U.
S.
Environmental
Protection
Agency
Technical
Support
Document
for
PM2.5
Designations
­
Supplemental
Notice
April
5,
2005
Technical
Support
Document
for
PM2.5
Designations
­
Supplemental
Notice
April
5,
2005
Table
of
Contents
Section
1.
PM2.5
Monitoring
Data
for
areas
changing
from
nonattainment
to
attainment
based
on
2002­
2004
data
Section
2.
Additional
discussion
for
areas
requesting
to
change
the
status
of
an
individual
county
from
nonattainment
to
attainment.

A.
Indiana
B.
Michigan
C.
Ohio
Section
3.
Chattanooga,
TN
request
to
invalidate
multiple
monitoring
samples
and
change
status
to
attainment.

A.
Summary
B.
Chattanooga
design
value
analysis
C.
EPA
memorandum
1
D.
EPA
memorandum
2
E.
Trajectory
analyses
F.
Analyses
of
chemical
composition
data
Section
4.
Columbus,
GA­
AL
spatial
averaging
proposal
and
request
to
change
status
to
attainment.
Section
2.
Additional
discussion
for
areas
requesting
to
change
the
status
of
an
individual
county
from
nonattainment
to
attainment.

This
section
discusses
four
areas
for
which
states
requested
to
change
specific
counties
from
nonattainment
to
attainment
because
they
now
have
monitors
with
2002­
2004
design
values
below
the
level
of
the
standards.
In
all
of
these
cases,
EPA
finds
that
the
counties
in
question
nevertheless
contribute
to
the
overall
air
quality
problem
in
the
area
and
should
remain
designated
as
nonattainment.

2.
A.
Indiana
Indiana
requested
that
EPA
limit
the
nonattainment
designations
to
only
counties
with
monitors
showing
a
violation
of
the
NAAQS.
EPA
notes
that
Section
107(
d)
of
the
Clean
Air
Act
requires
that
a
county
which
violates
the
National
Ambient
Air
Quality
Standards
(
NAAQS,)
or
that
contributes
to
a
violation
in
a
nearby
area,
must
be
included
in
the
nonattainment
area.
The
monitored
air
quality
for
each
county
is
just
one
of
the
factors
that
EPA
uses
to
determine
which
counties
are
included
in
the
nonattainment
area.
Indiana
objected
to
EPA's
inclusion
of
counties
that
were
monitoring
attainment,
or
that
have
no
monitor
at
all,
but
which
EPA
found
to
be
contributing
to
a
violating
monitor
in
another
county.

EPA
evaluated
counties
near
monitors
violating
the
fine
particule
standard
to
determine
if
those
counties
contribute
to
the
violation.
We
analyzed
counties
using
nine
factors
including
emissions,
commuting
patterns,
and
population.
EPA
did
not
designate
a
county
as
nonattainment
simply
because
it
contained
a
power
plant.
We
examined
emissions
data
as
a
first
indicator
of
a
county's
potential
to
contribute
to
violations.
However,
EPA
promulgated
nonattainment
designations
for
counties
based
on
an
overall
weight
of
evidence
analysis
of
the
nine
technical
factors
described
in
EPA
guidance.
In
a
number
of
cases,
when
the
contribution
from
a
specific
county
was
attributable
primarily
to
a
single
significant
emissions
source
and
the
rest
of
the
county
showed
little
contribution,
EPA
designated
a
partial
county
area
as
nonattainment.
Large
power
plants
have
significant
emissions
and
are
commonly
judged
to
be
important
contributors
not
just
to
regional
background
concentrations
but
also
to
local
PM2.5
concentrations.

Indiana
suggested
that
only
the
counties
with
violating
monitors
should
be
designated
as
nonattainment.
This
is
contrary
to
the
statutory
directive
that
EPA
should
designate
not
only
counties
with
violating
monitors,
but
also
those
that
contribute
to
violations
in
nearby
counties.
In
addition,
EPA
does
not
support
such
an
approach
because
it
would
provide
a
disincentive
for
States
to
monitor
air
quality.
Indiana
also
noted
that
the
monitoring
data
continues
to
show
a
trend
of
decreasing
fine
particulate
levels
across
the
state.
EPA
is
glad
that
Indiana's
ambient
air
quality
is
improving.
However,
this
is
not
a
basis
for
limiting
nonattainment
designations
only
to
those
counties
with
violating
monitors.
EPA
is
obligated
to
include
those
counties
that
contribute
to
violations
of
the
standard
in
nearby
areas
based
upon
the
data
before
the
Agency.

Indiana
also
stated
that
it
feels
that
a
nonattainment
designation
will
impede
economic
progress
and
that
future
rules
will
control
emissions
and
bring
all
areas
into
attainment.
We
believe
that
economic
progress
and
attainment
of
the
NAAQS
are
not
mutually
exclusive
goals.
EPA
agrees
that
current
or
near
term
regulatory
efforts
by
EPA,
such
as
the
Clear
Air
Interstate
Rule,
will
do
much
to
alleviate
regional
nonattainment,
but
there
will
continue
to
be
a
need
for
local
controls
in
some
areas
in
order
to
achieve
the
NAAQS.
In
addition,
future
rules
and
future
reductions
from
current
rules
are
not
considered
by
EPA
for
making
designations
because
the
area
analysis
uses
current
emissions
and
air
quality.

Lake
and
Porter
counties,
part
of
Chicago
Nonattainment
Area
Both
northwestern
Indiana
counties
are
in
the
Chicago
nonattainment
area.
Indiana
noted
that
the
2002
to
2004
design
values
for
all
sites
in
Lake
and
Porter
Counties
are
below
the
annual
PM2.5
standard.
The
2001
to
2003
design
value
for
Lake
County
was
above
the
standard.
Indiana
requested
that
EPA
change
the
designation
for
Lake
and
Porter
Counties
be
changed
to
attainment/
unclassifiable.

EPA
included
Lake
and
Porter
Counties
in
the
Chicago
nonattainment
area
because
of
the
violation
in
Lake
County
and
high
emissions,
populations,
and
significant
commuting
in
both
counties.
Although
the
design
value
for
the
monitor
in
Lake
County
is
now
below
the
level
of
the
standard
after
consideration
of
2004
data,
the
other
factors
indicate
that
Lake
County
and
Porter
County
both
contribute
to
the
overall
air
quality
problem
in
the
metropolitan
Chicago
area
and
that
these
counties
should
remain
in
the
Chicago
nonattainment
area.
As
EPA
explained
in
the
January
5,
2005,
notice,
it
believes
that
it
is
appropriate
to
alter
the
designation
for
a
nonattainment
area
based
upon
2004
data,
if
all
counties
within
that
area
are
monitoring
attainment
based
on
2002­
04
data.
When,
as
in
this
case,
there
is
a
continuing
violation
of
the
standard
in
the
area
and
the
counties
continue
to
contribute
to
that
violation,
EPA
believes
that
it
is
inappropriate
to
alter
the
designation
of
such
counties.

Evansville
Indiana
disagreed
with
the
designations
in
the
Evansville
area.
Previously,
Indiana
requested
splitting
the
Southwestern
Indiana
area
into
two
parts.
The
State
suggested
that
Vanderburgh,
Warrick,
and
part
of
Gibson
Counties
should
be
one
area
and
that
Dubois
County
and
parts
of
Pike
and
Spencer
Counties
should
be
another
area.
We
have
concluded
that
splitting
the
area
was
not
appropriate,
given
the
relative
geographic
proximity
of
the
counties
and
the
regional
nature
of
the
PM2.5
problem.
For
the
Evansville
Area,
EPA
designated
a
single
nonattainment
area
that
includes
Dubois,
Vanderburgh,
and
Warrick
Counties
as
well
as
portions
of
Gibson,
Pike,
and
Spencer
Counties.
Indiana
asked
us
to
reconsider
splitting
the
Evansville
area
into
two
parts
and
to
change
the
designation
for
Vanderburgh,
Warrick,
and
Gibson
Counties
to
attainment/
unclassified
based
upon
2004
data.
The
2004
monitoring
data
shows
that
all
monitors
in
Vanderburgh
County
are
now
below
the
annual
PM2.5
standard.
However,
Dubois
County
continues
to
have
a
design
value
above
15.0
µ
g/
m3.

