Recommendation
for
PM2.5
Designation
Technical
Support
Document
Connecticut
Department
of
Environmental
Protection
Bureau
of
Air
Management
February
2004
ii
Connecticut
Department
of
Environmental
Protection
Recommendation
for
PM2.5
Designation
Technical
Support
Document
Prepared
by
David
Wackter
Michael
Geigert
Randall
Semagin
Kurt
Kebschull
Paul
Bodner
February
2004
iii
TABLE
OF
CONTENTS
Page
Number
Executive
Summary              .   ..  
ES­
1
1.
Introduction               .     .  .
1
2.
Analysis
of
Monitoring
Data
for
Connecticut
PM2.5
NAAQS
Attainment
Boundary
Designations:
Classifying
the
Stiles
Street
Monitor
as
Microscale                  ...
4
A.
Background                      
4
B.
Site
Description                    ..
4
C.
Spatial
Distribution
of
Average
Fine
Particulate
Matter
Concentrations                    ...
5
D.
Fine
Particulate
Matter
Ambient
Air
Trends         .
7
E.
Particulate
Sulfate
Concentrations             .
9
F.
Particulate
Black
Carbon                ...
10
G.
Highway
Traffic
Density ...               
12
H.
Conclusions.                     ..
12
3.
Studies
to
Determine
the
Impact
of
Connecticut's
PM2.5
Emissions
on
NY
and
NJ
                      
14
A.
ISCST3
Area
Source
Modeling
of
Primary
PM2.5       
14
B.
Back
Trajectory
Analyses                ..
19
C.
Interstate
Air
Quality
Rule
Modeling            .
20
D.
Southwest
Connecticut
Commuting
Patterns
to
New
York
and
New
Jersey                      .
22
4.
Outreach
Activities
Undertaken
Regarding
the
Reclassifying
of
the
Stiles
Street
Monitor
to
Microscale            
24
5.
Appropriate
Use
of
Stiles
Street
Monitoring
Data
Based
on
EPA
Regulations
and
Guidance
Documents         .  ...
25
App
Appendix
K
Connecticut
DEP
PM
Monitoring
Network
Plan,
7/
1/
98,
New
Haven
Stiles
Street
iv
List
of
Figures
Figure
No.
Description
1.
PM2.5
Design
Values
in
CT
(
2000­
2002)

2.
Stiles
Street,
New
Haven
Monitoring
Shed,
View
Northwest
3.
Stiles
Street,
New
Haven
I­
95
Access
Ramp,
View
Northwest
4.
Stiles
Street,
New
Haven
Monitoring
Site
Location
5.
New
Haven
Area­
PM2.5
Monitoring
Sites
6.
State
Street,
New
Haven
Monitoring
Site
Location
7.
New
Haven
Area
PM2.5
Design
Values
(
2000­
2002)

8.
New
Haven
Area
PM2.5
Concentrations:
Average
of
3rd
day
samples
9.
State
Street
vs.
Stiles
Street
PM2.5
Concentrations
10.
West
Haven
vs.
Stiles
Street
PM2.5
Concentrations
11.
Mill
Rock
Basin,
Hamden
vs.
Stiles
Street
PM2.5
Concentrations
12.
CT
Agricultural
Station
vs.
Stiles
Street
PM2.5
Concentrations
13.
Woodward
Fire
House
vs.
Stiles
Street
PM2.5
Concentrations
14.
Estimated/
Measured
PM2.5
Concentrations
at
Stiles
Street
15.
Estimated/
Measured
PM2.5
Concentrations
at
State
Street
16.
Annual
Average
of
Estimated/
Measured
PM2.5
Concentrations
17.
Estimated
PM2.5
Trends,
New
Haven
vs.
Non­
New
Haven
Urban
Sites
18.
Measured
PM2.5
Concentrations
at
Mohawk
Mountain,
Cornwall,
CT
19.
PM2.5
Trends
Slopes
vs.
Initial
Concentrations
20.
Rural
vs.
Urban
Areas
Speciation
21.
Fine
Particulate
Sulfate
Concentrations
at
Brigantine
Refuge,
NJ
22.
Fine
Particulate
Sulfate
Concentrations
at
Mohawk
Mountain,
Cornwall,
CT
23.
Fine
Particulate
Sulfate
Concentrations
at
Quabbin
Summit,
Ware,
MA
24.
PM10
Sulfate
Concentrations
at
Stiles
Street,
New
Haven
25.
PM10
Sulfate
Concentrations
at
State
Street,
New
Haven
26.
Fine
Particulate
Total
Elemental
Carbon,
Mohawk
Mountain,
Cornwall
27.
Fine
Particulate
Elemental
Carbon,
Quabbin
Summit,
Ware
MA
28.
Day
and
Night
Black
Carbon
vs.
Wind
Octant
at
Stiles
Street,
New
Haven
29.
Day
and
Night
Black
Carbon
vs.
Wind
Octant
at
State
Street,
New
Haven
30.
Average
Diurnal
BC
Concentrations
and
Average
Hourly
Traffic,
Stiles
Street
31.
Average
Diurnal
BC
Concentrations
and
I­
91
on­
Ramp
Traffic,
State
Street
32.
Comparison
of
Site
Traffic,
Distance
to
Highway,
and
PM2.5
Concentrations
33.
Area
Sources
Used
In
ISC
Modeling
of
Primary
PM2.5
Emissions
34.
State
Primary
PM2.5
Emissions
Contributions
to
Total
PM2.5
Modeled
at
Selected
Receptors
35.
Modeling
Results
at
20
km
Grid
Points
using
LaGuardia
Surface
Meteorology
Data
36.
Modeling
Results
at
20
km
Grid
Points
using
Bridgeport
Surface
Meteorology
Data
37.
Grid
Point
Data
Represented
as
Concentration
Circles
(
LaGuardia
Meteorology).

38.
Grid
Point
Data
Represented
as
Concentration
Circles
(
Bridgeport
Meteorology).

39.
1994
LaGuardia
Airport
Surface
Met
Data
Wind
Rose
Diagram
40.
1974
Bridgeport
Surface
Met
Data
Wind
Rose
Diagram
41.
Back
Trajectories
on
Highest
10th
Percentile
PM2.5
Days
in
New
York
City
42.
Back
Trajectories
on
Lowest
10th
Percentile
PM2.5
Days
in
New
York
City
ES­
1
Executive
Summary
PM2.5
National
Ambient
Air
Quality
Standard
Recommended
Designations
for
Connecticut
The
United
States
Environmental
Protection
Agency
(
EPA)
promulgated
the
National
Ambient
Air
Quality
Standards
(
NAAQS)
for
fine
particulate
matter
(
PM2.5)
on
July
18,
1997.
The
annual
average
NAAQS
for
PM2.5
is
15
µ
g/
m3
(
micrograms
per
cubic
meter)
and
the
24­
hour
average
NAAQS
is
65
µ
g/
m3.
The
State
of
Connecticut
Department
of
Environmental
Protection
(
CTDEP
or
Department)
designed
a
network
and
began
PM2.5
monitoring
in
1999.

States
are
required
to
provide
EPA
with
recommendations
by
February
15,
2004
for
PM2.5
area
attainment
and
nonattainment
designations
based
on
three
years
of
monitored
data.
EPA
has
indicated
the
presumptive
boundaries
for
nonattainment
areas
will
be
based
on
combined
metropolitan
statistical
areas
(
or
CSA's).
EPA's
presumptive
use
of
CSA
boundaries
would
create
a
multi­
state
nonattainment
area
consisting
of
southwest
Connecticut
plus
southern
New
York
and
northern
New
Jersey,
however,
the
technical
analysis
herein
concludes
that
there
is
no
merit
to
such
outcome.

For
the
three­
year
period
ending
in
2002,
all
of
the
PM2.5
monitoring
sites
in
Connecticut
measured
levels
below
the
annual
and
24­
hour
NAAQS,
except
for
the
Stiles
Street
monitoring
site
in
New
Haven,
which
is
measuring
a
three
year
annual
average
(
or
"
design
value")
of
16.6
µ
g/
m3.

The
Department's
technical
review
concludes
that
the
high
annual
average
PM2.5
concentrations
at
the
Stiles
Street
site
are
the
result
of
"
microscale"
effects
even
though
this
site
is
currently
classified
by
EPA
as
a
middle
scale
site.
As
outlined
in
EPA
guidance,
data
from
sites
considered
microscale
should
not
be
used
to
determine
annual
PM2.5
nonattainment
status.
This
site
is
simply
not
representative
of
community
exposure
to
PM2.5
levels
in
New
Haven.
The
Department
has
also
demonstrated
that
emissions
from
Connecticut
sources
are
not
contributing
significantly
to
measured
nonattainment
in
New
York
City
and
northern
New
Jersey,
so
including
Connecticut
with
this
nonattainment
area
is
neither
technically
justified
nor
necessary
to
effectuate
attainment
in
those
areas.
Therefore,
the
purpose
of
this
document
is
to
demonstrate
the
following:

1)
The
Stiles
Street
New
Haven
PM2.5
monitor
should
be
classified
as
a
microscale
site
and
data
should
not
be
used
for
annual
nonattainment
designation;
and
2)
The
Connecticut
portion
of
the
New
York
City
CSA
should
not
be
included
with
the
New
York
City
nonattainment
area.
ES­
2
Measured
PM2.5
levels
in
Connecticut
are
below
the
PM2.5
NAAQS.

Ambient
PM2.5
levels
in
CT
are
produced
by
a
complicated
combination
of
transported
and
local
pollutants.
The
transported
components
can
travel
hundreds
to
thousands
of
miles
before
being
measured
by
an
air
quality
monitor.
The
local
components
include
emissions
from
vehicles,
industry
and
residences
across
urban
corridors.
We
have
compared
the
measured
levels
against
EPA's
National
Ambient
Air
Quality
Standards
of
15
µ
g/
m3
for
an
annual
average
and
65
µ
g/
m3
for
24­
hour
averages.
The
levels
are
shown
in
the
table
below
and
in
Figure
1.
The
measured
values
are
all
well
below
the
NAAQS
with
the
exception
of
the
16.6
µ
g/
m3
annual
value
at
the
Stiles
Street
monitor
in
New
Haven.
Data
from
Stiles
Street,
because
of
its
microscale
properties,
can
appropriately
be
used
only
for
24­
hr
designations
(
this
is
discussed
in
further
detail
below).
The
annual
PM2.5
values
at
monitors
representative
of
community
exposure
range
from
11.6
µ
g/
m3
in
East
Hartford
to
13.9
µ
g/
m3
at
the
New
Haven
State
Street
site.
Levels
along
the
I­
95
corridor
of
Fairfield
County
are
generally
well
below
the
NAAQS
of
15
µ
g/
m3.
PM2.5
design
values
for
the
24­
hour
averages
range
from
31
µ
g/
m3
in
Hartford
and
Norwich
to
41
µ
g/
m3
at
the
New
Haven
Stiles
Street
site,
well
below
the
24­
hour
NAAQS.

2000­
2002
PM2.5
Design
Values
in
CT
(
µ
g/
m3)
Monitor
Location
Annual
24hr*
Monitor
Location
Annual
24hr*
Bridgeport
Roosevelt
School
13.7
39
New
Haven
Stiles
St.**
16.6
41
Bridgeport
Congress
St.
12.8
35
Norwalk
13.0
35
Danbury
12.9
34
Norwich
11.8
31
East
Hartford
11.6
35
Stamford
13.1
36
Hamden
11.7
33
Waterbury
13.7
34
Hartford
12.6
31
Westport
12.4
34
New
Haven
State
St.
13.9
37
NAAQS
15.0
65.0
*
Design
values
for
24­
hr
levels
are
the
three­
year
average
of
the
98th
percentile
daily
values.
**
Microscale
effects
at
the
Stiles
Street
monitor
preclude
its
use
in
annual
attainment
designations.

The
PM2.5
monitor
at
Stiles
Street
is
strongly
influenced
by
microscale
phenomena
and
should
be
classified
as
a
microscale
monitor.

The
Stiles
Street
monitor
is
located
immediately
adjacent
to
the
southbound
I­
95
on­
ramp
approaching
the
Quinnipiac
River
Bridge.
Consequently,
the
monitor
is
significantly
influenced
by
microscale
phenomena,
particularly
diesel
truck
emissions
from
heavily
loaded
trucks
accelerating
up
the
steeply
graded
on­
ramp
and
approach
to
the
Q­
bridge
(
see
Figures
2
and
3).
The
footprint
of
this
hotspot
is
on
the
order
of
tens
of
meters,
much
smaller
than
a
football
field,
and
does
not
include
residential
areas.
As
such,
it
is
not
representative
of
community
exposure
and,
consistent
with
EPA's
guidance,
should
be
treated
as
a
microscale
site
for
PM2.5
classification
purposes.
Data
from
the
Stiles
Street
site
should
not
be
used
for
annual
average
attainment
designations.
Upon
EPA
Region
I's
recommendation,
DEP
conducted
an
outreach
campaign,
providing
several
presentations
to
various
groups
in
New
Haven
and
around
the
State.
The
intent
was
to
reach
the
regulated
community,
environmental
groups,
and
the
general
public
to
inform
them
that
a
microscale
monitor
classification
could
lead
to
an
attainment
designation.
ES­
3
One
presentation
has
been
televised
statewide
on
a
public
information
cable
network.
The
technical
support
materials
included
in
this
document
provide
details
of
the
Stiles
Street
site
along
with
data
supporting
the
microscale
argument.

