented
in
Table
I.
Since
the
pattern
of
results
observed
on
board
RJV
Acania
on
both
days
were
similar
it
suffices to
discuss
the
experiment
conducted
on
22
July.
A
more
detailed
discussion
of
the
concentration
levels
measured
at
the
coastal
monitoring
stations
is
con-
tained
in
Shair
eta
!. (1981).
Prior
to
0530
PDT,
when
the
mixing
depth
was
below
200m,
the
ship
passed
under
the
calc
ulated
plume
positions
at 0100, 0325
and
0438
and
no
significant
concentrations
of
SF
6
were
observed.
At
0530,
when
the
ship
was
6.4
km
south
of
the
plume,
the
first significant
peak
(80
ppt)
was
recorded
at
a
time
when
the
mixed
layer
was
growing
above
the
200 m level.
From
0600
onward
all
the
concentration
peaks
at
0730, 0835
and
0925
were
observed
when
the
ship
was
in
the
vicinity
of
the
plume
and
the
mixed-layer
height
was
above
200m.
From
0830
to
1130
the
SF
6
exceeded
20
ppt
and
the
ship
was
always
within
3
km
of
the
plume.
Lower
concentrations
were
observed
when
the
ship