EPA
previously
concluded
that
the
violations
in
Vanderburgh
and
DuBois
Counties
arise
in
part
from
contributions
from
Gibson,
Pike,
Spencer,
and
Warrick
Counties.
EPA
continues
to
believe
that
these
counties
contribute
to
violations
in
DuBois
County.
While
EPA's
prior
inclusion
of
Vanderburgh
County
in
the
nonattainment
area
reflected
both
the
violation
within
the
county
and
the
contribution
to
the
broader
ambient
air
problem
in
the
area
and
the
violating
monitor
in
DuBois
County,
EPA
believes
that
the
contribution
of
emissions
in
Vanderburgh
County
to
violations
in
DuBois
County
by
itself
warrants
inclusion
of
this
county
in
the
nonattainment
area.
Indeed,
Vanderburgh
County,
which
includes
the
core
city
of
Evansville,
has
well
over
half
of
the
metropolitan
area
population,
slightly
under
half
of
the
metropolitan
area
vehicle
miles
traveled,
and
a
significant
fraction
of
the
area's
emissions.
EPA
continues
to
believe
that
a
single
area
with
three
full
counties
and
three
partial
counties
is
appropriate.
Since
all
the
counties
in
the
area
are
not
attaining,
even
after
inclusion
of
2004
data,
and
Vanderburgh
county
contributes
to
the
air
quality
violation
in
the
metro
area,
EPA
has
determined
that
the
entire
Evansville
area
will
remain
designated
as
nonattainment
for
fine
particulate.
This
is
consistent
with
EPA's
position
taken
with
respect
to
inclusion
of
2004
data
in
the
January
5,
2005
notice.

2.
B.
Michigan
Detroit
area
In
response
to
the
January
5,
2005,
Federal
Register
notice,
the
Michigan
Department
of
Environmental
Quality
(
MDEQ)
requested
a
change
of
designation
status
to
attainment
for
Livingston,
Oakland,
Macomb,
Monroe,
St.
Clair,
and
Washtenaw
Counties
within
Southeast
Michigan.
Michigan
worked
diligently
to
have
the
2004
monitoring
data
completed,
quality
assured,
and
certified
within
the
time
frame
indicated
in
the
January
5,
2005
notice.
However,
EPA
explained
that
it
would
consider
modification
of
the
initial
designations
only
if
each
county
in
the
area
is
monitoring
attainment
based
upon
inclusion
of
2004
data.
Because
Wayne
County
is
still
monitoring
violations
of
the
PM2.5
standard,
EPA
concludes
that
a
change
of
designation
from
nonattainment
to
attainment
is
not
warranted
for
the
counties
listed
above.
Although
Monroe
County
now
shows
attainment
with
the
2002­
2004
data,
EPA
has
concluded
that
Monroe
County
and
the
other
nearby
counties
contribute
to
the
violation
in
Wayne
County.
Indeed,
in
ranking
composite
emission
scores
for
counties
within
the
Detroit
area,
Monroe
County
has
a
score
that
is
second
only
to
Wayne
County,
and
the
county
is
generally
upwind
from
violations
recorded
in
Wayne
County.
Based
upon
analysis
of
all
of
the
factors,
EPA
concluded
that
Monroe
County
should
remain
nonattainment
because
Monroe
County
contributes
to
the
violating
monitor
in
Wayne
County.

As
we
have
previously
stated,
once
an
area
has
a
monitor
violating
the
NAAQS,
EPA
evaluates
emissions
data,
along
with
other
information
for
nine
technical
factors,
to
help
determine
which
counties
in
the
area
are
contributing
to
the
violation.
The
PM2.5
(
air
quality)
weighted
emissions
scores
are
considered
in
the
context
of
all
the
relevant
factors
in
determining
the
boundary
of
a
nonattainment
area.
EPA
must
follow
the
Clean
Air
Act's
prescription
to
include
both
the
violating
area
and
all
nearby
areas
that
contribute
to
the
violation,
thereby
providing
for
implementation
of
the
full
range
of
Clean
Air
Act
provisions
(
including
but
not
limited
to
the
attainment
planning
requirement)
that
help
address
nonattainment
problems.

2.
C.
Ohio
Huntington­
Ashland,
OH­
KY­
WV
In
a
letter
dated
February
14,
2005,
the
State
of
Ohio
submitted
and
certified
PM2.5
data
for
2004.
This
letter
requested
that
the
Toledo
and
Youngstown
areas
and
Ohio's
portion
of
the
Huntington/
Ashland
area
be
designated
as
attainment
for
the
PM2.5
NAAQS
based
on
inclusion
of
2004
data.

Based
on
data
submitted
by
Ohio
(
and
data
from
Pennsylvania
pertinent
to
the
Youngstown
area),
EPA
has
determined
that
all
monitors
in
the
Toledo
and
Youngstown
areas
now
show
attainment
of
the
standard.
EPA
is
modifying
the
designation
for
these
areas
to
reflect
the
inclusion
of
2004
data.

Ohio's
request
for
the
Huntington/
Ashland
area
presents
more
complicated
issues.
The
Ohio
portion
of
the
Huntington/
Ashland
PM2.5
nonattainment
area
consists
of
Lawrence
and
Scioto
Counties
and
portions
of
Adams
and
Gallia
Counties.
Ohio
states
that
the
highest
3­
year
average
PM2.5
concentration
in
the
Ohio
portion
of
this
nonattainment
area
for
the
years
2002­
2004
is
14.80
µ
g/
m3
(
the
National
Ambient
Air
Quality
Standard
for
PM2.5
is
15
µ
g/
m3).
The
State
recognized
that
a
monitor
in
Cabell
County,
West
Virginia
(
which
is
part
of
the
Huntington/
Ashland
nonattainment
area)
is
not
measuring
attainment
for
2002­
2004,
which
prevents
the
entire
nonattainment
area
from
being
considered
as
attainment.
The
State
claimed
that
the
Ohio
portion
of
the
Huntington/
Ashland
nonattainment
area
has
demonstrated
significant
local
emission
reductions,
which
have
improved
the
air
quality
in
the
Ohio
portion
of
the
nonattainment
area.
The
State
also
claimed
that
the
Cabell
County
violations
result
from
local
sources,
including
the
AK
Steel
facility
and
the
Marathon/
Ashland
refinery,
rather
than
from
sources
elsewhere
in
the
area
designated
as
part
of
this
nonattainment
area
by
EPA.
Also
at
issue
is
whether
Scioto
and
Adams
Counties,
which
Ohio
labels
"
the
Portsmouth
area",
should
be
identified
as
a
separate
area
from
the
Huntington/
Ashland
area.

In
the
January
5,
2005,
notice,
EPA
explained
that
it
would
consider
changes
to
the
designation
of
an
area
only
if
every
county
within
that
area
would
be
deemed
in
compliance
with
the
NAAQS
as
a
result
of
inclusion
of
the
data
from
2004.
Because
the
Cabell
County,
West
Virginia,
monitor
continues
to
show
nonattainment,
Ohio's
submittal
does
not
meet
the
criteria
EPA
identified
for
mofification
of
the
designation
based
on
2004
data.

Nevertheless,
EPA
examined
Ohio's
recommendation
to
revise
the
boundaries
of
the
Huntington/
Ashland
PM2.5
nonattainment
area.
For
Adams
and
Gallia
Counties,
Ohio
provided
no
new
information
to
support
a
revision
to
the
designation.
These
counties
do
not
have
monitoring
data,
but
both
have
very
high
emissions
levels
that
EPA
judged
to
contribute
to
violations
in
the
Huntington
area,
and
EPA
has
no
reason
to
change
that
judgment.
For
Lawrence
County,
the
county
is
within
the
main
portion
of
the
metropolitan
area
(
MSA),
and
has
sufficient
emissions
to
warrant
continued
inclusion
in
the
nonattainment
area
because
of
its
contribution
to
nonattainment
in
the
area
as
a
whole.