PM2.5
emissions
from
CT
sources
are
contributing
insignificantly
to
measured
PM2.5
nonattainment
in
New
York
and
New
Jersey.

The
New
York
City
CSA
includes
much
of
northern
New
Jersey
and
southern
New
York,
plus
Litchfield,
Fairfield
and
New
Haven
Counties
in
Connecticut.
Monitors
in
New
York
City,
NY
and
Union
City,
NJ
are
measuring
annual
PM2.5
levels
above
the
NAAQS
of
15
µ
g/
m3.
Air
quality
in
primarily
rural
Litchfield
County
and
along
the
urbanized
corridor
of
New
Haven
and
Fairfield
Counties
is
well
below
the
level
of
the
NAAQS.
Furthermore,
computer
modeling
of
the
transport
and
dispersion
of
pollutants
conducted
by
both
the
Department
and
EPA
conclude
that
emissions
from
Connecticut
sources
contribute
insignificantly
to
elevated
pollution
levels
in
New
York
City
and
northern
New
Jersey.
Therefore
it
is
appropriate,
according
to
EPA's
guidance,
that
EPA
will
consider
reducing
the
size
of
nonattainment
areas
from
the
presumptive
metropolitan
area
boundaries
provided
that
certain
factors
are
adequately
addressed.
The
factors
to
be
assessed
include
air
quality
levels,
emissions,
population
distribution,
traffic,
growth
patterns,
meteorology,
topography,
jurisdiction,
and
control
programs.
This
technical
support
document
addresses
these
factors
in
more
detail.

Recently
adopted
programs
are
expected
to
provide
significant
air
quality
benefits
to
Connecticut
citizens
regardless
of
the
attainment
designation.

If
any
part
of
Connecticut
is
designated
as
nonattainment
for
PM2.5
it
then
becomes
subject
to
a
number
of
planning
requirements.
These
plans,
in
large
part,
will
only
document
the
effectiveness
of
existing
and
expected
programs.
For
examples,
see
the
table
below.

Existing
and
expected
federal
programs
designed
to
reduce
PM2.5
levels
Tier
1
and
tier
2
vehicle
emission
standards
Low
sulfur
gasoline
fuel
standards
Heavy
duty
diesel
truck
and
bus
engine
standards
Ultra
low
sulfur
diesel
fuel
standards
Non­
road
compression
ignition
and
diesel
engine
standards
Non­
road
fuel
standards
NOx
SIP
call
IAQR
or
equivalent
transport
rule
Planning
efforts,
if
required,
have
high
administrative
costs
and
would
divert
resources
away
from
the
Department's
environmental
goal
of
achieving
additional
PM2.5
pollution
reductions
in
urban
areas
through
local
measures.
The
federal
programs
identified
above
are
already
in
place,
or
expected
to
be
adopted
in
the
near
future,
and
emission
reductions
from
these
programs
will
occur
regardless
of
Connecticut's
planning
efforts.
EPA's
modeling
indicates
significant
air
ES­
4
quality
improvements
will
be
realized
from
the
federal
programs
that
are
already
promulgated
as
well
as
those
recently
proposed.

The
best
use
of
Agency
resources
is
to
support
additional
PM
reductions.

There
is
no
"
bright
line"
below
which
PM2.5
levels
are
healthy.
In
fact,
DEP
provides
daily
forecasts
of
the
Air
Quality
Index
(
AQI),
which
occasionally
fall
in
the
"
unhealthy
for
sensitive
groups"
(
USG)
category
for
PM2.5,
even
though
the
levels
are
less
than
the
NAAQS.
These
levels
need
to
be
reduced
further
to
adequately
protect
sensitive
individuals.
Given
limited
departmental
resources,
an
attainment
designation
would
more
appropriately
provide
the
opportunity
to
focus
its
efforts
on
further
reducing
urban
area
PM2.5
levels
statewide
through
programs
such
as
school
bus
retrofit
initiatives,
widespread
use
of
clean
fuels,
anti­
idling
(
outreach
and
enforcement),
targeted
retrofits
on
diesel
fleets,
opacity
testing,
etc.
DEP
has
already
been
developing
a
number
of
these
programs
and
would
like
to
expand
upon
the
following:

­
Coordinating
with
municipalities
on
school
bus
retrofits,
clean
fuels
and
anti­
idling
efforts;
­
Coordinating
with
Connecticut
Department
of
Transportation
(
CTDOT)
on
opacity
testing,
diesel
truck
retrofits
and
expanded
retrofits
of
off­
road
vehicles,
clean
fuels
and
anti­
idling
requirements
in
construction
contracts;
and
­
Coordinating
with
Connecticut
Department
of
Motor
Vehicles
(
CTDMV)
on
random
roadside
truck
opacity
testing
and
anti­
idling
(
outreach
and
enforcement).

Additionally
the
Department
is
already
pursuing
a
number
of
potential
programs
to
further
reduce
PM2.5
levels
including,
limiting
off­
road
diesel
fuel
and
home
heating
oil
sulfur
levels
to
less
than
500
parts
per
million
(
ppm),
anti­
idling
enforcement
and
CTDMV
targeted
opacity
testing
for
urban
fleets.

Conclusion
In
light
of
the
demonstrated
microscale
characteristics
of
the
New
Haven
Stiles
Street
site
and
measured
compliance
with
the
PM2.5
standards
throughout
the
state,
EPA
should
designate
the
entire
State
of
Connecticut
as
attainment
for
the
PM2.5
NAAQS.
A
statewide
attainment
classification
will
free
DEP
from
the
resource
burden
of
fulfilling
numerous
planning
requirements
providing
limited
air
quality
benefits,
thus
allowing
scarce
resources
to
be
committed
to
implementing
several
programs
that
will
provide
earlier
and
more
effective
air
quality
improvements,
especially
in
urban
areas.
1
1.
Introduction
The
United
States
Environmental
Protection
Agency
(
EPA)
promulgated
the
National
Ambient
Air
Quality
Standards
(
NAAQS)
for
fine
particulate
matter
(
less
than
2.5
microns
in
diameter,
or
PM2.5)
on
July
18,
1997.
After
years
of
litigation,
the
standards
were
recently
upheld
by
the
United
States
Supreme
Court.
The
annual
average
NAAQS
for
PM2.5
is
15
µ
g/
m3
(
micrograms
per
cubic
meter)
and
the
24­
hour
average
NAAQS
is
65
µ
g/
m3.
The
State
of
Connecticut
Department
of
Environmental
Protection
(
DEP
or
the
Department)
designed
a
network
and
began
PM2.5
monitoring
in
1999.

Pursuant
to
section
107(
d)(
1)
of
the
Clean
Air
Act
(
CAA)
and
section
6102(
c)
of
the
Transportation
Equity
Act
for
the
21st
Century
(
as
modified
by
Sec
425(
6)
of
the
Consolidated
Appropriations
Bill,
H.
R.
2673,
signed
by
the
President
on
January
23,
2004),
Governors
are
required
by
February
15,
2004
to
submit
recommendations
regarding
attainment
designations
and
geographic
boundaries
for
the
PM2.5
NAAQS.
Following
states
recommendations,
EPA
is
required
to
promulgate
designations
by
December
31,
2004.

Section
107(
d)(
1)
of
the
CAA
requires
areas
to
be
designated
nonattainment
if
they
do
not
meet
the
NAAQS
or
if
they
contribute
to
ambient
air
quality
in
a
nearby
area
that
does
not
meet
the
standard.
EPA
guidance
further
recommends
that
metropolitan
areas
(
identified
by
the
federal
Office
of
Management
and
Budget
based
on
U.
S.
Census
Bureau
data)
serve
as
the
presumptive
boundaries
for
PM2.5
nonattainment
areas.
EPA
recommends
that
the
presumptive
boundaries
for
nonattainment
areas
be
metropolitan
statistical
areas
(
MSA's)
or
combined
statistical
areas
(
CSA's).
For
example,
parts
of
Connecticut
have
been
included
with
southern
New
York
and
northern
New
Jersey
with
respect
to
the
1­
hour
ozone
nonattainment
designation.

EPA
will
allow
deviations
from
these
boundaries
provided
that
a
number
of
factors
are
adequately
addressed.
As
described
briefly
below
and
in
more
detail
later,
available
air
quality
data
indicate
that
the
citizens
of
Connecticut
are
not
being
exposed
to
PM2.5
levels
above
the
NAAQS
and
emissions
from
Connecticut
sources
are
not
contributing
significantly
to
measured
PM2.5
nonattainment
in
the
New
York
and
New
Jersey
portions
of
the
New
York
City
CMSA.
Therefore,
all
of
Connecticut
should
be
designated
attainment
with
respect
to
the
24­
hour
average
and
annual
average
PM2.5
NAAQS.
This
variance
from
the
metropolitan
area
boundary
is
consistent
with
the
recommended
boundaries
being
submitted
by
the
States
of
New
York
and
New
Jersey.

For
the
three­
year
period
ending
in
2002,
all
of
the
PM2.5
monitoring
sites
in
Connecticut
are
measuring
levels
below
the
annual
and
24­
hour
NAAQS,
except
for
the
Stiles
Street
microscale
site
in
New
Haven,
which
is
measuring
a
three­
year
annual
average
of
16.6
µ
g/
m3.
Similarly,
PM2.5
values
below
the
level
of
the
NAAQS
have
been
monitored
at
most
locations
in
New
York
and
New
Jersey
with
the
exception
that
a
number
of
monitors
in
New
York
City
and
in
urban
areas
of
northern
New
Jersey
have
recorded
PM2.5
levels
above
the
annual
NAAQS
of
15
µ
g/
m3.

The
Department
believes
that
PM2.5
concentrations
at
the
Stiles
Street
site
in
excess
of
the
annual
PM2.5
NAAQS
are
the
result
of
microscale
effects
even
though
this
site
is
currently
2
classified
by
EPA
as
a
middle
scale
site.
As
a
microscale
site,
data
from
this
monitor
should
not
be
used
to
determine
annual
nonattainment
status.
This
site
is
simply
not
representative
of
community
exposure
to
PM2.5
levels
in
New
Haven.
The
Department
also
believes
that
emissions
from
Connecticut
sources
are
not
contributing
significantly
to
measured
nonattainment
in
New
York
City
and
northern
New
Jersey,
so
including
Connecticut
with
this
nonattainment
area
is
neither
technically
justified
nor
necessary
to
effectuate
attainment
in
those
areas.
Therefore,
the
purpose
of
this
document
is
to
demonstrate
the
following:

1)
The
Stiles
Street
New
Haven
PM2.5
monitor
should
be
classified
as
a
microscale
site;
and
2)
The
Connecticut
portion
of
the
New
York
City
CSA
should
not
be
included
with
the
New
York
City
nonattainment
area.

The
evidence
presented
in
Section
2
of
this
document
suggests
that
unique
characteristics
of
the
Stiles
Street
monitor
location
classify
the
site
as
"
microscale",
meaning
it
is
significantly
affected
by
sources
at
distances
between
10
and
100
meters
and
should
not
be
used
for
determining
compliance
with
the
annual
PM2.5
NAAQS.
Information
presented
in
Section
2
includes
an
examination
of
Stiles
Street
site
characteristics
as
well
as
analyses
of
the
spatial
and
temporal
distributions
of
PM2.5
and
contributing
chemical
species.

Section
3
of
this
document
discusses
three
ambient
impact
analyses
that
support
the
conclusion
that
Connecticut
does
not
significantly
contribute
to
PM2.5
concentrations
in
the
New
York
City
area.
Section
3
also
discusses
commuting
and
transportation
pattern
from
Connecticut
to
New
York
and
New
Jersey.

According
to
the
April
1,
2003
EPA
guidance
memorandum
from
Jeffrey
R.
Holmstead,
EPA
will
consider
a
number
of
factors
in
assessing
whether
to
exclude
portions
of
a
metropolitan
area
from
a
nonattainment
designation.
One
of
the
factors
listed
is
meteorology
(
weather/
transport
patterns).
Section
3
includes
results
obtained
with
the
ISCST3
model,
demonstrating
that
because
of
transport
patterns,
primary
particulate
matter
emissions
have
a
low
impact
on
receptors
in
New
York
City
and
Hudson
County,
New
Jersey.

Maximum
daily
PM2.5
concentrations
for
over
a
nearly
five
year
period
(
January
1999
through
September
2003)
at
P.
S.
59
in
New
York
City
were
also
analyzed
and
rank­
ordered
from
highest
to
lowest.
The
dates
of
the
top
and
bottom
10
percentile
were
obtained.
Back
trajectory
winds
were
run
once
a
day
for
each
of
those
days
at
three
height
levels
(
10m,
500m,
and
1000m
above
ground
level).
All
72
hourly
positions
of
the
model
run
were
saved
for
the
high
and
low
categories.
Results
of
this
study,
also
summarized
in
Section
3,
show
the
air
mass
during
the
dirtiest
days
originated
from
and
passed
through
locations
in
a
sector
from
SSW
and
SW
through
W
and
WNW
from
New
York
City
(
i.
e.,
did
not
pass
over
Connecticut).