For
Scioto
County,
although
the
county
is
outside
the
presumptive
boundaries
nonattainment
area
because
it
is
outside
the
CMSA
for
Huntington/
Ashland,
EPA
concludes
that
the
emissions
levels
within
the
county
and
other
factors
justify
inclusion
of
the
county
because
of
its
contribution
to
nonattainment
in
the
area.
EPA
notes
that
the
emissions
from
the
county
are
comparable
to
those
of
other
counties
that
EPA
included
in
the
nonattainment
area,
thus
warranting
continued
inclusion
of
this
county
because
it
is
contributing
to
nonattainment
in
the
Huntington/
Ashland
area.
Although
EPA
concurs
with
Ohio
that
air
quality
improvements
at
the
Portsmouth
monitor
site
correlate
closely
with
the
shutdown
of
the
New
Boston
Coke
facility,
EPA
nevertheless
believes
that
emissions
in
Scioto
County
continue
to
contribute
to
violations
elsewhere
in
the
area.
In
particular,
even
if
the
emissions
in
Scioto
County
have
been
reduced
by
the
shutdown
of
the
New
Boston
coke
plant,
they
have
been
and
will
continue
to
be
increased
by
construction
and
operation
of
a
new
coke
plant
that
is
even
closer
geographically
to
the
monitored
violations
in
this
nonattainment
area.
Therefore,
EPA
believes
that
it
would
be
inappropriate
to
treat
Scioto
County
or
Scioto
and
Adams
Counties
as
a
separate
air
quality
planning
area.
EPA
believes
that
the
Huntington/
Ashland
area
should
remain
nonattainment
and
retain
the
same
boundaries
as
published
on
January
5,
2005.
Section
3.
Chattanooga,
TN
request
to
invalidate
multiple
monitoring
samples
and
change
status
to
attainment.

3.
A.
Summary
In
December
2004,
EPA
designated
Hamilton
County,
TN,
and
Catoosa
and
Walker
Counties,
GA
as
nonattainment.
The
monitors
in
Hamilton
and
Walker
counties
had
three
years
(
01
­
03)
of
data
showing
design
values
above
the
standard.
Catoosa
was
included
due
to
its
contribution
to
both
Hamilton
and
Walker
Counties.
As
allowed
by
EPA's
final
designations
rule,
both
TN
and
GA
submitted
their
2004
quality
assured
and
certified
PM
air
quality
data
to
EPA
for
the
counties
in
question.
The
States
requested
that
fifteen
days
during
2003
and
2004
of
data
be
"
flagged"
due
to
influence
from
agricultural
fires
and
wildfires.
Previously,
TN
had
requested
that
10
days
in
2002
be
flagged
and
Region
4
rejected
the
flags.
This
new
submittal
included
a
request
that
the
revised
monitoring
data
be
considered
and
the
designation
of
the
area
changed
to
attainment
or
unclassifiable
prior
to
April
5,
2005.

EPA
has
determined
that
at
least
7
of
these
fifteen
days
should
not
be
flagged
as
exceptional
events.
The
trajectory
analyses
conducted
by
OAQPS
do
not
support
the
contention
that
these
data
are
affected
by
the
cited
agricultural
or
wildfires.
For
the
remaining
seven
days,
trajectory
analyses
do
not
immediately
rule
out
the
possibility
that
agricultural
fires
and
wildfires
had
an
effect
on
the
air
quality
monitors
in
the
Chattanooga
area.
However,
EPA
does
not
have
sufficient
supporting
data
from
the
State
to
determine
whether
the
fires
on
these
days
affected
air
quality
in
Chattanooga
and
if
they
did,
whether
they
should
be
flagged
as
exceptional
events
and
removed
from
the
data
set
of
air
quality
considered
for
designation
purposes.
Moreover,
even
if
these
7
days
were
flagged
and
removed
from
the
air
quality
data
set
because
EPA
agreed
that
they
should
qualify
as
exceptional
events
that
may
properly
be
excluded
from
designation
decisions,
the
Hamilton
County
monitor
would
continue
to
be
nonattainment.

On
those
seven
days
that
EPA's
trajectory
analysis
indicated
that
there
may
have
been
impacts
resulting
from
a
fire
event,
EPA
looked
at
speciation
data
that
was
available.
Of
the
seven
days
that
may
have
been
impacted,
only
three
of
those
days
had
speciation
data
available.
The
sulfates
on
those
three
days
ranged
from
12
to
15
µ
g/
m3
while
the
organic
carbon
(
a
wildfire
marker)
ranged
from
5
to
9
µ
g/
m3.
Neither
of
these
ranges
was
unusual
as
compared
to
any
other
summer
day
with
high
values.
Wildfires
are
not
the
only
source
of
organic
carbon.
Chattanooga
also
used
potassium
as
a
wildfire
marker.
The
use
of
potassium
has
been
questioned
by
EPA
scientists,
but
even
if
it
were
used,
the
potassium
levels
were
not
any
higher
on
a
percentage
basis
on
these
alleged
event
days
than
other
days
with
high
values.
Since
the
speciation
data
did
not
support
Chattanooga's
request,
we
determined
the
data
to
be
inconclusive.
It
is
more
plausible
to
believe
that
these
days
were
typical
summer
days,
high
temperatures
resulting
in
the
conversion
of
SO2
to
sulfates.
If
one
assumes
that
the
sulfates
and
nitrates
were
ammonium
sulfates
and
nitrates,
their
contribution
would
be
even
greater
than
the
ranges
given
above.

The
supporting
data
provided
by
Chattanooga
to
qualify
the
elevated
and/
or
exceedance
measurements
as
exceptional
events
is
neither
sufficient
nor
conclusive
for
this
determination.
The
frequency
of
fires,
the
distant
locations
for
the
fires,
and
the
lack
of
specific
detailed
consequence
analysis
for
each
firemeasurement
event
make
the
provided
justification
insufficient
and/
or
inconclusive
to
exempt
the
measured
data
as
exceptional
fire­
caused
events.
Additional,
more
detailed
consequence
specific
information
is
needed
to
make
this
determination.
The
new
information
in
the
November
4,
2004,
Chattanooga­
Hamilton
County
Air
Pollution
Control
Bureau
letter
does
not
change
the
conclusions
provided
in
EPA's
December
1,
2003,
memorandum
on
the
original
request.
The
evidence
provided
is
insufficient
to
conclusively
support
the
request
to
define
the
April,
June,
and
August
2003
events
as
exceptional
because
of
the
influence
of
distant
agricultural
and
wild
fires.
Additional
detailed
analyses
and
information
are
needed
to
support
this
exceptional
event
request.
See
EPA's
November
30,
2004,
memorandum,
and
forward
and
back
trajectories,
for
detailed
information.

The
information
submitted
by
Chattanooga
in
support
of
their
request
for
the
June,
July,
and
August
2004
events
was
inadequate.
Among
the
problems
with
their
request
are:
the
trajectory
analyses
were
done
at
such
high
levels
of
the
atmosphere
that
mixing
of
fire
emissions
with
ground
level
air
was
highly
improbable;
there
was
no
comprehensive
analysis
of
the
speciated
air
quality
data
in
the
Chattanooga
area
and
receptor
modeling
techniques
were
not
used
to
try
and
identify
the
sources
of
the
PM2.5
mass
in
the
area;
and
there
was
no
assessment
of
the
impact
of
regional
and
local
sources
of
emissions
on
PM2.5
concentrations
in
Chattanooga.

EPA
includes
the
following
documents
in
support
of
the
decision
for
the
Chattanooga
area:

3.
B.
Chattanooga
design
value
analysis
3.
C.
Memorandum
from
Stanley
Krivo,
EPA
Region
4,
to
Richard
Guillot,
EPA
Region
4,
Regarding
Exceptional
Events
for
Exceedances/
Elevated
Ozone
and
PM2.5
Measurements,
Jefferson
County,
AL
and
Chattanooga­
Hamilton
County,
TN;
December
1,
2003
Office
Memorandum
Air
Quality
Modeling
and
Transportation
Section
To:
Richard
Guillot
Information:
Scott
Davis
Rick
Gillam
Brenda
Johnson
From:
Stan
Krivo
Date:
01
December
2003
Subject:
Exceptional
Events
for
Exceedances/
Elevated
Ozone
and
PM2.5
Measurements
(
Jefferson
County,
AL
and
Chattanooga­
Hamilton
County,
TN)

The
following
are
my
review
comments
on
the
justification
provided
to
exempt
the
monitored
measurements
of
ozone
and/
or
PM2.5
because
measurements
are
considered
exceptional
events.

October
2000
for
Jefferson
County,
AL
1.
Time
Series
Measurements
­
The
provided
measurements
of
PM2.5
for
the
Wylam
and
N.
Birmingham
monitors
reveals
similar
pattern
of
measurements
for
21­
28
October
2000.
These
measurements
do
not
appear
to
be
outliers.
If
the
ozone
8­
hour
measurements
follow
the
same
pattern
as
the
24­
hour
PM2.5,
the
exceedance
measurements
of
concern
will
also
not
be
outliers.

2.
Fire
Locations
­
The
surface
winds
for
the
dates
of
concern
show
very
little
transport
so
only
local
fires
could
contribute
to
the
concentration
measurements.
The
specific
location
of
the
fire
and
the
start/
stop
dates
and
times
were
not
provided
to
relate
to
the
time
series
measurements.
To
determine
the
affect
of
the
fires
on
the
measurements,
the
total
time
series
of
measurements
for
all
Jefferson
monitors
should
be
review
for
the
period
when
the
fires
were
occurring.

Based
on
the
information
provided,
only
local
fires
could
possibly
affect
the
measurements
of
concern.
More
specific
information
on
the
fire(
s)
location,
start
time
and
end
time
are
needed
to
relate
the
fire
emissions
to
measurements
of
ozone
and
PM
in
the
Birmingham
area.
The
provided
supporting
information
is
not
sufficient
nor
conclusive
enough
to
eliminate
the
elevated/
exceedance
measurements.
2002/
2003
for
Chattanooga­
Hamilton
County,
TN.

1.
Number
of
Exception
Events/
Region
of
Concern
­
It
appears
that
every
elevated
or
exceedance
measurement
of
ozone
or
PM
is
being
exempted
based
on
the
potential
that
emissions
from
fires
could
have
contributed
to
the
concentrations.
Because
the
location
of
the
fires
range
from
the
local
county
to
northern
Canada,
Minnesota,
to
Mexico,
it
is
likely
that
a
fire
would
have
occurred
somewhere
in
this
large
region
during
the
period
of
concern.
Therefore,
it
is
most
important
that
the
transport
mechanism
exist
and
the
resultant
contribution
from
the
fires
be
large
enough
to
significantly
impact
the
measured
values.

2.
Back
Trajectories
­
The
back
trajectory
calculations
are
used
to
show
that
the
transport
mechanism
exists
during
the
period
of
elevated
measurement.
Given
the
transport
mechanism
exists,
the
fire's
emissions
could
contribute
to
the
measured
concentration.
Back
trajectories
calculations
were
not
performed
in
a
consistent
manner
for
each
event.
It
appears
that
the
only
justification
needed
to
show
that
a
fire
contributed
to
an
observed
elevated
measurement
is
that
a
back
trajectory
calculation
from
any
atmospheric
level
must
past
near
the
location
of
a
fire
during
some
period
near
the
time
of
the
measurement.
The
atmospheric
levels
used
in
the
back
trajectory
calculations
range
from
the
surface
to
5,000
meters.
It
should
be
noted
that
even
give
this
broad,
liberal
criteria,
the
provided
trajectories
for
some
events
still
do
not
past
close
enough
to
the
fire(
s)
to
support
the
conclusion
that
transport
of
fire
emissions
to
the
monitor
is
possible.

3.
Concentrations
­
The
back
trajectories
and
the
fire
maps
with
the
location
of
possible
smoke
plumes
are
not
detailed
enough
to
provide
conclusive
transport
information
and
provide
no
information
of
the
magnitude
of
the
potential
contribution.
Given
the
large
distances
that
the
fire
emissions
must
travel
to
reach
the
location
of
concern,
the
magnitude
of
the
fire
plume's
concentrations
must
be
small.

4.
Routine
Fires
­
The
fires
in
KS
and
OK
that
are
indicated
to
have
affected
the
April
2003
measurements
in
Chattanooga­
Hamilton
County
are
annual
events.
These
same
fires
should
have
caused
problem
measurements
in
the
past
but
the
report
indicates
that
since
1990
no
other
year's
measurements
were
a
problem.
One
exceedance
in
this
period
(
on
04/
25/
98)
was
noted
and
it
was
attributed
to
fires
in
Mexico.
The
annual
nature
of
the
fires
and
the
lack
of
past
impacts
to
the
measurements,
along
with
the
large
distances
between
the
fires
and
Chattanooga­
Hamilton
County,
bring
into
question
the
source
as
well
as
the
magnitude
of
concentration
contributions
associated
with
the
KS/
OK
fires.

5.
Time
Series
Measurements
­
To
support
the
request
for
exemption,
seasonal
time
series
plots
of
all
measurements
should
be
provided
to
demonstrate
that
the
requested
values
are
outliers
from
the
rest
of
the
measurements
and
that
their
large
magnitudes
are
caused
by
the
noted
fires.
Should
the
time
series
plots
show
that
the
requested
elevated
concentrations
or
exceedances
are
within
the
normal
range
of
measurements,
than
the
events
may
not
exceptional
events.

In
summary,
I
believe
the
supporting
data
provided
to
qualify
the
elevated
and/
or
exceedance
measurements
as
exceptional
events
is
not
sufficient
nor
conclusive
for
this
determination.
It
appears
from
the
frequency
of
fires,
the
distant
locations
for
the
fires,
and
the
lack
of
specific
detailed
consequence
analysis
for
each
fire­
measurement
event
make
the
provided
justification
insufficient
and/
or
inconclusive
to
exempt
the
measured
data
as
exceptional
fire­
caused
events.
Additional,
more
detailed
consequence
specific
information
is
needed
to
make
this
determination.

Please
let
me
know
if
you
have
any
questions.

3.
D.
Memorandum
from
Stanley
Krivo,
EPA
Region
4,
to
Richard
Guillot,
EPA
Region
4,
Regarding
2003
Exception
Events
for
Exceedances/
Elevated
Ozone
and
PM2.5
Measurements,
Chattanooga­
Hamilton
County
Air
Pollution
Control
Bureau
(
APCB)
November
4,
2004
Letter;
December
2,
2004
Office
Memorandum
Air
Quality
Modeling
and
Transportation
Section
To:
Richard
Guillot
Information:
Joel
Hansel
Rick
Gillam
Brenda
Johnson
From:
Stan
Krivo
Date:
02
December
2004
Subject:
2003
Exception
Events
for
Exceedances/
Elevated
Ozone
and
PM2.5
Measur
ements
Chattanooga­
Hamilton
County
Air
Pollution
Control
Bureau
(
APCB)
November
4,
2004
Letter
The
following
are
my
review
comments
on
the
additional
information
provided
in
the
referenced
APCB
letter
to
justification
the
exemption
of
three
2003
periods
of
monitored
ozone
and
PM2.5
measurements
because
they
are
considered
exceptional
events.
The
original
December
2003
exemption
request
included
additional
periods.

6.
Exception
Events/
Region
of
Concern
­
Three
events
during
2003
with
elevated
or
exceedance
measurements
of
ozone
or
PM2.5
are
requested
for
exemption
based
on
the
belief
that
emissions
from
distance
fires
caused
or
significantly
contributed
to
the
measured
concentrations.
The
three
periods
are:

­
Ozone
April
12,
14,
and
15
PM2.5
April
15
­
Ozone
June
24,
25,
26
PM2.5
June
26,
29
­
Ozone
August
26
PM2.5
August
19,
22,
25,
28
It
appears
that
the
selected
days
for
exemption
were
based
on
the
elevated
magnitude
of
the
measured
concentrations
(
e.
g.,
top
10
measurements
during
year
or
values
exceeding
the
standards).