In
the
proposed
"
Rule
To
Reduce
Interstate
Transport
of
Fine
Particulate
Matter
and
Ozone
(
Interstate
Air
Quality
Rule)",
Federal
Register
/
Vol.
69,
No.
20
/
published
on
Friday,
January
30,
2004,
EPA
conducted
an
evaluation
of
the
upwind
contributions
to
downwind
PM2.5
nonattainment
In
this
study,
Connecticut
was
found
to
make
an
insignificant
contribution
of
PM2.5
concentration
to
a
downwind
site
in
New
York
City.
Relevant
results
from
EPA's
analysis
are
also
included
in
Section
3.
3
EPA
Region
I
had
recommended
that
Connecticut
DEP
conduct
public
outreach
prior
to
the
reclassification
of
the
Stiles
Street
PM2.5
monitor
to
"
microscale."
Section
4
of
this
document
provides
an
overview
of
several
presentations
that
were
conducted
in
this
regard.

Section
5
presents
a
summary
of
guidance
from
the
EPA
relating
to
PM2.5
monitor
siting
criteria
and
the
NAAQS
designations
for
PM2.5.
4
2.
Analysis
of
Monitoring
Data
for
Connecticut
PM2.5
NAAQS
Attainment
Boundary
Designations:
Classifying
the
Stiles
Street
Monitor
as
Microscale
A.
Background
The
EPA
promulgated
PM2.5
NAAQS
on
July
18,
1997.
The
State
of
Connecticut
designed
a
monitoring
network
and
began
implementing
PM2.5
monitoring
in
1999.
Per
statute,
evaluation
of
the
PM2.5
standards
requires
three
years
of
concurrent
data.
As
such,
a
partial
evaluation
of
three
concurrent
years
of
monitoring
data
was
first
made
for
the
1999­
2001
period.
A
more
complete
three­
year
data
set
was
available
for
2000­
2002.
Among
the
Connecticut
PM2.5
State/
Local
Air
Monitoring
Sites
(
SLAMS
sites)
and
Special
Purpose
sites,
only
the
Stiles
Street,
New
Haven
site
exceeded
the
annual
average
NAAQS.
The
2000­
2002
PM2.5
annual
average
design
value
(
DV)
for
Stiles
Street
was
16.6
µ
g/
m3,
while
the
other
Connecticut
monitors
were
well
below
the
annual
average
NAAQS
with
design
values
ranging
from
11.6
µ
g/
m3
to
13.7
µ
g/
m3
(
Figure
1).
All
Connecticut
monitors
including
Stiles
Street,
New
Haven,
are
well
below
with
the
24­
hour
average
PM2.5
NAAQS.
(
24­
hour
design
values
are
determined
as
the
98th
percentile
of
daily
values
over
a
three­
year
period).

The
timetable
for
implementation
of
the
standards
requires
State
recommendations
for
designation
of
PM2.5
NAAQS
non­
compliance
areas
be
submitted
by
February
15,
2004.
This
section
provides
technical
information
and
data
in
support
of
the
Department's
boundary
designation
recommendations.

The
Department
believes
that
the
high
annual
average
PM2.5
concentrations
at
the
Stiles
Street
site
result
from
both
large­
scale
regional
sources
and
local
sources.
The
regional,
transported
components
can
travel
hundreds
to
thousands
of
miles
before
being
measured
by
an
air
quality
monitor.
The
local
components
can
include
emissions
from
vehicles,
industry,
and
residences
across
urban
corridors.
Local
sources
influencing
the
Stiles
Street
monitor
include:
heavy
duty
diesel
vehicles
accessing
Interstate­
95
(
I­
95)
from
Stiles
Street,
the
high
volume
of
traffic
on
the
interstate
highway,
and
regular
and
frequent
traffic
congestion
due
to
insufficient
carrying
capacity
of
the
Veterans
Memorial
Bridge
over
the
Quinnipiac
River
(
Q
Bridge).
The
evidence
presented
in
this
document
suggests
that
unique
characteristics
of
the
Stiles
Street
monitor
location
classify
the
site
as
"
microscale,"
significantly
affected
by
sources
at
distances
between
10
and
100
meters.
Further,
the
state,
including
the
City
of
New
Haven,
is
embedded
in
a
very
large
region
of
downward­
trending
fine
particle
concentrations,
which
suggests
that
Stiles
Street
PM2.5
concentrations
will
be
lower
than
the
NAAQS
within
a
few
years.

B.
Site
Description
The
Stiles
Street
PM2.5
monitoring
site
is
located
in
a
commercial/
industrial
area
near
I­
95
and
US
Route
1,
at
the
juncture
of
the
Quinnipiac
River
and
New
Haven
harbor.
Photographs
of
the
monitoring
shed
and
the
I­
95
access
ramp
are
given
in
Figures
2
and
3.
The
geographic
coordinates
of
the
site
are
at
latitude
41.2937
°
and
longitude
 
72.9007
°
.
The
PM2.5
sample
intake
is
located
on
the
roof
of
the
monitoring
shed.
The
shed
is
located
on
a
grass
parcel
bounded
by
I­
95
to
the
west,
the
I­
95
southbound
on­
ramp
to
the
north
and
east,
and
Stiles
Street
to
the
south
and
southeast.
The
respective
distances
to
the
bounding
roadways
are
Stiles
Street:
5
less
than
10
meters,
I­
95
access
ramp:
less
than
10
meters
and
I­
95:
less
than
30
meters.
Figure
4
is
a
site
map
showing
the
relationship
of
the
site
to
the
major
roadways
and
the
New
Haven
harbor
terminal
area.
The
site
relationship
is
also
documented
in
the
attached
Appendix,
which
is
"
Appendix
K
Connecticut
DEP
PM
Monitoring
Network
Plan,
7/
1/
98,
New
Haven
Stiles
Street."

Some
of
the
unique
characteristics
that
distinguish
the
Stiles
Street
site
from
other
Connecticut
urban
areas
with
potentially
high
PM2.5
levels
include:
(
1)
extremely
close
distances
to
traveled
roadways,
(
2)
high
volumes
of
heavy
duty
diesel
trucks
serving
the
New
Haven
Terminal
and
other
local
industries
accessing
the
interstate
highway
via
Stiles
Street,
and
(
3)
significant
uphill
grades
on
both
the
I­
95
southbound
entrance
ramp
and
the
subsequent
I­
95
approach
to
the
QBridge
that
requires
vehicles
to
accelerate
under
higher
loads,
producing
greater
particulate
matter
emissions.

In
addition
to
the
Stiles
Street
site,
there
are
several
PM2.5
monitors
within
the
greater
New
Haven
area.
Figure
5
shows
the
seven
current
New
Haven
area
monitors,
including
the
new
station
at
Criscuolo
Park.
Three
monitors,
at
the
West
Haven
toll
booth,
at
the
Connecticut
Agricultural
Experiment
Station,
and
at
the
Woodward
Fire
House,
were
implemented
beginning
the
second
quarter
of
2003
as
special
purpose
monitors
for
this
study.
The
State
Street,
Stiles
Street,
and
Hamden
Mill
Rock
Basins
sites
have
been
used
for
monitoring
PM2.5
since
1999.
Black
carbon,
which
is
considered
an
indicator
of
primary
PM2.5
emissions,
has
been
monitored
at
both
the
Stiles
Street
and
State
Street
sites
during
2003.
A
detailed
view
of
the
State
Street
site
is
provided
in
Figure
6.

C.
Spatial
Distribution
of
Average
Fine
Particulate
Concentrations
The
2000­
2002
PM2.5
design
values
for
24­
hour
and
annual
time
periods
are
presented
in
Table
2.1.
As
demonstrated
by
the
annual
design
values,
the
Stiles
Street
annual
average
PM2.5
concentrations
are
an
anomaly
among
the
Connecticut
monitoring
sites.
Figure
7
shows
the
2000­
2002
PM2.5
design
values
for
the
three
regular
New
Haven
area
sites:
Stiles
Street
and
State
Street,
New
Haven,
and
Mill
Rock
Basins,
Hamden.
The
design
values
indicate
a
steep
PM2.5
concentration
gradient
between
Stiles
Street
and
the
other
two
nearby
sites.
The
Stiles
Street
and
Mill
Rock
Basins
sites
are
5.3
km
(
3.3
mi)
apart,
with
a
difference
in
design
values
of
4.9
µ
g/
m3.
State
Street,
which
is
1.5
miles
from
Stiles
Street
and
approximately
midway
between
the
Stiles
and
Mill
Rock
sites,
has
a
design
value
of
2.7
µ
g/
m3
less
than
Stiles
Street.

As
part
of
this
study,
three
additional
special
purpose
monitors
were
installed
and
operated
within
the
New
Haven
area
starting
in
April
2003.
The
average
PM2.5
concentrations
for
the
period
April
through
September
2003
are
shown
in
Figure
8.
These
values
are
once
every
thirdday
samples
(
intermediate
Stiles
Street
one­
day
samples
are
omitted)
that
include
only
monitoring
days
for
which
there
are
valid
Stiles
Street
values.
The
average
PM2.5
values
for
the
six­
month
period
substantiate
the
evidence
of
a
strong
spatial
gradient
between
Stiles
Street
and
the
other
local
monitors.
The
period
of
record
is
not
representative
of
annual
averages,
as
summertime
PM2.5
levels
tend
to
be
higher
than
annual
averages.
Note
that
the
three
sites
with
values
above
13
µ
g/
m3
(
Stiles
Street,
State
Street
and
the
former
West
Haven
toll
booth)
are
those
that
are
located
less
than
50
meters
to
an
interstate
highway
or
on­
ramp.
The
former
toll
booth
is
within
50
meters
of
I­
95,
but
is
not
within
close
proximity
to
any
access
or
exit
ramps
as
6
at
Stiles
Street,
and
State
Street
is
within
75
meters
of
two
high­
volume
access
ramps,
but
is
about
200
meters
distance
from
I­
91.
Regardless
of
the
proximity
to
the
interstate
highway,
the
Stiles
Street
monitored
levels
are
10
to
15%
higher
than
the
other
near­
highway
sites.
The
Woodward
Fire
House
average
concentration
of
11.9
µ
g/
m3
is
3.8
µ
g/
m3
less
than
the
concentration
measured
approximately
one
half
mile
away
at
Stiles
Street
even
though
this
site
is
only
about
one
quarter
mile
from
I­
95.

Daily
Stiles
Street
PM2.5
data
for
the
six­
month
2003
period
of
record
was
compared
with
the
five
other
New
Haven
area
monitors.
Statistical
plots
are
provided
for
each
site
paired
with
Stiles
Street
in
Figures
9­
13.
The
linear
regression
lines
for
each
pair
of
sites
have
correlation
coefficients
greater
than
0.95,
and
have
slopes
close
to
a
value
of
1.
This
suggests
that
all
of
these
sites
monitor
approximately
the
same
New
Haven
area
background
PM2.5,
and
that
differences
among
sites
result
from
an
additive
local
PM2.5
contributions
that
are
roughly
2­
4
µ
g/
m3
higher
at
Stiles
Street
than
the
other
sites.
This
is
an
indication
of
the
magnitude
of
the
microscale
component
being
measured
at
Stiles
Street.
There
is
minimal
variability
between
Stiles
Street
and
the
other
sites
and
this
is
an
indication
that
the
magnitude
of
the
urban
to
regional
scale
component
of
PM2.5
in
the
New
Haven
area
can
range
from
5
to
50
µ
g/
m3
Table
2.1:
2000­
2002
Annual
and
24­
HourPM2.5
Design
Values
in
Connecticut
Monitor
Location
Annual
Mean
PM2.5
Design
Values
(
µ
g/
m3)
24­
Hour
Average
PM2.5
Design
Values**

(
µ
g/
m3)
Bridgeport
Roosevelt
School
13.7
39
Bridgeport
Congress
Street
12.8
35
(
35)
Danbury
12.9
34
East
Hartford
11.6
36
(
35)
Hamden
11.7
33
Hartford
12.6
31
(
31)
New
Haven
Stiles
Street
16.6*
41
New
Haven
State
Street
13.9
38
(
37)
Norwalk
13.0
35
(
35)
Norwich
11.8
28
(
31)
Stamford
13.1
37
(
36)
Waterbury
13.7
34
Westport
12.4
34
NAAQS
15.0
65
Notes:
*
Microscale
siting
at
Stiles
Street
precludes
its
use
in
annual
attainment
designations.
**
24
hour
design
values
computed
using
98th
percentile
values
only
for
calendar
years
having
75%
quarterly
data
completeness,
per
CFR40
Part
50
App
N.
Parenthetical
values
are
computed
using
98th
percentile
values
from
all
calendar
years
(
2000­
2002).
7
Table
2.1
also
includes
computed
24­
hour
average
PM2.5
design
values.
The
design
values
are
the
arithmetic
average
of
the
yearly
98th
percentiles
of
24­
hour
PM2.5
values
for
three
consecutive
calendar
years
(
i.
e.:
2000­
2002).
Statute
requires
that
each
annual
data
set
has
a
minimum
of
75
percent
data
completeness.
For
the
period
2000­
2002,
approximately
one­
half
of
the
sites
have
at
least
one
year
with
less
than
75
percent
complete
quarters.
As
such,
for
purposes
of
analysis,
24­
hour
design
values
are
presented
in
Table
2.1
computed
two
ways,
(
a)
using
only
annual
98th
percentile
values
from
years
with
four
"
complete"
quarters,
and
(
b)
using
all
three
annual
98th
percentile
values
regardless
of
data
completeness.