For
all
the
events
only
distant
fires
were
noted
as
significant
reasons
for
the
elevated
measurements.
Review
of
more
local
causes
for
these
measurements
were
not
indicated
to
have
been
performed.
To
understand
and
more
conclusively
attribute
the
elevated
measurements
to
these
distant
fires,
the
following
are
suggested
needed
studies
or
information.

­
Dates
and
locations
of
the
identified
controlled
Kansas
fires
and
Canadian
wildfires
to
correlate
with
periods
of
high
measurements
in
region.
­
Identification
of
any
other
Kansas
and
Canadian
fires
during
2003
and
corresponding
Chattanooga
area
ambient
ozone
and
PM2.5
measurements
to
determine
the
uniqueness
of
these
events
to
ambient
Chattanooga
conditions.
­
Duration
of
the
elevated
pollutant
measurements
in
Chattanooga
area
needs
to
be
supported.
The
start
and
end
dates
for
the
burns
were
not
provided.
TOMS
aerosol
observations
during
each
of
these
events
do
not
provided
conclusive
evidence
of
fire
plumes
transportation
to
the
Chattanooga
area.
­
Magnitude
of
the
Kansas
and
Canadian
fires
contribution
to
Chattanooga's
measurements
should
be
considered.
When
comparing
the
ozone
and
PM2.5
measurements,
provided
in
the
new
time
series
plots,
on
either
side
of
the
requested
exemption
periods
to
the
maximum
values
on
the
requested
exception
days,
the
distance
fires
would
have
to
contribute
20
to
40
ppb
to
the
ozone
measurements
and
15
to
30
ug/
m3
to
the
PM2.5
measurements.
Considering
the
large
distance,
it
appears
unlikely
that
this
large
a
contribution
would
come
from
such
a
distant
source.
­
More
local
causes
for
the
identified
elevated
measurements
must
be
investigated
and
eliminated.
For
example,
the
large
power
plants
in
NW
Georgia
and
NE
Alabama
should
be
eliminated
as
possible
cause
for
these
elevated
measurements.
[
Note:
The
TOMS
visual
for
April
event
indicated
large
aerosol
concentrations
in
an
area
of
NE
Alabama/
NW
Georgia,
general
location
of
large
power
plants
­
a
possible
source
of
pollutants
that
could
be
transported
to
the
Chattanooga
area
causing
the
April
elevated
measurements.
The
TOMS
observations
should
be
related
to
the
back
trajectory
analyses
for
a
more
conclusive
argument.]
­
Synoptic
analyses
of
the
weather
events
during
the
identified
exceptional
periods
should
be
provided.
The
synoptic
conditions
along
the
expected
transportation
pathways
during
the
events
would
provide
additional
information
that
would
be
of
value
in
evaluating
the
possibility
of
long
range
transport
of
pollutants
from
controlled
and
wildfire
burns.
­
Other
areas
of
the
SE
should
be
reviewed
for
these
same
exceptional
event
days.
If
plumes
from
the
distant
fires
affected
the
Chattanooga
area,
they
should
have
also
affected
other
measurements
in
the
SE
(
e.
g.,
Knoxville,
Nashville,
and
Atlanta).
Have
these
areas
also
requested
exceptional
events
for
the
identified
days?

7.
TOMS
Observations
­
Review
of
the
TOMS
videos
did
not
conclusively
demonstrate
transport
of
burn
emissions
to
the
Chattanooga
area.
This
is
especially
true
considering
the
TOMS
observations
are
at
10,000
feet
or
more
elevation.
The
analysis
appears
to
assume
that
high
TOMS
concentrations
on
the
days
of
concern
over
SE
TN
are
representative
of
surface
concentrations.
It
also
assumes
that
low
TOMS
concentrations
over
SE
TN
on
the
days
of
concern
just
mean
that
the
fire
plume
is
lower
than
10,000
feet
­
a
can't
lose
situation.
Left
unanswered
is
the
question
of
magnitude
of
the
fire
plume's
contribution
to
the
measurements.

8.
Back
Trajectories
­
Nothing
new
was
provided
on
the
back
trajectory
calculations.
Our
previously
provided
comments
on
the
bask
trajectory
analysis
are
still
applicable.
Back
trajectories
calculations
were
not
performed
in
a
consistent
manner
for
each
event.
It
appeared
that
the
only
justification
needed
to
show
that
a
fire
contributed
to
an
observed
elevated
measurement
is
that
a
back
trajectory
calculation
from
any
atmospheric
level
must
past
near
the
location
of
a
fire
during
some
period
near
the
time
of
the
measurement.

9.
Routine
Fires
­
The
fires
in
KS
and
OK
that
are
indicated
to
have
affected
the
April
2003
measurements
in
Chattanooga­
Hamilton
County
are
annual
events.
These
same
fires
should
have
caused
problem
measurements
in
the
past
but
the
report
indicates
that
since
1990,
no
other
year's
measurements
were
a
problem.
One
exceedance
in
this
period
(
on
04/
25/
98)
was
noted
and
it
was
attributed
to
fires
in
Mexico.
The
annual
nature
of
the
fires
and
the
lack
of
past
impacts
to
the
measurements,
along
with
the
large
distances
between
the
fires
and
Chattanooga­
Hamilton
County,
bring
into
question
this
source
as
the
cause
of
the
elevated
measurement
event.
This
is
especially
true
when
more
local
causes
of
the
elevated
concentrations
were
not
eliminated.

10.
Speciation
Data
­
Graphs
of
speciation
data
were
provided
for
the
2003
ozone
season.
It
was
indicated
that
the
biomass
markers
were
provided
however
there
is
no
discussion
indicating
support
or
non­
support
for
the
fires
causing
the
elevated
measurements
on
the
requested
exceptional
event
days.

In
summary,
I
believe
the
new
information
in
the
4
November
2004
Chattanooga­
Hamilton
County
Air
Pollution
Control
Bureau
letter
does
not
change
the
conclusions
provided
in
my
1
December
2003
memorandum
on
the
original
request.
The
evidence
provided
is
insufficient
to
conclusively
support
the
request
to
define
the
April,
June,
and
August
2003
events
as
exceptional
because
of
the
influence
of
distance
agricultural
and
wild
fires.
Additional
detailed
consequence
specific
analyses
and
information,
such
as
that
suggested
in
item
1
above,
are
needed
to
support
this
exceptional
event
request.

Please
let
me
know
if
you
have
any
questions.
3.
E.
EPA
Review
of
Trajectory
Analysis,
March
29
2005
April
15,
2003
Kansas
Agricultural
Fires
Chattanooga
Tennessee
did
not
provide
any
trajectory
analyses.

Back
trajectory
analysis
was
performed
using
start
heights
of
500
m,
1500
m
and
2468
m.
The
trajectories
indicate
that
except
at
the
top
of
the
mixed
layer,
air
within
the
mixed
layer
over
Chattanooga
came
from
Georgia
and
circled
around
back
through
Tennessee,
Kentucky
and
Illinois
but
did
not
originate
in
Kansas.
The
top
trajectory
indicates
it
could
have
originated
over
Kansas
4
to
5
days
prior
to
April
15.
Fires
over
Kansas
around
April
10,
2003
would
need
to
be
shown
in
order
to
provide
any
evidence
of
an
impact
upon
Chattanooga.

June
26,
2003
Canadian
Fires
from
Western
Ontario
Chattanooga
did
not
provide
any
trajectory
analyses.

Back
trajectory
analysis
was
performed
using
start
heights
of
500
m,
1000
m
and
1908
m.
The
trajectories
indicate
air
within
the
mixed
layer
over
Chattanooga
was
rather
stagnant
and
came
from
the
south
and
southeast
around
Georgia
and
Florida
coastal
areas,
not
from
Canada.

June
29,
2003
Canadian
Fires
from
Western
Ontario
Chattanooga
did
not
provide
any
trajectory
analyses.

Back
trajectory
analysis
was
performed
using
start
heights
of
500
m,
1000
m
and
2020
m.
The
trajectories
indicate
that
air
at
low
levels
of
the
atmosphere
was
nearly
stagnant
and
meandered
around
Alabama
and
Georgia.
However,
near
the
top
of
the
mixed
layer
the
air
was
shown
to
have
come
from
central
Canada.
Although
only
one
trajectory
supports
it,
it
does
indicate
that
smoke
from
the
fires
in
Ontario
could
have
transported
down
to
Tennessee
and
could
have
entrained
into
the
mixed
layer
to
the
surface
in
Chattanooga.

August
19,
2003
Canadian
Fires
Chattanooga
did
not
provide
any
trajectory
analyses.

Back
trajectory
analysis
was
performed
using
start
heights
of
500
m,
1000
m,
and
1679
m.
Trajectories
do
indicate
that
the
air
within
the
mixed
layer
over
Chattanooga
may
have
originated
in
south
central
Canada
3
to
5
days
prior
to
August
19.
However,
there
were
no
satellite
photographs
in
the
supporting
documentation
to
indicate
whether
smoke
was
over
south
central
Canada
or
not
during
that
same
time
period.
August
22,
2003
Canadian
Fires
Chattanooga
did
not
provide
any
trajectory
analyses.

Back
trajectory
analysis
was
performed
using
start
heights
of
500
m,
1000
m
and
1697
m.
The
trajectories
do
not
show
any
evidence
of
originating
in
Canada
within
the
120
hour
run.
They
remain
within
the
southeast
and
midwestern
regions
of
the
U.
S.
Although
TOMS
satellite
photographs
show
smoke
from
Canada
traveling
near
Tennessee,
the
trajectory
evidence
does
not
support
the
smoke
entraining
down
into
the
mixed
layer.

August
25,
2003
Canadian
Fires
Chattanooga
did
not
provide
any
trajectory
analyses.

Back
trajectory
analysis
was
performed
using
start
heights
of
500
m,
1000
m
and
1294
m.
This
analysis
does
provide
evidence
that
smoke
from
the
Canadian
fires
may
have
impacted
Chattanooga.
Trajectories
originate
in
south
central
Canada
3
to
5
days
prior
to
their
potential
impact
with
Tennessee.
The
TOMS
satellite
photographs
indicate
smoke
in
south
central
Canada
at
the
same
location
as
the
trajectories
at
the
same
time
3
to
5
day
period
prior
to
the
potential
impact
over
Tennessee.
It
is
uncertain
whether
the
smoke
over
that
region
was
at
the
same
height
as
the
trajectories
though.

August
28,
2003
Canadian
Fires
Chattanooga
did
not
provide
any
trajectory
analyses.

Back
trajectory
analysis
was
performed
using
start
heights
of
500
m,
1000
m
and
1723
m.
The
trajectories
do
not
show
any
evidence
of
originating
in
Canada.
They
remain
within
the
southeast
and
midwestern
regions
of
the
U.
S.

June
8,
2004
Arkansas
Agricultural
Wheat
Fires
Chattanooga
performed
a
trajectory
analysis
using
FNL
low
resolution
data
which
is
not
the
recommended
data
set
for
this
analysis.
They
used
start
heights
of
7000
m,
6500
m
and
6750
m.
These
start
heights
are
inappropriate
because
they
are
well
above
the
calculated
mixed
layer.

More
appropriate
back
trajectories
were
performed
using
EDAS
high
resolution
data
and
start
heights
of
500
m,
1000
m
and
1834
m.
These
trajectories
show
evidence
against
any
smoke
from
Arkansas
moving
over
Chattanooga
and
affecting
the
mixed
layer.
The
trajectories
come
from
a
southeast
direction
near
the
Georgia
and
Florida
coasts,
not
from
a
westward
direction
from
Arkansas.
June
11,
2004
Arkansas
Agricultural
Wheat
Fires
Chattanooga
performed
a
trajectory
analysis
using
FNL
low
resolution
data
which
is
not
the
recommended
data
set
for
this
analysis.
They
used
start
heights
of
9000
m,
8000
m
and
7000
m.
These
start
heights
are
inappropriate
because
they
are
well
above
the
calculated
mixed
layer.

More
appropriate
back
trajectories
were
performed
using
EDAS
high
resolution
data
and
start
heights
of
500
m,
1500
m
and
2154
m.
These
trajectories
do
not
show
any
evidence
of
originating
in
Arkansas.
They
indicate
that
the
air
meandered
throughout
eastern
Tennessee,
Alabama
and
Georgia
within
3
days
prior
to
June
11.

July
17,
2004
Alaskan
Fires
Chattanooga
performed
a
trajectory
analysis
using
FNL
low
resolution
data
which
is
not
the
recommended
data
set
for
this
analysis.
They
used
start
heights
of
4000
m,
5000
m
and
7000
m.
These
start
heights
are
inappropriate
because
they
are
well
above
the
calculated
mixed
layer.

More
appropriate
back
trajectories
were
performed
using
EDAS
high
reolution
data
and
start
heights
of
500
m,
1000
m
and
1484
m.
These
trajectories
originate
over
south
central
Canada
about
5
days
prior
to
July
17.
TOMS
satellite
data
shows
smoke
from
Alaska
traveling
down
into
south
central
Canada
about
2
to
3
days
prior
to
July
17.
According
to
the
trajectories,
the
timing
appears
to
be
off
to
provide
evidence
that
the
Alaskan
smoke
impacted
Chattanooga.

July
20,
2004
Alaskan
Fires
Chattanooga
performed
a
trajectory
analysis
using
FNL
low
resolution
data
which
is
not
the
recommended
data
set
for
this
analysis.
They
also
used
inappropriate
start
heights
of
3000
m,
7000
m
and
9000
m,
which
are
all
above
the
calculated
mixed
layer
height.

Appropriate
back
trajectory
analysis
was
performed
using
high
resolution
EDAS
data
and
start
heights
of
500
m,
1000
m
and
1834
m.
This
analysis
does
provide
evidence
that
smoke
from
the
Alaskan
fires
may
have
impacted
Chattanooga
Tennessee.
Trajectories
originate
in
south
central
Canada
5
days
prior
to
their
potential
impact
with
Tennessee.
The
TOMS
satellite
photographs
indicate
smoke
from
Alaska
at
the
same
location
in
south
central
Canada
as
the
trajectories
at
the
same
time
5
days
prior
to
the
potential
impact
over
Tennessee.
It
is
uncertain
whether
the
smoke
over
that
region
was
at
the
same
height
as
the
trajectories
though.

August
4,
2004
Alaskan
and
Canadian
Fires
Chattanooga
performed
a
trajectory
analysis
using
FNL
low
resolution
data
which
is
not
the
recommended
data
set
for
this
analysis.
They
also
used
inappropriate
start
heights
of
2000
m,
3000
m
and
4000
m,
which
are
all
above
the
calculated
mixed
layer
height.
Appropriate
back
trajectory
analysis
was
performed
using
start
heights
fo
500
m,
1000
m
and
1516
m.
Comparing
these
trajectories
with
the
satellite
photographs
does
indicate
that
the
smoke
from
the
Alaskan
and
Canadian
fires
could
have
impacted
Chattanooga.
The
trajectories
intersect
the
smoke
on
the
photographs.
There
is
some
uncertainty
about
the
height
of
the
smoke
and
whether
it
was
at
the
same
levels
as
the
trajectories.

August
10,
2004
Alaskan
and
Canadian
Fires
Chattanooga
performed
a
trajectory
analysis
using
FNL
low
resolution
data
which
is
not
the
recommended
data
set
for
this
analysis.
They
also
used
one
inappropriate
start
height
of
4000
m,
which
is
above
the
calculated
mixed
layer
height.