The
Connecticut
24­
hour
design
values
are
uniformly
lower
than
the
NAAQS.
This
is
most
likely
due
to
the
nature
of
high
24­
hour
PM2.5
events,
which
tend
to
occur
on
a
regional
scale.
In
contrast,
monitors
that
are
in
close
proximity
to
heavily­
traveled
interstate
highways
or
other
known
sources
show
daily
average
PM2.5
concentrations
that
are
a
few
micrograms
per
cubic
meter
higher
than
regional
or
background
levels,
independent
of
what
those
background
levels
are.

D.
Fine
Particulate
Matter
Ambient
Air
Trends
Available
ambient
air
monitoring
data
supports
the
view
that
fine
particle
concentrations
in
Connecticut
and
the
surrounding
region
are
in
steady
decline
since
the
late
1980'
s.
This
is
observed
from
declining
concentrations
of
both
surrogate
parameters
(
i.
e.
PM10)
and
component
parameters,
such
as
particulate
sulfate.

Monitoring
of
PM2.5
in
Connecticut
by
federal
reference
or
equivalence
methods
commenced
during
the
first
quarter
of
1999.
This
time
scale
of
approximately
four
years
is
not
of
sufficient
length
to
identify
overall
trends
in
PM2.5
concentrations
in
the
state,
due
to
the
high
degree
of
year­
to­
year
variability
caused
by
the
effects
of
meteorology
on
measured
pollutant
concentrations.
Alternative
methods
used
to
assess
longer­
term
fine
particle
trends
in
this
study
include
site­
specific
estimation
using
PM10
as
a
surrogate
and
speciated­
particle
analysis
from
IMPROVE
type
samplers.

Time
series
trend
plots
of
24­
hour
FRM
PM
concentrations
were
developed
to
show
statewide
patterns
for
the
period
1988
to
2003
for
seven
selected
sites.
The
selected
sites
are:

 
Stiles
Street,
New
Haven
(
adjacent
to
I­
95
southbound
on­
ramp);

 
State
Street,
New
Haven
(
adjacent
to
I­
91
northbound
on­
ramp);

 
Roosevelt
School,
Bridgeport
(
adjacent
to
I­
95/
Rt
8
interchange);

 
West
Avenue/
Interstate­
95
(
PM10)
and
Health
Department
(
PM2.5),
Norwalk;

 
Bank
and
Meadow
Streets,
Waterbury
(
adjacent
to
I­
84
eastbound
exit
ramp);

 
Sheldon
Street
(
PM2.5)
and
Capital
Community
Technical
College
(
PM10),
Hartford;

 
McAuliffe
Park
(
PM2.5)
and
High
Street
(
PM10),
East
Hartford.

These
sites
were
selected
for
comparison
primarily
to
provide
a
relatively
well­
distributed
network
of
monitoring
points
over
the
western
Connecticut
region,
with
an
emphasis
on
8
locations
in
close
proximity
to
busy
interstate
highways.
The
sites
were
also
chosen
because
they
had
both
PM2.5
and
PM10
monitoring
either
collocated
at
the
site
or
at
nearby
sites
since
1999.
The
two
New
Haven
sites,
Bridgeport
and
Waterbury
had
collocated
PM10
and
PM2.5
for
the
period
1999
through
2001.

The
Norwalk,
Hartford
and
East
Hartford
sites
had
PM10
and
PM2.5
sampling
at
nearby,
noncollocated
sites.
In
Norwalk,
PM10
was
monitored
at
the
West
Avenue/
I­
95
ramp
site,
and
PM2.5
was
monitored
approximately
0.7
miles
to
the
Northeast
at
the
Norwalk
Health
Department.
In
East
Hartford,
PM2.5
is
monitored
at
McAuliffe
Park,
and
PM10
is
monitored
approximately
3
miles
south
at
High
Street.
Prior
to
1999,
PM10
was
monitored
at
East
Hartford
City
Hall,
about
1.5
miles
southwest
of
McAuliffe
Park.
In
Hartford,
PM2.5
monitoring
takes
place
approximately
25
miles
east
of
PM10
monitoring
at
the
former
Capital
Community
Technical
College
(
CCTC)
on
Flatbush
Avenue.

Using
the
1999­
2001
PM10
and
PM2.5
24­
hour
sample
data
from
the
above
sites,
linear
relationships
for
each
site
(
or
pair
of
sites
if
not
collocated)
were
derived
from
least
squares
linear
regression
curves
for
each
meteorological
season
(
i.
e.:
Dec­
Feb,
Mar­
Jun,
etc.).
Slopes
of
the
linear
functions
generally
ranged
from
0.6
to
0.9,
and
correlation
coefficients
were
in
the
range
of
0.6
to
0.95.
Linear
regression
data
are
summarized
in
Table
2.2.

Table
2.2:
Site­
Specific
PM2.5:
PM10
Regression
Relationship
Data
for
Meteorological
Quarters.

Spring
Summer
Fall
Winter
Site:
m
b
r2
N
m
b
r2
N
m
b
r2
N
m
b
r2
N
Norwalk
0.6371
­
4.4141
0.6904
35
0.7316
­
7.2775
0.6941
43
0.7256
­
5.5492
0.893
29
0.5794
­
2.2313
0.6144
22
East
Hartford
0.4897
1.22
0.5833
41
0.8826
­
2.8929
0.8117
41
0.9453
­
3.2815
0.8361
40
0.3976
3.9913
0.5665
39
Hartford
0.6731
0.3777
0.5965
46
0.9279
­
2.8012
0.8442
40
0.8678
­
1.3419
0.8536
37
0.7529
2.3813
0.7003
42
Waterbury
0.6132
0.0237
0.6022
52
0.8827
­
3.0559
0.8231
57
0.8344
­
2.0919
0.7509
44
0.6664
1.4064
0.6861
51
Bridgeport
0.6571
0.037
0.7538
58
0.9168
3.368
0.9095
53
0.7948
­
1.7872
0.9333
49
0.819
­
0.6144
0.9139
50
New
Haven­
Stiles
St.
0.403
4.008
0.4294
56
0.7875
­
4.2573
0.7155
56
0.6452
­
0.9148
0.683
49
0.5465
2.4873
0.5719
49
New
Haven­
State
St.
0.6919
­
0.7381
0.7733
59
0.9827
­
4.4719
0.9445
54
0.8038
­
1.262
0.9509
48
0.7894
­
0.652
0.8769
51
In
Figures
14
and
15,
the
estimated
and
measured
data
are
combined
to
show
"
PM2.5
trends"
from
1988
to
the
present
for
Stiles
Street
and
State
Street.
The
estimated
and
measured
annual
average
concentrations
are
given
in
Figure
16,
which
also
includes
estimated
PM2.5
concentrations
for
the
Burlington
and
Torrington
sites.
These
concentrations
are
based
on
seasonal
PM2.5
/
PM10
relationships
from
the
nearest
site
having
both
parameters
(
Waterbury).
Note
that
while
average
concentrations
vary
from
year
to
year,
the
differences
between
the
sites
remains
relatively
constant.
All
sites
have
decreasing
concentrations
indicated
by
least
squares
9
linear
regression
lines
with
slopes
ranging
from
 
0.33
µ
g/
m3­
yr
in
East
Hartford
to
 
0.77
µ
g/
m3­
yr
at
Stiles
Street,
New
Haven.
Figure
17
presents
the
estimated
and
measured
combined
annual
average
trends
for
the
two
New
Haven
vs.
five
non­
New
Haven
urban
sites,
comparing
the
downtrending
of
these
two
sectors.

In
addition
to
these
FRM­
based
measurements
and
estimations,
NESCAUM
(
NorthEast
States
for
Coordinated
Air
Use
Management)
sampling
from
1988
through
1993
and
IMPROVE
(
Interagency
Monitoring
of
PROtected
Visual
Environments)
sampling
starting
in
late
2001
show
PM2.5
average
concentrations
at
Mohawk
Mountain,
Cornwall,
gradually
declining
over
this
period
(
Figure
18).
This
rural
site
is
representative
of
regional
background
PM2.5
in
northwest
Connecticut.

Analysis
of
trends
of
the
5th
and
95th
percentiles
of
quarterly
periods
indicates
declining
concentrations
of
the
entire
range
of
data,
supporting
the
validity
of
the
downward
concentration
trends.
In
addition,
the
average
concentration
regression
line
equations
were
used
to
estimate
the
2000­
2002
DVs,
which
are
presented
on
the
figures
together
with
the
calculated
DVs.
The
estimated
and
calculated
DVs
are
in
close
agreement.
It
should
be
noted
that
the
Stiles
Street
estimated
PM2.5
data
may
be
biased
high
for
the
earlier
period
because
of
local
management
practices
implemented
during
the
1990s
to
reduce
levels
of
coarse
particulate
matter
(
i.
e.:
concrete
barricades
were
installed
to
prevent
passage
of
heavy
duty
trucks
across
an
unpaved
lot
that
entrained
dust
and
contributed
to
monitored
PM10
concentrations).

Figure
19
compares
trend
line
characteristics
of
the
seven
sites.
Initial
(
1988)
concentration
is
plotted
against
slope,
showing
that
the
dirtiest
sites
are
improving
at
the
fastest
rate.
The
relationship
between
these
parameters
appears
to
be
close
to
linear
for
these
sites.

E.
Particulate
Sulfate
Concentrations
Particulate
matter
elemental
and
ionic
speciation
data
from
the
EPA
STN
(
Speciation
Trends
Network),
IMPROVE
and
NESCAUM
network
samplers
were
examined
to
provide
information
about
potential
PM2.5
sources
and
concentration
trends.
Results
of
an
EPA
analysis
of
fine
particle
speciation
data,
shown
in
Figure
20,
indicate
that
sulfate
is
a
smaller
fraction,
and
total
carbon
is
a
larger
fraction,
of
fine
particulate
matter
for
urban
sites
compared
to
rural
sites.

Figures
21­
25
present
time
series
sulfate
concentrations
for
three
regional
rural
sites
(
Brigantine
National
Wildlife
Refuge
in
Oceanville,
NJ;
Mohawk
Mountain,
Cornwall,
CT;
and
Quabbin
Summit,
Ware,
MA)
and
two
urban
sites
(
Stiles
Street,
New
Haven,
and
State
Street,
New
Haven).
The
24­
hour
samples
were
collected
on
three
to
six
day
schedules.
The
Brigantine
site
had
a
nearly
continuous
record
of
3­
4
day
interval
monitoring
from
1992
through
first
quarter
2003.
Fine
particulate
sulfate
was
only
monitored
at
Mohawk
Mountain
and
Quabbin
Summit
from
2001­
2003.
However,
PM2.5
samples
collected
from
1988­
1993
were
analyzed
for
total
sulfur,
as
were
the
most
recent
samples.
As
such,
site­
specific
ratios
of
fine
particulate
sulfur
to
sulfate
were
established
from
this
data,
and
fine
particle
sulfate
was
estimated
from
these
ratios
using
the
total
sulfur
data
for
the
earlier
periods.
At
the
two
New
Haven
sites,
the
sulfate
analyses
were
from
PM10
filters,
which
may
result
in
higher
concentrations
if
there
is
significant
coarse
fraction
sulfate.
10
Comparison
of
the
sulfate
data
linear
regression
plots
among
the
five
sites
reveals
similarity
in
both
the
rate
of
concentration
decline
over
time
and
average
concentration
values.
At
the
rural
sites,
average
sulfate
is
declining
at
rates
of
from
0.063
to
0.094
µ
g/
m3/
yr,
while
at
the
New
Haven
sites,
it
is
declining
at
a
rates
of
0.12
and
0.15
µ
g/
m3/
yr
for
Stiles
Street
and
State
Street,
respectively.
Estimated
average
sulfate,
from
the
linear
regression
line
on
January
1,
2002,
ranged
from
about
2.6
µ
g/
m3
at
Quabbin
and
Mohawk
to
3.8
µ
g/
m3
at
New
Haven
and
Brigantine.
These
results
indicate
that
the
New
Haven
area
is
imbedded
in
a
large
region
of
relatively
uniform
sulfate
concentrations,
and
that
these
concentrations
are
declining
across
the
region
uniformly.
The
slightly
lower
sulfate
concentrations
at
Mohawk
and
Quabbin
may
be
due
in
part
to
those
site's
higher
elevations,
which
promotes
a
greater
degree
of
atmospheric
mixing
at
times
when
morning
inversions
cap
pollutants
near
the
surface
at
lower
elevations.

F.
Particulate
Black
Carbon
Ongoing
monitoring
of
particulate
black
carbon
(
BC)
has
been
conducted
in
New
Haven
at
the
Stiles
Street
site
since
December
2002,
and
at
the
State
Street
site
since
April
2003.
At
each
site,
continuous
samples
are
analyzed
using
Magee
Scientific
model
AE2100
aethalometers
that
provide
five­
minute
average
concentrations.
Hourly
and
24­
hour
arithmetic
average
concentrations
are
computed
using
five­
minute
and
validated
one­
hour
values,
respectively.

Black
Carbon,
which
is
emitted
from
many
types
of
combustion
sources,
is
most
notably
associated
with
diesel
fuel
combustion
in
the
absence
of
significant
biomass
combustion.
As
such,
an
urban
site
having
a
high
BC
to
PM2.5
ratio
would
most
likely
be
impacted
by
high
diesel
emissions.
A
comparison
of
the
average
of
the
daily
BC/
PM2.5
ratios
for
the
Stiles
Street
and
State
Street
sites
presented
in
Table
2.3
shows
a
33
percent
higher
average
BC/
PM2.5
ratio
(
0.16
at
Stiles
Street
compared
to
0.12
at
State
Street).
Note
also
that
black
carbon
at
Stiles
Street
is
71%
(
2.28
µ
g/
m3)
higher
than
at
State
Street
(
1.33
µ
g/
m3).

Table
2.3:
Black
Carbon:
PM
2.5
Ratios
for
Stiles
Street
and
State
Street,
New
Haven
4/
2003
through
9/
2003
Stiles
Street
State
Street
BC
PM2.5
BC/
PM2.5*
BC
PM2.5
BC/
PM2.5*

Average
2.28
16.24
0.16
1.33
14.36
0.12
Count
44
44
29
29
44
29
Max
6.16
53.9
0.26
3.15
48.8
0.2
Min
0.56
4.7
0.06
0.4
4.7
0.05
St
Deviation
1.19
10.53
0.06
0.83
10.34
0.04
*
Averages
of
the
ratios
computed
using
only
days
with
valid
BC
and
PM
data
for
both
sites.
11
An
indication
of
regional
background
black
carbon
is
provided
by
time
series
plots
of
total
elemental
carbon
data
from
Mohawk
Mountain
and
Quabbin
Summit
(
Figures
26
and
27).
Average
black
carbon
at
these
sites
is
in
the
range
of
0.3
 
0.4
µ
g/
m3,
a
fraction
of
the
average
black
carbon
in
New
Haven.

The
above
data
suggests
that
New
Haven
black
carbon
is
strongly
dependent
on
local
sources.
To
assess
this
further,
hourly
black
carbon
concentrations
were
compared
to
hourly
average
wind
directions
to
investigate
the
direct
impact
of
potential
local
sources.
Average
black
carbon
concentrations
were
computed
for
each
45
°
wind
direction
octant
for
daytime
(
6AM­
6PM)
and
nighttime
(
6PM­
6AM)
hours.
The
results
for
Stiles
Street
(
Figure
28)
and
State
Street
(
Figure
29)
show
distinct
patterns
between
the
two
sites,
and
between
day
and
night
for
Stiles
Street.

The
Stiles
Street
and
State
Street
nighttime
black
carbon
vs.
wind
direction
patterns
are
similar
to
each
other
with
regard
to
relative
average
black
carbon
maxima
and
minima
occurring
for
certain
wind
directions.
When
winds
are
from
the
south
to
south­
southwest,
nighttime
average
black
carbon
concentrations
are
about
the
same
at
the
two
sites.
At
State
Street,
the
day
and
night
black
carbon
levels
are
similar
for
most
wind
directions,
whereas
at
Stiles
Street,
black
carbon
concentrations
are
significantly
greater
during
daytime
hours
than
nighttime
hours
when
winds
are
blowing
from
the
south­
west­
northeast
sector.
As
can
be
seen
from
the
Stiles
Street
site
map
(
Figure
4),
this
western
segment
includes
the
closest
approach
to
I­
95,
which
encircles
the
site
to
its
west
from
the
south
to
the
north.
The
highest
nighttime
black
carbon
concentrations
are
found
in
the
north­
northeast
to
east­
northeast
directions,
which
are
the
closest
approach
of
the
I­
95
southbound
entrance
ramp.
As
discussed
below,
traffic
count
studies
conducted
for
the
entrance
ramp
indicate
heavy
traffic
for
the
4
AM
to
10
AM
period,
which
includes
some
nighttime
hours.
At
State
Street
(
Figure
6),
the
closest
approach
to
the
I­
91
on
ramp
is
less
than
30
meters
to
the
south,
which
corresponds
to
the
higher
black
carbon
levels
for
wind
directions
from
that
sector.

Figures
30
and
31
show
diurnal
hourly
average
black
carbon
and
average
hourly
traffic
for
the
Stiles
Street
and
State
Street
access
ramps,
respectively.
These
data
(
obtained
from
the
Connecticut
Department
of
Transportation)
show
weekday
traffic
volume
in
the
8:
00
hour
of
about
400
vehicles
for
Stiles
Street,
and
about
330
vehicles
for
State
Street.
However,
average
black
carbon
at
Stiles
Street
is
greater
than
4
ug/
m3
from
7:
00
 
9:
00
A.
M.,
while
at
State
Street
black
carbon
is
about
2.2
ug/
m3
during
this
time.
Factors
that
could
contribute
to
the
higher
observed
black
carbon
at
Stiles
Street
for
similar
ramp
traffic
densities
include:
higher
fraction
of
heavy­
duty
diesel
truck
traffic,
shorter
distance
between
ramps
and
monitors,
and
shorter
distance
of
interstate
highways
to
monitors.

On
Wednesday,
January
21,
2004,
CTDEP
personnel
conducted
traffic
counts
at
the
Stiles
Street
site
for
vehicles
traveling
the
I­
95
southbound
access
ramp
from
5:
00
A.
M.
to
9:
30
A.
M.
(
Table
2.4).
The
counts
were
of
10
to
15
minute
durations,
distinguishing
among
light­
duty,
mediumduty
and
heavy­
duty
vehicles.
Heavy
trucks
(
GVW
>
26,000
lbs)
are
prohibited
from
using
the
ramp
from
7:
00
to
8:
30
A.
M.,
apparently
to
ease
traffic
congestion
and
slowdowns
on
the
Q
Bridge
during
morning
rush
hours.
The
percentage
of
trucks
using
the
ramp
ranged
from
79
percent
in
the
5:
00
A.
M.
hour
to
23
percent
in
the
9:
00
A.
M.
hour,
with
the
exception
of
the
12
prohibition.
During
the
prohibition
period,
the
fraction
of
observed
heavy
vehicles
was
from
3
percent
to
5
percent.

Table
2.4:
I­
95
Southbound
Access
Ramp
Traffic
Count
Data,
Stiles
Street
New
Haven,
January
21,
2004
Start
Time
End
Time
Scaling
Factor
Hourly
LDV
Hourly
MDV
Hourly
HDV
Hourly
Total
Vehicles
%
HDV
5:
21
5:
36
4
12
0
44
56
79
6:
04
6:
19
4
44
4
56
104
54
6:
50
7:
00
6
114
0
60
174
34
7:
15
7:
30
4
276
16
16
308
5
8:
10
8:
25
4
516
8
16
540
3
9:
01
9:
16
4
304
4
92
400
23
G.
Highway
Traffic
Density
The
proximity
of
several
selected
monitoring
sites
to
high­
volume
highways
was
reviewed.
Average
PM2.5
concentrations
were
computed
for
the
period
April
2003
through
September
2003
to
include
the
three
new
monitors
in
the
New
Haven
area
(
Woodward
Street
Fire
House,
West
Haven
Toll
Plaza,
and
CT
Agricultural
Experiment
Station).
For
each
site,
lateral
distance
from
the
monitor
to
the
nearest
major
interstate
or
state
highway
was
determined
from
available
GIS
mapping,
and
average
daily
traffic
for
the
nearest
segment
of
highway
was
obtained
from
Connecticut
DOT
2001
Traffic
Data.
The
values
of
these
parameters
are
provided
in
Figure
32.
The
Stiles
Street
site,
with
the
highest
PM2.5
value,
has
daily
traffic
in
the
higher
range
(
approximately
130,000
vehicles
per
day)
along
with
the
other
New
Haven
area
sites,
Fairfield
County
and
Waterbury.
Stiles
Street
is
also
the
closest
site
to
a
highway
at
less
than
100
meters,
followed
by
West
Haven
and
Norwalk
at
under
200
meters.
Other
factors
that
may
elevate
PM2.5
at
sites
such
as
Hartford,
Bridgeport,
and
State
Street,
are
the
proximity
of
busy
highway
ramps
and
connectors
(
Hartford,
State
Street),
and
additional
nearby
highways
(
Norwalk,
Bridgeport).

H.
Conclusions
A
review
of
the
data
and
analyses
presented
in
this
document
may
be
summarized
in
the
following
points:

 
Although
most
of
the
regular
Connecticut
PM2.5
monitors
are
sited
in
industrial/
urban
areas
in
proximity
to
major
interstate
highways,
only
one
site
(
Stiles
Street,
New
Haven)
exceeds
PM2.5
NAAQS
(
annual
arithmetic
mean).

 
The
Stiles
Street
monitor
location
has
the
local
maximum
average
PM2.5
concentrations,
with
a
strong
decreasing
gradient
extending
out
to
the
surrounding
New
Haven
area
monitors,
indicating
the
existence
of
a
strong
local
source
at
Stiles
Street.

 
A
combination
of
actual
and
estimated
PM2.5
data
in
Connecticut
show
reductions
in
average
PM2.5
for
all
sites
over
the
most
recent
18­
year
period,
with
higher
concentration
13
sites
improving
faster
that
lower
concentration
sites.
Rural
background
PM2.5
from
measured
data
also
shows
declines,
suggesting
that
PM2.5
in
Connecticut
is
derived
from
ubiquitous
sources,
likely
large
power
plants
and
mobile
sources.

 
Observed
sulfate
trends
have
been
declining,
albeit
at
a
slower
rate
than
PM2.5
concentrations,
reflecting
reductions
in
sulfur
emissions
from
acid
rain
programs.

 
Regional
rural
background
elemental
carbon
is
about
0.3­
0.4
ug/
m3,
while
urban
New
Haven
black
carbon
was
4
to
7
times
higher,
suggesting
that
local
sources
are
the
primary
contributor
of
black
carbon.
Higher
ratios
of
black
carbon
to
PM2.5
at
Stiles
Street
compared
to
State
Street
suggest
that
the
Stiles
Street
monitor
is
more
highly
impacted
by
local
diesel
tailpipe
emissions
than
State
Street.

 
Hourly
black
carbon
and
wind
direction
data
indicates
that
wind
direction
has
a
more
significant
impact
on
black
carbon
at
Stiles
Street
than
at
State
Street.
Also,
daytime
black
carbon
was
highest
with
winds
from
the
direction
of
I­
95,
while
black
carbon
was
highest
with
winds
from
the
direction
of
the
access
ramp
during
nighttime
hours
at
Stiles
Street.
The
State
Street
black
carbon
did
not
exhibit
as
strong
a
directional
or
time
of
day
dependence
as
Stiles
Street.

 
Traffic
counts
and
diurnal
black
carbon
data
suggest
a
relationship
between
high­
density
heavy
duty
diesel
truck
traffic
and
morning
black
carbon
maxima
at
Stiles
Street.
Since
black
carbon
is
acknowledged
as
an
indicator
of
primary
PM2.5
emissions,
it
is
plausible
that
the
unusual
nature
of
the
Stiles
Street
monitor
(
i.
e.,
high
volume
of
heavy­
duty
transport
vehicles
serving
harbor
terminal
bulk
fuel
farms
and
other
commodities,
proximity
and
uphill
grade
of
the
area's
major
I­
95
southbound
access
ramp)
are
contributory
to
the
PM2.5
levels
observed
at
the
site.
14
3.
Studies
to
Determine
the
Impact
of
Connecticut's
PM2.5
Emissions
on
New
York
and
New
Jersey
This
section
presents
the
results
from
four
studies
that
will
further
the
case
that
Connecticut
does
not
significantly
contribute
PM2.5
concentrations
to
receptors
in
New
York
and
New
Jersey.
Three
of
the
studies
were
undertaken
by
staff
at
CTDEP,
while
another
study
conducted
by
EPA
for
the
proposed
Interstate
Air
Quality
Rule,
is
referenced
in
this
section.
Taken
together,
these
studies
present
a
weight
of
evidence
leading
to
the
conclusion
that
Connecticut
counties
should
not
be
included
in
the
greater
New
York
City
CSA
non­
attainment
area.
Section
A
describes
ISCST3
(
Industrial
Source
Complex
Simple
Terrain)
area
source
modeling
that
was
used
to
show
State's
contribution
of
PM2.5
to
receptors
in
CT,
NY
and
NJ.
Section
B
shows
the
results
of
back
trajectory
analysis
using
the
HYSPLIT4
(
HYbrid
Single­
Particle
Lagrangian
Integrated
Trajectory)
model.
Section
C
excerpts
a
portion
of
the
proposed
EPA
Interstate
Air
Quality
Rule
technical
support
document
that
shows
Connecticut's
contribution
of
PM2.5
to
NYC.
Section
D
examines
United
States
Census
Bureau
data
describing
southwest
Connecticut
commuting
patterns
to
New
York
City
and
New
Jersey.