More
appropriate
back
trajectory
analysis
was
performed
using
start
heights
of
500
m,
1500
m
and
2138
m,
all
within
the
calculated
mixed
layer
height.
The
trajectories
do
not
provide
evidence
that
smoke
from
the
Alaskan
and
Canadian
fires
impacted
Chattanooga
based
on
the
satellite
photographs
provided.
The
day
before,
on
August
9,
the
trajectories
meandered
to
the
south
and
east
of
Chattanooga
when
the
satellite
photographs
indicate
the
smoke
was
north
and
west
of
Chattanooga
that
day.
Satellite
photographs
show
the
smoke
moving
across
the
state
of
Tennessee
from
the
northwest
to
the
southeast
which
appear
to
be
more
indicative
of
the
winds
at
higher
heights
above
the
mixed
layer.
There
is
some
uncertainty
that
the
trajectories
could
have
intersected
the
smoke
several
days
before
since
they
came
from
the
north,
but
satellite
photographs
were
not
provided
for
the
previous
days
so
it
could
not
be
verified.

August
16,
2004
Alaskan
and
Canadian
Fires
Chattanooga
performed
a
trajectory
analysis
using
high
resolution
EDAS
data.
They
used
an
inappropriate
start
height
of
6000
m,
well
above
the
mixed
layer.
Appropriate
back
trajectory
analysis
was
performed
using
start
heights
of
500
m,
1000
m
and
1784
m.
One
trajectory
does
originate
from
south
central
Canada
but
two
trajectories
do
not
and
they
remain
in
the
southeast
region
of
the
U.
S.
This
indicates
that
smoke
from
Canada
could
have
transported
south
to
Tennessee
although
only
one
trajectory
at
one
level
supports
it.

August
19,
2004
Alaskan
and
Canadian
Fires
Chattanooga
performed
a
trajectory
analysis
using
FNL
low
resolution
data
which
is
not
the
recommended
data
set
for
this
analysis.
They
used
start
heights
of
1000
m,
2000
m
and
3000
m.
The
3000
m
start
height
is
inappropriate
because
it
is
above
the
calculated
mixed
layer.
The
3000
m
start
height
trajectory
is
the
only
trajectory
that
originated
in
Canada.
The
other
lower
level
trajectories
remained
in
the
southern
U.
S.

More
appropriate
back
trajectory
analysis
was
performed
using
start
heights
of
500
m,
1000
m
and
2234
m.
These
trajectories
were
also
performed
using
high
resolution
EDAS
data.
These
trajectories
do
not
provide
any
evidence
of
smoke
transport
from
Canada
or
Alaska
and
the
trajectories
remained
in
the
south
and
central
regions
of
the
U.
S.
3.
F.
PowerPoint
Presentation
by
Venkatesh
Rao,
EPA
Office
of
Air
Quality
Planning
and
Standards,
Regarding
Speciation
Data
Analysis
for
Chattanooga;
April
4,
2005
Section
4.
Columbus,
GA­
AL
spatial
averaging
proposal
and
request
to
change
status
to
attainment.

EPA
evaluates
requests
for
spatial
averaging
on
a
case
by
case
basis,
in
light
of
the
particular
facts
and
circumstances
in
each
area.
The
general
regulatory
requirements
for
spatial
averaging
are
set
forth
in
40
CFR
Part
58
Appendix
D.
There
are
three
basic
technical
requirements
for
spatial
averaging,
all
of
which
must
be
met
for
spatial
averaging
to
be
appropriate
and
approvable
for
a
given
geographic
area:

1.
Monitor
site
annual
means
need
to
be
+/­
20%
of
spatial
annual
mean
2.
Monitor
sites
should
show
"
similar
day­
to­
day
variability"
e.
g.
0.60
correlation)
3.
Monitor
sites
should
reflect
impacts
by
the
same
types
of
emissions
sources.

The
purpose
for
these
requirements
is
to
insure
that
the
monitor
network
is
appropriate
for
consideration
for
spatial
averaging
and
properly
reflects
the
ambient
conditions
within
the
area.
More
specific
guidance
concerning
spatial
averaging
is
provided
in
Evaluating
Network
Adequacy
for
Spatial
Averaging,
Guidance
For
Network
Design
and
Optimum
Site
Exposure
For
PM2.5
And
PM10,
December
1997;
and
Attachment
C
of
the
Guideline
on
Data
Handling
Conventions
for
the
PM
NAAQS,
EPA­
454/
R­
99­
008,
April
1999.

In
accordance
with
40
CFR
Part
58
and
applicable
monitoring
guidance,
EPA
has
performed
a
detailed
review
of
the
GA
and
AL
spatial
averaging
plans
submitted
by
the
States
and
the
data
submitted
to
EPA's
Air
Quality
System
(
AQS)
for
the
Columbus
GA­
AL
area.
Based
upon
consideration
of
a
number
of
factors,
EPA
has
decided
to
approve
spatial
averaging
for
this
area.
Using
spatial
averaging
and
data
from
the
three
year
period
from
2002­
2004,
EPA
has
also
determined
that
this
area
is
in
attainment
with
the
PM
NAAQS.
The
factors
considered
by
EPA
are
discussed
below.

Network
Design
°
In
July
2004,
the
States
of
AL
and
GA
submitted
to
EPA
a
spatial
averaging
plan
covering
three
monitors
in
the
Columbus
GA­
AL
area.
Agencies
from
both
States
held
a
joint
public
hearing
to
meet
the
public
hearing
requirements
of
the
applicable
EPA
regulations
and
received
no
adverse
comments
on
that
spatial
averaging
plan.
The
states
received
supportive
comments
from
the
US
Fish
and
Wildlife
Service
and
a
private
citizen.
EPA
denied
the
July
2004
spatial
averaging
request
because
EPA
concluded
that
one
of
the
three
monitors,
located
outside
the
city
center,
did
not
properly
represent
the
same
emission
sources
as
the
other
two
monitors.

°
In
December
2004,
the
States
submitted
a
revised
spatial
averaging
plan
that
includes
only
the
two
urban
core
monitors.
The
States
conducted
a
public
hearing
on
the
revised
spatial
averaging
plan
on
March
17,
2005
and
no
comments
were
received.
°
Additional
data
from
monitors
in
Muscogee
County(
the
GA
portion
of
Columbus),
which
consists
of
an
additional
FRM,
a
speciation
and
a
continuous
monitor,
all
provide
data
that
meet
the
NAAQS
and
therefore
tend
to
support
an
attainment
classification
for
the
area.

°
The
two
monitors
included
in
the
December
2004
spatial
averaging
plan
are
both
located
in
the
urban
core
of
the
city
and
are
separated
by
a
distance
of
only
1.7
km
(
1.1
miles).
The
monitors
are
part
of
a
multi­
state
PM2.5
network,
with
one
monitor
located
in
Alabama
and
the
other
located
in
Georgia.
In
accordance
with
the
applicable
PM2.5
regulations,
PM2.5
monitors
which
are
used
to
make
comparisons
to
the
annual
PM2.5
NAAQS
must
be
of
a
neighborhood
scale,
which
typically
represent
areas
from
0.5
­
4
km
in
diameter.
Given
the
close
proximity
of
the
two
monitors
in
the
spatial
averaging
plan,
EPA
believes
that
the
two
monitors
in
the
December
2004
plan
represent
the
same
neighborhood.
Consideration
of
the
location
of
the
two
monitors
thus
tends
to
support
a
spatial
averaging
approach.

Annual
average
concentrations
at
the
monitors
located
in
Phenix
City,
AL
and
Columbus,
GA
for
2002­
2004
As
part
of
evaluating
the
December
2004
spatial
averaging
plan
for
this
area,
EPA
examined
the
data
from
the
two
monitors
to
determine
whether
they
meet
the
applicable
criteria.

The
table
below
presents
the
two
sets
of
data.
It
should
be
noted
that
the
second
set
excludes
one
reported
data
value
of
0.8ug/
m3
for
June
22,
2002
at
the
Phenix
City
site.
EPA
has
determined
that
the
data
value
for
this
day
erroneous
and
must
be
invalidated,
because
it
is
plainly
inconsistent
with
other
ambient
readings
in
the
region
for
this
date.
The
annual
average
results
are
the
same
after
rounding
concentrations
to
the
nearest
tenth.
On
the
day
of
this
Federal
Register
notice,
the
June
22
data
value
was
in
EPA's
Air
Qualilty
Subsystem.