A.
ISCST3
Area
Source
Modeling
of
Primary
PM2.5
I.
Introduction
States
must
provide
recommendations
for
area
designations
by
February
15,
2004.
In
the
past,
some
parts
of
Connecticut
were
included
in
the
1­
hour
ozone
nonattainment
area
with
southern
New
York
and
northern
New
Jersey.
According
to
the
April
1,
2003
memo
from
Jeffrey
R.
Holmstead,
EPA
will
consider
a
number
of
factors
in
assessing
whether
to
exclude
portions
of
a
metropolitan
area
from
the
boundaries
of
a
PM2.5
nonattainment
area.
One
of
the
factors
listed
was
meteorology
(
weather/
transport
patterns).
Connecticut
has
recently
used
the
ISCST3
model
to
demonstrate
that
because
of
transport
patterns,
primary
fine­
particulate
matter
emissions
from
Connecticut
sources
have
a
low
impact
on
receptors
in
New
York
City
and
Hudson
County
New
Jersey.

II.
Methodology
It
has
been
recognized
for
a
number
of
years
that
air
quality
models
using
fugitive
dust
emission
inventories
substantially
overestimate
the
ambient
PM2.5
crustal
material
actually
found
in
ambient
samples
(
EPA
memo,
Thomas
G.
Pace,
August
22,
2003).
It
was
suggested
that
most
of
a
dust
plume
remains
close
to
the
ground
and
that
air
quality
models
"
do
not
adequately
account
for
injection
height,
deposition
losses
and
impaction
losses
near
fugitive
dust
emission
sources."
Because
of
this,
fugitive
emissions
from
the
following
SCCs
(
source
classification
codes)
were
reduced
by
90
percent
to
provide
a
better
estimate
of
PM2.5:

 
SCC
2311020000
(
Industrial
Processes,
Construction,
Heavy
Construction,
Total)
 
SCC
2311030000
(
Industrial
Processes,
Construction,
Road
Construction,
Total)
 
SCC
2325000000
(
Industrial
Processes,
Mining
and
Quarrying,
All
Processes,
Total)
15
 
SCC
2801000003
(
Misc.
Area
Sources,
Agricultural
Production,
Agricultural
Crops,
Tilling)
 
SCC
2294000000
(
Mobile
Sources,
Paved
Roads,
All
Paved
Roads,
Total:
Fugitives)
 
2296000000
(
Mobile
Sources,
Unpaved
Roads,
All
Unpaved
Roads,
Total:
Fugitives)

For
all
the
residential
wood
burning
(
stationary
source
fuel
combust,
residential,
wood)
and
all
SCC8
categories:
SCCs
(
210008030,
2104008004,
210400850,
2104008003,
2104008001,
2104008010,
and
2104008002),
the
estimated
emissions
were
reduced
to
zero
pending
a
new
report
from
MARAMA
(
Mid­
Atlantic
Regional
Air
Management
Association)
that
is
expected
to
revise
the
estimates.

Finally,
the
following
SCCs
were
revised
using
estimates
from
the
MANE­
VU
(
Mid­
Atlantic/
Northeast
Visibility
Union)
report
on
open
burning
"
Open
Burning
in
Residential
Areas,
Emissions
Inventory
Development
Report"
prepared
by
EH
Pechan
and
Assoc.
Inc.
January
31,
2003
for
MANE­
VU:

 
SCC
2610000400
(
Waste
disposal/
treatment,
open
burning,
all
categories,
yard
wastebrush
spec.
unspecified)
 
SCC
2610000100
(
Waste
disposal/
treatment,
open
burning,
all
categories,
yard
wasteleaf
spec.
unspecified)
 
SCC
2610030000
(
Waste
disposal/
treatment,
open
burning,
residential,
household
waste)

 
SCC
2610000500
(
Waste
disposal/
treatment,
open
burning,
all
categories,
land
clearing
debris)
was
reduced
to
zero,
since
it
was
not
accounted
for
in
the
report.

For
simplicity,
it
was
decided
to
add
the
point,
non­
road,
and
mobile
source
emissions
to
the
area
source
emissions
from
each
county,
to
obtain
a
total
area
source
emissions
to
input
into
the
model.
Each
county
was
converted
into
an
approximately
shaped
rectangular
area
and
placed
on
a
map
grid
of
the
region
(
Figure
33).
All
counties
were
oriented
horizontally,
except
the
Manhattan
County
source,
which
was
rotated
at
a
25­
degree
clockwise
angle.
The
southwest
corner
coordinates
for
each
rectangular
county
area
were
then
obtained
and
input
to
the
model.
A
Summary
of
Emission
sources
is
provided
in
Table
3.1.

The
approximate
locations
for
the
five
discrete
Cartesian
receptor
points
were
also
obtained
from
the
modeling
grid
and
the
input
from
all
source
counties
was
run
for
each
of
the
receptor
points.
The
county
containing
the
receptor
was
included
since
the
model
allows
it
(
except
for
very
small
areas
of
a
few
meters
across).
The
modeled
source
receptor
locations
are
also
shown
in
Figure
33.

In
order
to
test
the
variability
from
meteorological
conditions,
each
receptor
was
run
using
two
different
meteorological
data
sets.
One
run
used
1994
surface
data
from
LaGuardia,
NY
and
upper
air
data
from
Atlantic
City,
NJ,
and
the
other
run
used
1974
surface
data
from
Sikorsky
Airport
in
Bridgeport,
CT
and
upper
air
data
from
Kennedy
Airport,
NY.

The
ISCST3
dispersion
model
was
run
using
the
area
source
subroutine
with
the
following
parameters:
16
 
Regulatory
DEFAULT
option
 
URBAN
dispersion
parameter
 
CONCENTRATION
output
units
 
NO
COMPLEX
terrain
 
24
Hour
averaging
PERIOD
 
No
particle
deposition
 
Output=
annual
average
concentration
from
each
source
III.
Results
The
emissions
and
modeled
estimates
are
for
primary
PM2.5
and
are
not
intended
to
represent
any
secondarily
formed
PM,
such
as
sulfates,
nitrates
or
organic
aerosols.
Also
the
modeling
is
not
designed
to
replicate
any
localized
neighborhood
or
microscale
effects.

Table
3.2
shows
results
for
the
New
York
City
receptor.
The
model
results
indicate
that
Connecticut
source
contributions
ranged
from
1.7%
of
the
total
using
LaGuardia
surface
met
data
to
2.1%
of
the
total
when
Bridgeport
surface
met
data
was
input
to
the
ISCST3
model.
Connecticut
source
contributions
for
the
Union
City,
New
Jersey
receptor
ranged
from
2.9%
of
the
total
(
LaGuardia
met
data)
to
2.3%
of
the
total
for
Bridgeport
met
data
(
Table
3.2).
For
comparison
purposes,
results
to
three
Connecticut
receptors
are
also
included.
Contributions
from
the
individual
States
to
each
receptor
are
also
plotted
in
pie
charts
(
Figure
34).
For
the
receptors
in
Bridgeport
and
New
Haven,
Connecticut
sources
contributed
more
than
half
the
primary
PM2.5
totals,
with
New
Jersey
and
New
York
contributing
significant
percentages.

As
stated
earlier,
the
area
source
emissions
were
adjusted
for
specific
categories,
and
this
resulted
in
the
substantial
reduction
of
area
source
emissions
from
all
the
counties.
Most
of
the
differences
can
be
attributed
to
the
90%
reduction
of
the
primary
fugitive
dust
categories.
In
order
to
compare
the
effect
that
this
adjustment
had
on
the
modeling
output,
Table
3.1
include
a
column
with
results
from
running
the
model
with
the
unadjusted
annual
emissions
for
the
NYC
receptor.
The
concentration
at
this
receptor
increased
by
47%
overall,
however,
as
displayed
on
the
pie
chart
(
Figure
34),
the
contribution
from
CT
sources
only
increased
from
1.7%
to
2.8%
of
the
total.
Clearly,
adjusting
the
emissions
inventory
did
not
affect
the
significance
of
the
CT
contribution.

The
model
was
also
run
using
20km
grid
points
as
receptors
and
the
results
are
displayed
in
Figures
35
and
36.
Grid
point
concentration
values
are
generally
higher
around
the
immediate
NYC
area
when
using
the
Bridgeport
(
Sikorsky
Airport)
meteorological
surface
data.
To
better
visualize
the
relative
impacts,
these
grid
point
data
are
plotted
as
concentration
circles
in
Figures
37
and
38.
Wind
rose
diagrams
have
been
plotted
for
the
LaGuardia
and
Bridgeport
surface
meteorological
inputs
in
Figures
39
and
40.
These
show
the
wind
patterns
for
their
respective
years
to
be
noticeably
difference
in
appearance.
Regardless
of
this
fact,
the
Connecticut
contribution
to
the
NYC
receptor
has
been
shown
to
increase
only
slightly
when
using
the
Bridgeport
meteorological
data.
17
Table
3.1:
Area
Source
Parameters
as
Inputs
to
the
ISCST3
Model.

PM
2.5
Modeling
Area
Source
1999
NEI
Emissions
(
Using
LaGuardia
Met
Data)

County
State
County
Code
Km
2
x
Km
y
Km
Rectangle
Km
2
SW
Corner
(
x,
y)
Point
Source
Emissions
TPY
Area
Source
Emissions
TPY
Adjusted
Area
Source
TPY
Highway
Emissions
TPY
Non­
road
Emissions
TPY
Total
Emissions
TPY
Adjusted
Total
Emissions
TPY
Adjusted
Emissions
g/
s/
m
2
Unadjusted
Emissions
g/
s/
m
2
NYC
Adjusted
Annual
Average
µ
g/
m
3
NYC
Unadjusted
Annual
Average
µ
g/
m
3
New
Haven
CT
9
1603
40
40
1600
186,130
718
2334
484
410
620
4082
2232
4.01E­
08
7.33E­
08
0.051
0.094
Litchfield
CT
5
2448
43
57
2451
140,150
22
1978
347
91
98
2189
558
6.56E­
09
2.57E­
08
0.008
0.032
Fairfield
CT
1
1671
38
44
1672
147,112
267
3626
682
454
746
5093
2149
3.70E­
08
8.77E­
08
0.085
0.200
Hartford
CT
3
1945
40
48
1920
183,157
276
4752
767
460
337
5825
1840
2.72E­
08
8.62E­
08
0.033
0.106
Middlesex
CT
7
997
28
35
980
206,120
190
1478
199
112
111
1891
612
1.77E­
08
5.46E­
08
0.008
0.024
New
London
CT
11
1793
51
35
1785
232,126
1223
2202
370
192
189
3806
1974
3.17E­
08
6.11E­
08
0.015
0.028
Tolland
CT
13
1080
24
45
1080
223,160
9
1327
207
113
51
1500
380
1.01E­
08
4.00E­
08
0.006
0.024
Windham
CT
15
1349
32
42
1344
248,160
31
1180
191
106
44
1361
372
7.93E­
09
2.90E­
08
0.004
0.016
Dutchess
NY
27
2137
39
55
2145
100,150
7
3983
527
204
151
4345
889
1.20E­
08
5.85E­
08
0.010
0.047
Putnam
NY
79
637
36
18
648
100,130
0.04
1460
135
63
47
1570.04
245.04
1.11E­
08
7.09E­
08
0.004
0.028
Westchester
NY
119
1230
28
44
1232
100,84
68
4657
734
239
505
5469
1546
3.62E­
08
1.28E­
07
0.069
0.245
Bronx
NY
5
111
11
10
110
100,70
37
2261
577
338
218
2854
1170
3.03E­
07
7.40E­
07
0.254
0.620
New
York
NY
61
72
3.5
21
73.5
92.6,58.4
439
3080
2125
418
1171
5108
4153
1.66E­
06
2.04E­
06
9.753
11.986
Queens
NY
81
280
17
16
272
101,51
463
4787
1751
548
786
6584
3548
3.65E­
07
6.77E­
07
0.258
0.479
Kings
NY
47
179
13
13
169
94,46
226
4363
1357
647
411
5647
2641
4.24E­
07
9.08E­
07
0.516
1.104
Nassau
NY
59
717
21
34
714
118,46
135
7314
949
361
481
8291
1926
7.73E­
08
3.33E­
07
0.043
0.184
Suffolk
NY
103
2388
100
24
2400
140,60
324
9040
1228
413
1017
10794
2982
3.59E­
08
1.30E­
07
0.017
0.062
Richmond
NY
85
150
12
12
144
77,40
242
1929
485
106
144
2421
977
1.87E­
07
4.64E­
07
0.081
0.200
Orange
NY
71
2173
55
40
2200
44,113
136
4529
492
264
158
5087
1050
1.39E­
08
6.74E­
08
0.020
0.099
Rockland
NY
87
515
20
26
520
85,96
195
1768
379
74
118
2155
766
4.28E­
08
1.20E­
07
0.026
0.073
Sussex
NJ
37
1388
32
43
1376
26,87
7
1519
210
110
101
1737
428
8.87E­
09
3.60E­
08
0.017
0.071
Passaic
NJ
31
510
18
29
522
66,77
26
1713
399
210
245
2194
880
4.96E­
08
1.24E­
07
0.066
0.164
Bergen
NJ
3
640
20
32
640
80,72
207
3558
912
383
817
4965
2319
1.04E­
07
2.23E­
07
0.203
0.436
Hudson
NJ
13
132
10
13
130
85,55
470
1842
552
257
876
3445
2155
4.70E­
07
7.51E­
07
0.605
0.966
Essex
NJ
13
332
17
20
340
66,60
270
3145
699
363
654
4432
1986
1.72E­
07
3.84E­
07
0.194
0.432
Union
NJ
39
269
19
14
266
60,46
97
2559
565
230
451
3337
1343
1.44E­
07
3.57E­
07
0.048
0.119
Morris
NJ
27
1246
42
30
1260
27,63
56
3215
541
253
408
3932
1258
2.90E­
08
9.08E­
08
0.054
0.170
Warren
NJ
41
940
22
43
946
0,56
227
1275
201
123
91
1716
642
1.96E­
08
5.25E­
08
0.012
0.031
Hunterdon
NJ
19
1134
33
34
1122
0,30
52
1772
224
156
130
2110
562
1.43E­
08
5.35E­
08
0.007
0.025
Somerset
NJ
35
790
20
40
800
32,24
171
2369
444
148
286
2974
1049
3.82E­
08
1.08E­
07
0.019
0.053
Middlesex
NJ
23
818
25
32
800
48,12
400
4584
832
347
661
5992
2240
7.88E­
08
2.11E­
07
0.068
0.182
Mercer
NJ
21
593
24
25
600
22,0
174
3043
1046
214
335
3766
1769
8.58E­
08
1.83E­
07
0.032
0.068
Monmouth
NJ
25
1235
49
25
1225
48,0
31
3950
724
325
565
4871
1645
3.83E­
08
1.13E­
07
0.050
0.148
Sum
Total
12.64
18.52
CT
0.21
0.52
NY
11.05
15.13
NJ
1.37
2.87
18
Table
3.2:
Summary
of
Annual
Average
Modeled
PM2.5
Concentrations
NYC
(
Manhattan)
Bridgeport
CT
New
Haven
CT
Greenwich
CT
Union
City
NJ
Source
County
Adjusted
Annual
Average
µ
g/
m3
Contribution
LaGuardia/
Atlc
City
1994
Met
Data
Adjusted
Annual
Average
µ
g/
m3
Contribution
Sikorsky/
Kennedy
1974
Met
Data
Adjusted
Annual
Average
µ
g/
m3
Contribution
LaGuardia/
Atlc
City
1994
Met
Data
Adjusted
Annual
Average
µ
g/
m3
Contribution
Sikorsky/
Kennedy
1974
Met
Data
Adjusted
Annual
Average
µ
g/
m3
Contribution
LaGuardia/
Atlc
City
1994
Met
Data
Adjusted
Annual
Average
µ
g/
m3
Contribution
Sikorsky/
Kennedy
1974
Met
Data
Adjusted
Annual
Average
µ
g/
m3
Contribution
LaGuardia/
Atlc
City
1994
Met
Data
Adjusted
Annual
Average
µ
g/
m3
Contribution
Sikorsky/
Kennedy
1974
Met
Data
Adjusted
Annual
Average
µ
g/
m3
Contribution
LaGuardia/
Atlc
City
1994
Met
Data
Adjusted
Annual
Average
µ
g/
m3
Contribution
Sikorsky/
Kennedy
1974
Met
Data
New
Haven
CT
0.051
0.043
0.192
0.214
0.578
0.686
0.068
0.068
0.046
0.042
Litchfield
CT
0.008
0.007
0.012
0.016
0.015
0.019
0.010
0.010
0.009
0.007
Fairfield
CT
0.085
0.067
0.488
0.585
0.079
0.159
0.265
0.301
0.078
0.067
Hartford
CT
0.033
0.025
0.046
0.040
0.051
0.056
0.045
0.033
0.033
0.024
Middlesex
CT
0.008
0.009
0.016
0.018
0.026
0.030
0.008
0.011
0.007
0.009
New
London
CT
0.015
0.022
0.023
0.031
0.028
0.041
0.017
0.022
0.014
0.022
Tolland
CT
0.006
0.004
0.011
0.008
0.014
0.010
0.007
0.006
0.006
0.004
Windham
NY
0.004
0.003
0.007
0.006
0.009
0.008
0.005
0.005
0.004
0.003
Dutchess
NY
0.010
0.013
0.021
0.019
0.029
0.017
0.015
0.020
0.009
0.012
Putnam
NY
0.004
0.006
0.015
0.009
0.006
0.007
0.008
0.012
0.004
0.005
Westchester
NY
0.069
0.088
0.040
0.085
0.022
0.062
0.343
0.426
0.073
0.069
Bronx
NY
0.254
0.308
0.018
0.045
0.012
0.032
0.071
0.104
0.210
0.204
New
York
NY
9.753
5.445
0.062
0.144
0.042
0.107
0.197
0.261
1.453
1.467
Queens
NY
0.258
0.192
0.077
0.099
0.053
0.083
0.154
0.156
0.102
0.237
Kings
NY
0.516
0.163
0.047
0.063
0.035
0.055
0.084
0.101
0.259
0.168
Nassau
NY
0.043
0.065
0.055
0.067
0.038
0.050
0.147
0.081
0.032
0.068
Suffolk
NY
0.017
0.041
0.080
0.062
0.081
0.062
0.025
0.050
0.015
0.039
Richmond
NY
0.081
0.068
0.014
0.022
0.010
0.018
0.025
0.028
0.100
0.091
Orange
NY
0.020
0.029
0.020
0.021
0.015
0.019
0.045
0.028
0.021
0.028
Rockland
NJ
0.026
0.044
0.015
0.031
0.009
0.026
0.057
0.047
0.026
0.036
Sussex
NJ
0.017
0.011
0.004
0.009
0.003
0.008
0.008
0.010
0.018
0.011
Passaic
NJ
0.066
0.064
0.009
0.031
0.006
0.024
0.022
0.048
0.061
0.061
Bergen
NJ
0.203
0.465
0.025
0.095
0.018
0.066
0.077
0.179
0.200
0.280
Hudson
NJ
0.605
0.678
0.027
0.063
0.019
0.048
0.072
0.103
3.497
3.968
Essex
NJ
0.194
0.288
0.017
0.060
0.014
0.048
0.033
0.104
0.270
0.306
Union
NJ
0.048
0.129
0.012
0.030
0.009
0.024
0.025
0.049
0.068
0.175
Morris
NJ
0.054
0.064
0.008
0.032
0.007
0.025
0.015
0.050
0.065
0.060
Warren
NJ
0.012
0.018
0.003
0.013
0.002
0.011
0.006
0.016
0.013
0.016
Hunterdon
NJ
0.007
0.022
0.003
0.010
0.002
0.009
0.004
0.013
0.007
0.022
Somerset
NJ
0.019
0.053
0.007
0.020
0.006
0.017
0.011
0.029
0.021
0.059
Middlesex
NJ
0.068
0.089
0.023
0.039
0.018
0.034
0.038
0.049
0.077
0.105
Mercer
NJ
0.032
0.047
0.014
0.026
0.012
0.023
0.022
0.032
0.036
0.053
Monmouth
NJ
0.050
0.042
0.019
0.024
0.016
0.020
0.025
0.030
0.054
0.047
TOTAL
(
µ
g/
m3)
12.64
8.61
1.43
2.04
1.29
1.90
1.95
2.48
6.89
7.76
CT
Total
(
µ
g/
m3)
0.21
0.18
0.79
0.92
0.80
1.01
0.43
0.46
0.20
0.18
NY
Total
(
µ
g/
m3)
11.05
6.46
0.46
0.67
0.35
0.54
1.17
1.31
2.30
2.42
NJ
Total
(
µ
g/
m3)
1.37
1.97
0.17
0.45
0.13
0.36
0.36
0.71
4.39
5.16
%
CT
1.7
2.1
55.6
45.1
62.2
53.0
21.8
18.4
2.9
2.3
%
NY
87.5
75.0
32.4
32.8
27.5
28.3
59.9
52.9
33.4
31.2
%
NJ
10.9
22.9
12.0
22.2
10.3
18.7
18.3
28.7
63.7
66.5
19
B.
Back
Trajectory
Analysis
of
Days
with
High
and
Low
PM2.5
Concentrations
in
New
York
City
The
highest
PM2.5
concentrations
measured
in
New
York
City
are
at
the
P.
S.
59
monitor
in
Manhattan.
The
daily
PM2.5
concentrations
over
a
nearly
5­
year
period
(
1/
99­
9/
03)
at
P.
S.
59
were
analyzed
and
rank­
ordered
from
highest
to
lowest.
The
dates
for
the
top
and
bottom
10
percentile
were
obtained.
The
National
Oceanic
and
Atmospheric
Administration's
HYSPLIT4
model
(
http://
www.
arl.
noaa.
gov/
ready/
hysplit4.
html)
was
then
used
to
produce
back
trajectories
once
a
day
for
each
of
those
days
at
three
height
levels
(
10,
500,
and
1000
meters
above
ground
level).
All
72
hourly
positions
(
or
back
trajectories)
of
the
model
run
at
each
vertical
level
were
saved
each
day
for
the
"
dirty"
and
"
clean"
categories.

The
back
trajectories
for
the
"
dirtiest"
or
highest
10
percentile
PM2.5
concentration
days
are
plotted
in
Figure
41
and
the
"
cleanest"
or
lowest
10
percentile
PM2.5
concentration
days
are
plotted
in
Figure
42.
One
can
see
immediately
that
the
source
regions
from
which
air
is
being
transported
into
New
York
City
are
distinctly
different
for
the
two
scenarios.
These
figures
demonstrate
that
during
the
dirtiest
days,
the
air
arriving
in
New
York
City
comes
from
and
passes
through
locations
in
a
sector
from
the
south
and
west
(
Figure
41).
Conversely,
during
the
cleanest
days
the
air
arriving
in
New
York
City
comes
from
and
passes
through
locations
in
a
sector
from
the
north
and
east.

Since
Connecticut
is
northeast
of
New
York
City,
this
back
trajectory
analysis
supports
the
conclusion
that
Connecticut's
emissions
are
not
contributing
significantly
to
the
highest
PM2.5
levels
measured
in
New
York
City.
20
C.
Interstate
Air
Quality
Rule
Modeling
In
the
proposed
"
Rule
To
Reduce
Interstate
Transport
of
Fine
Particulate
Matter
and
Ozone
(
Interstate
Air
Quality
Rule)",
Federal
Register
/
Vol.
69,
No.
20
/
published
on
Friday,
January
30,
2004,
EPA
conducted
an
evaluation
of
the
upwind
contributions
to
downwind
PM2.5
non­
attainment.

EPA
used
the
REgional
Modeling
System
for
Aerosols
and
Deposition
(
REMSAD)
as
the
tool
for
simulating
base
year
and
future
concentrations
of
PM,
visibility,
and
deposition
in
support
of
the
IAQR
air
quality
assessments.
According
to
the
Technical
Support
Document
for
the
Interstate
Air
Quality
Rule
Air
Quality
Modeling
Analyses
(
January
2004):

"
The
basis
for
REMSAD
is
the
atmospheric
diffusion
equation
(
also
called
the
species
continuity
or
advection/
diffusion
equation).
This
equation
represents
a
mass
balance
in
which
all
of
the
relevant
emissions,
transport,
diffusion,
chemical
reactions,
and
removal
processes
are
expressed
in
mathematical
terms.
REMSAD
employs
finite­
difference
numerical
techniques
for
the
solution
of
the
advection/
diffusion
equation."

The
REMSAD
model
is
much
more
sophisticated
than
the
modeling
that
CTDEP
staff
conducted
with
the
ISCST3
model,
in
that
it
involves
secondary
PM2.5
formation.
The
aforementioned
Technical
Support
Document
describes
it
as
follows:

"
Primary
PM
emissions
in
REMSAD
are
treated
as
inert
species.
They
are
advected
and
deposited
without
any
chemical
interaction
with
other
species.
Secondary
PM
species,
such
as
sulfate
and
nitrate
are
formed
through
chemical
reactions
within
the
model.
SO2
is
the
gas
phase
precursor
for
particulate
sulfate,
while
nitric
acid
is
the
gas
phase
precursor
for
particulate
nitrate.
Several
other
gas
phase
species
are
also
involved
in
the
secondary
reactions.
There
are
two
pathways
for
sulfate
formation;
gas
phase
and
aqueous
phase.
Aqueous
phase
reactions
take
place
within
clouds,
rain,
and/
or
fog.
In­
cloud
processes
can
account
for
the
majority
of
atmospheric
sulfate
formation
in
many
areas."

EPA
used
REMSAD
to
perform
State­
by­
State
zero­
out
modeling
to
quantify
the
contribution
from
emissions
in
each
State
to
future
PM2.5
nonattainment
in
other
States.
They
analyzed
a
total
of
41
States
on
a
State­
by­
State
basis
in
different
model
runs.
EPA
is
proposing
to
use
a
criterion
of
0.15
µ
g/
m3
for
determining
whether
emissions
in
a
State
make
a
significant
contribution
(
before
considering
cost)
to
PM2.5
nonattainment
in
another
State.
Of
the
States
analyzed
for
this
proposal,
28
States
and
the
District
of
Columbia
contribute
0.15
µ
g/
m3
or
more
to
nonattainment
in
other
States
and
therefore
would
be
found
to
make
a
significant
contribution
to
PM2.5.