EPA
notes
that
the
annual
average
concentrations
at
Phenix
City
during
2002­
04
are
consistently
higher
than
Columbus.
The
concentrations
are
also
higher
in
2004
at
both
locations.
(
See
graphics
below.)
However,
EPA
believes
that
these
variations
in
the
monitor
data
may
be
a
consequence
of
better
monitoring
performance
during
2004
after
sampler
problems
were
corrected
as
discussed
below.
Thus
the
2004
data
particularly
at
Phenix
City
may
be
more
representative
of
current
conditions
than
the
previous
2
years.

EPA's
evaluation
of
the
data
from
these
monitors
for
2002­
2004
indicates
that
spatial
averaging
is
appropriate.
The
annual
monitored
concentrations
at
the
sites
in
Phenix
City
AL
and
Columbus
GA
are
within
+/­
5
%
of
the
annual
spatial
average.
The
3­
year
design
value
is
within
+/­
2%.
Both
of
these
relationships
are
well
within
the
regulatory
requirement
of
+/­
20%.
EPA
does
not
believe
that
the
malfunctions
at
one
of
the
monitors
or
the
variations
reflected
in
the
data
affect
this
conclusion.
13.0
13.5
14.0
14.5
15.0
15.5
16.0
16.5
2002
2003
2004
Phenix
City
Columbus
Figure
1.
Quarterly
and
annual
average
concentrations,
2002­
2004
at
Phenix
City
AL
and
Columbus
GA
monitor
sites.

Review
of
Monitoring
changes
and
reported
concentrations
During
2001­
2002,
the
Phenix
City,
AL
monitor
had
operational
problems.
At
this
monitor,
28
of
92
potential
samples
were
missing
or
had
very
low
values
during
2001
Q4,
2002
Q1
and
2002
Q2.
Consequently,
the
PM2.5
sampler
was
replaced
4
times
during
this
period.

After
state
certification
and
submission
of
the
data
to
EPA,
the
State
deleted
7
measurements
(
with
abnormally
low
readings
ranging
from
0
to
1.8
ug/
m3)
which
it
determined
to
be
invalid.
As
indicated
above,
there
is
still
one
data
value
(
June
22,
2002)
in
the
Phenix
City
data
set
on
AQS
with
concentration
of
0.8
ug/
m3.
This
measured
value
is
highly
unlikely
to
occur
during
the
summer
when
PM
2.5
concentrations
are
regionally
homogeneous
throughout
large
areas
of
the
southeastern
U.
S.
at
far
higher
levels.
Such
an
abnormally
low
reading
is
indicative
of
an
invalid
measurement
and
should
likewise
be
excluded.

EPA
believes
that
the
remaining
readings
from
these
monitors
are
adequate
to
evaluate
the
spatial
averaging
plan,
and
tend
to
support
approval
of
the
plan.
year
quarter
Quarterly
Correlation
2001
1
0.50
2001
2
0.93
2001
3
0.84
2001
4
0.83
2002
1
0.89
2002
2
0.98
2002
3
0.95
2002
4
0.81
2003
1
0.49
2003
2
0.69
2003
3
0.96
2003
4
0.93
2004
1
0.89
2004
2
0.98
2004
3
0.99
2004
4
0.91
Day­
to­
day
variability
As
required
by
the
monitoring
regulations,
EPA
also
evaluated
whether
the
data
for
the
two
monitors
in
the
December
2004
spatial
averaging
plan
exhibit
similar
day­
to­
day
variability.
Based
on
the
data,
EPA
has
concluded
that
the
correlation
between
the
monitors
for
the
three
years
of
data
from
2002­
2004
is
0.85,
which
is
greater
than
the
correlation
of
0.6
suggested
in
EPA's
regulations.
However,
EPA
notes
that
when
the
data
are
examined
on
a
quarterly
basis,
rather
than
an
annual
basis,
there
were
2
calendar
quarters
during
the
past
4
years
when
the
quarterly
correlation
was
less
than
0.6.
Both
of
these
instances
occurred
during
the
first
calendar
quarter.
(
See
table
below.)

EPA
believes
that
evaluation
of
the
data
on
a
quarterly
basis,
rather
than
an
annual
basis,
is
appropriate
because
ambient
PM2.5
levels
are
typically
dominated
by
regional
emission
sources
during
the
summer
season
(
contributing
to
uniformly
high
urban
concentrations
of
sulfates
and
high
upwind
concentrations
of
carbon).
As
a
result,
summer­
time
concentrations
are
very
similar
throughout
urban
areas,
making
an
annual
evaluation
of
the
correlation
between
monitors
less
indicative
of
local
emissions
impacts.
This
regional
impact
is
less
pervasive
during
winter
periods,
making
winter
readings
potentially
more
reflective
of
local
source
impacts
in
this
area,
and
hence
more
reflective
of
the
correlation
between
the
monitors.
EPA
believes
that
it
is
appropriate
to
consider
quarterly
correlations
as
part
of
evaluating
spatial
homogeneity
of
monitors
in
evaluating
spatial
averaging
plans.
In
this
area,
EPA
notes
that
the
annual
correlation
between
the
monitors
is
high
and
meets
the
suggested
degree
of
correlation
that
is
appropriate
for
spatial
monitoring.
The
quarterly
correlation,
at
least
in
two
winter
quarters,
is
less
than
the
degree
suggested
in
the
regulations.
Nevertheless,
EPA
has
concluded
that
the
degree
of
correlation
between
the
two
monitors
in
the
December
2004
spatial
averaging
plan
is
acceptable,
in
light
of
EPA's
conclusions
with
respect
to
the
other
factors
considered
in
this
analysis.

Influencing
Emissions
EPA
has
also
examined
whether
the
two
monitors
in
the
December
2004
spatial
averaging
plan
are
affected
by
comparable
sources.
EPA
notes
that
the
two
monitors
are
only
1.7
km
(
1.1
miles)
apart,
and
this
tends
to
suggest
that
they
are
probably
affected
by
comparable
sources.
Information
provided
to
the
Agency
by
the
States
and
otherwise
available
to
the
Agency
indicates
that
the
predominant
local
sources
of
emissions
in
this
area
are
related
to
transportation
(
gas
and
diesel
mobile
sources),
and
related
to
commercial/
residential
fuel
combustion,
which
is
predominantly
natural
gas.
Electricity
is
the
remaining
energy
source
for
commercial
and
home
heating/
cooling.
There
is
very
little
wood,
oil
or
coal
combustion
in
the
area.
There
are
few
large
local
stationary
sources
of
emissions.
Based
on
this
universe
of
sources,
and
the
proximity
of
the
monitors,
EPA
believes
that
the
impacts
on
both
monitors
probably
result
from
comparable
sources.

Conclusion
EPA
is
approving
the
December
2004
spatial
averaging
plan
for
the
Columbus
GA­
AL
area,
based
upon
consideration
of
all
of
the
factors
discussed
above.
EPA
has
concluded
that
the
plan
meets
the
basic
regulatory
requirements
for
such
plans
with
respect
to
important
factors
such
as
the
relationship
of
the
annual
means
between
the
monitors,
the
days
to
day
variability
between
the
monitors,
and
the
impacts
from
comparable
sources.
Most
significantly,
EPA
believes
that
the
particularly
close
geographic
relationship
of
the
two
monitors
confirms
that
the
monitors
are
suitable
for
spatial
averaging
because
this
proximity
tends
to
support
the
conclusions
with
respect
to
the
suitability
of
the
network
design
and
the
impacts
of
comparable
sources.
EPA
notes
that
the
evaluation
of
spatial
averaging
plans
must
be
conducted
on
a
case
by
case
basis,
on
the
facts
and
circumstances
of
each
situation.
In
this
instance,
EPA
has
concluded
that
spatial
averaging
is
appropriate.