The
maximum
downwind
contribution
from
each
upwind
State
to
a
downwind
nonattainment
county
is
provided
in
Table
3.3
(
from
page
4608
Federal
Register
/
Vol.
69,
No.
20)
and
Connecticut
was
found
to
contribute
.07
µ
g/
m3
of
PM2.5
concentration
to
a
New
York
City
receptor.
According
to
the
EPA
criteria
described
above,
Connecticut
does
not
significantly
contribute
to
non­
attainment
in
New
York,
NY.
21
TABLE
3.3.
MAXIMUM
DOWNWIND
PM
2.5
CONTRIBUTION
(
µ
g/
m3)
FOR
EACH
OF
41
UPWIND
STATES
Maximum
Downwind
Upwind
state
Downwind
nonattainment
contribution
county
of
maximum
µ
g/
m3
contribution
Alabama
...........................................................................................................................................
1.17
Floyd,
GA.
Arkansas
...........................................................................................................................................
0.29
St.
Clair,
IL.
Connecticut
.......................................................................................................................................
0.07
New
York,
NY.
Colorado
............................................................................................................................................
0.04
Madison,
IL.
Delaware
...........................................................................................................................................
0.17
Berks,
PA.
Florida
...............................................................................................................................................
0.52
Russell,
AL.
Georgia
.............................................................................................................................................
1.52
Russell,
AL.
Illinois
...............................................................................................................................................
1.50
St.
Louis,
MO.
Indiana
..............................................................................................................................................
1.06
Hamilton,
OH.
Iowa
..................................................................................................................................................
0.43
Madison,
IL.
Kansas
...............................................................................................................................................
0.15
Madison,
IL.
Kentucky
...........................................................................................................................................
1.10
Clark,
IN.
Louisiana
...........................................................................................................................................
0.25
Jefferson,
AL.
Maryland/
District
of
Columbia
.........................................................................................................
0.85
York,
PA.
Maine
.................................................................................................................................................
0.03
New
Haven,
CT.
Massachusetts
............................................................................................................................ .....
0.21
New
Haven,
CT.
Michigan
............................................................................................................................................
0.88
Cuyahoga,
OH.
Minnesota
................................................................................................................................... ....
0.39
Cook,
IL.
Mississippi
..........................................................................................................................................
0.30
Jefferson,
AL.
Missouri
..............................................................................................................................................
0.89
Madison,
IL.
Montana
....................................................................................................................................... ...
0.03
Cook,
IL.
Nebraska
........................................................................................................................................ .
0.08
Madison,
IL.
New
Hampshire
............................................................................................................................... 
0.06
New
Haven,
CT.
New
Jersey
..................................................................................................................................... .
0.45
New
York,
NY.
New
Mexico
.......................................................................................................................................
0.03
Knox,
TN.
New
York
...........................................................................................................................................
0.85
New
Haven,
CT.
North
Carolina
....................................................................................................................................
0.41
Sullivan,
TN.
North
Dakota
......................................................................................................................................
0.12
Cook,
IL.
Ohio
................................................................................................................................................. 
1.90
Hancock,
WV.
Oklahoma
........................................................................................................................................ 
0.14
Madison,
IL.
Pennsylvania
......................................................................................................................................
1.17
New
Castle,
DE.
Rhode
Island
......................................................................................................................................
0.01
New
Haven,
CT.
South
Carolina
...................................................................................................................................
0.72
Richmond,
GA.
South
Dakota
.....................................................................................................................................
0.04
Madison,
IL.
Tennessee
............................................................................................................................... ........
0.57
Floyd,
GA.
Texas
........................................................................................................................................ ......
0.37
St.
Clair,
IL.
Vermont
.............................................................................................................................................
0.06
New
Haven,
CT.
Virginia
..............................................................................................................................................
0.67
Washington,
DC.
West
Virginia
.....................................................................................................................................
0.89
Allegheny,
PA.
Wisconsin
..........................................................................................................................................
1.00
Cook,
IL.
Wyoming
...........................................................................................................................................
0.05
Madison,
IL.
22
D.
Southwest
Connecticut
Commuting
Patterns
to
New
York
and
New
Jersey
EPA
guidance
on
nonattainment
area
designations
for
PM2.5
states
"
a
nonattainment
area
must
be
defined
not
only
to
include
the
area
that
is
violating
the
standard,
but
also
to
include
the
nearby
source
areas
that
contribute
to
the
violation"
(
see
page
4
of
Attachment
2
to
J.
Holmstead's
memorandum
of
April
1,
2003).
Discussion
in
the
sections
above
provides
evidence
from
air
quality
modeling
and
air
parcel
trajectory
analyses
indicating
that
emissions
emanating
from
within
the
borders
of
Connecticut
do
not
significantly
contribute
to
PM2.5
levels
in
either
the
New
York
or
New
Jersey
portions
of
the
New
York
City
Consolidated
Statistical
Area
(
CSA).
In
order
to
develop
a
more
complete
picture
of
the
Connecticut's
potential
impact
on
PM2.5
levels
in
the
New
York
and
New
Jersey
portions
of
the
CSA,
it
is
also
important
to
examine
Connecticut's
contribution
to
motor
vehicle
traffic
traveling
within
the
remainder
of
the
CSA.

Available
2000
Census
Bureau
data
on
work­
trip
origins
and
destinations
were
judged
to
provide
a
reasonable
surrogate
for
assessing
Connecticut's
contribution
to
traffic
levels
in
the
New
York
City
CSA.
Ideally,
detailed
and
current
traffic
survey
data
would
have
been
used
to
determine
the
fraction
of
vehicular
traffic
traveling
in
the
non­
Connecticut
portion
of
the
New
York
City
CSA
originating
from
within
Connecticut's
portion
of
the
CSA
(
i.
e.,
Fairfield,
New
Haven,
and
Litchfield
Counties).
However,
efforts
to
identify
and
obtain
such
data
were
unsuccessful.
Excel
spreadsheets
were
developed
to
process
county­
level
work­
trip
data
for
each
of
the
three
states
to
determine
the
fraction
of
all
work
trips
to
various
portions
of
the
New
York
City
CSA
that
originated
in
Connecticut.
Results
of
that
analysis
are
summarized
below
in
Table
3.4.

Table
3.4
Work
Trips
from
the
Connecticut
Portion
of
the
CSA
as
a
Percentage
of
Total
Work­
Trips
Into:
Combined
NY
and
NJ
Portion
of
CSA
New
Jersey
Portion
of
CSA
New
York
State
Portion
of
CSA
New
York
City
portion
of
CSA
0.7%
0.1%
1.0%
0.9%

Overall,
Connecticut's
portion
of
the
CSA
contributes
only
0.7%
of
total
work­
trips
destined
for
the
combined
New
York
and
New
Jersey
portions
of
the
CSA.
When
the
work­
trip
data
are
examined
on
a
smaller
geographic
scale,
Connecticut's
contribution
ranges
from
0.1%
in
the
New
Jersey
portion
to
1.0%
in
the
New
York
State
portion
of
the
CSA.
For
all
work­
trips
headed
into
the
five
boroughs
of
New
York
City,
only
0.9%
of
work
trips
originate
from
Connecticut's
portion
of
the
CSA.

The
above
data
do
not
differentiate
between
the
various
modes
of
travel
available
to
commuters,
such
as
motor
vehicles
and
mass
transit.
The
2000
Census
Bureau
travel
data
were
not
differentiated
between
travel
modes,
however,
CTDEP
previously
analyzed
1990
travel
data
as
part
of
the
redesignation
SIP
package
prepared
for
the
Southwest
Connecticut
carbon
monoxide
nonattainment
area.
That
analysis
found
0.6%
of
all
motor
vehicle
work­
trips
to
the
five
boroughs
of
New
York
City
originated
in
Connecticut's
Fairfield
County,
consistent
with
the
figures
in
the
above
table.
23
Based
on
these
travel
data,
it
is
logical
to
conclude
that
Connecticut's
motor
vehicle
emissions
within
the
New
York
and
New
Jersey
portions
of
the
CSA
are
an
insignificant
fraction
of
the
total.
When
viewed
together
with
the
dispersion
modeling
and
trajectory
analyses
presented
earlier,
it
is
clear
that
Connecticut
emission
sources
do
not
significantly
affect
ambient
PM2.5
levels
throughout
the
remainder
of
the
New
York
City
CSA.
24
4.
Outreach
Activities
Undertaken
Regarding
the
Reclassifying
of
the
Stiles
Street
Monitor
to
Microscale.

Upon
the
recommendation
of
EPA
Region
I,
CTDEP
conducted
an
outreach
campaign,
providing
several
presentations
to
various
groups
in
New
Haven
and
around
the
State.
The
intent
of
the
outreach
was
to
inform
the
regulated
community,
environmental
groups,
and
the
general
public
that
reclassification
of
the
Stiles
Street
monitor
could
lead
to
an
attainment
designation
for
PM2.5.

In
addition
to
addressing
the
consequences
of
monitor
reclassification,
the
outreach
presentations
included
a
discussion
of
the
rational
for
reclassification.
The
Stiles
Street
monitor
is
significantly
influenced
by
microscale
phenomena,
particularly
diesel
truck
emissions
from
heavily
loaded
trucks
accelerating
up
the
steeply
graded
I­
96
southbound
access
ramp.
The
access
ramp
and
nearest
section
of
I­
95,
the
approach
to
the
Q­
Bridge,
are
within
distances
on
the
order
of
tens
of
meters
from
the
monitor.
The
immediate
area
of
the
monitor
is
highly
industrial,
and
does
not
include
residential
areas.
As
such,
it
is
not
representative
of
community
exposure
and,
consistent
with
EPA's
guidance,
should
be
treated
as
a
microscale
site
for
PM2.5
classification
purposes.

The
following
presentations
were
given
to
inform
the
public
of
the
reclassification
of
the
Stiles
Street
PM2.5
monitor:

a)
June
6,
2003­
Presentation
to
the
CTDEP
Air
Bureau
b)
June
16,
2003­
Presentation
to
EPA
Region
1,
quarterly
meeting
c)
November
12,
2003­
Presentation
to
the
New
Haven
EQ
Group
d)
January
8,
2004­
Presentation
to
SIPRAC
meeting
e)
January
22,
2004­
Presentation
to
a
New
Haven
public
meeting
Note,
item
(
d)
above
contains
the
CTDEP
Website
link
to
the
January
8,
2004
SIPRAC
meeting
presentation.
25
5.
Appropriate
Use
of
Stiles
Street
Monitoring
Data
Based
on
EPA
Regulations
and
Guidance
Documents
The
purpose
of
this
section
is
to:
1)
examine
available
EPA
regulations
and
guidance
to
provide
some
general
perspective
on
the
siting
of
PM2.5
monitoring
sites
and
the
appropriate
use
of
collected
data
for
determining
compliance
with
the
PM2.5
NAAQS;
and
2)
consider
how
EPA's
guidance
should
be
applied
to
the
New
Haven
Stiles
Street
monitor.
Excerpts
from
the
following
EPA
documents
are
reproduced
and
discussed
below.

Doc.
A
Guidance
for
Network
Design
and
Optimum
Site
Exposure
for
PM2.5
and
PM10
(
EPA­
454/
R­
99­
022,
December
1997)

Doc.
B
Designations
for
the
Fine
Particle
National
Ambient
Air
Quality
Standards
(
EPA
memo,
J.
Holmstead,
April
1,
2003)

Doc.
C
40CFR58,
Appendix
D.
Network
Design
for
State
and
Local
Air
Monitoring
Stations
(
SLAMS),
National
Air
Monitoring
Stations
(
NAMS),
and
Photochemical
Assessment
Monitoring
Stations
(
PAMS)

Doc.
D
40CFR58,
Appendix
E.
Probe
and
Monitoring
Path
Siting
Criteria
for
Ambient
Air
Quality
Monitoring
The
Jeff
Holmstead
memorandum
regarding
PM2.5
designations
includes
a
brief
section
desribing
how
NAAQS
violations
are
to
identified
when
examining
monitoring
data
(
see
Section
3
of
Doc.
B).
It
notes
exceptional
circumstances
when
concentrations
above
the
level
of
the
standard
are
not
to
be
interpreted
as
violations,
stating:

"
Sites
that
monitor
source­
oriented
hot
spots
in
some
cases
should
be
assessed
only
with
respect
to
the
24­
hour
standard,
not
the
annual
average
standard.
In
40CFR
Part
58
(
Appendix
D,
section
2.8.1.2.3),
EPA
states
that
monitoring
sites
representing
unique
localized
conditions
not
found
elsewhere
in
the
area
should
not
be
compared
with
the
annual
average
standard."

The
following
discussion
takes
a
closer
look
at
EPA
background
information
directly
affecting
the
applicability
of
this
potential
exception
to
the
Stiles
Street
monitoring
site.

A.
EPA
Changes
in
Monitoring
Objectives
Under
the
New
PM2.5
Standards
EPA's
guidance
for
network
design
(
see
Doc.
A,
section
1.0)
describes
how
monitoring
objectives
have
changed
with
the
implementation
of
the
new
PM2.5
standards,
pointing
out
distinct
differences
between
the
objectives
for
PM2.5
versus
PM10
networks:

"
Previously,
the
PM
NAAQS
applied
to
the
highest
24­
hour
or
annual
averages
found
within
a
monitoring
planning
area,
and
monitoring
networks
were
often
designed
to
measure
these
highest
values.
These
networks
did
not
necessarily
26
represent
the
overall
exposure
of
populations
to
excessive
PM
concentrations.
Some
data
from
these
networks
were
disregarded
by
epidemiologists
as
being
unrelated
to
health
indicators
such
as
hospital
admissions
and
death.
Air
quality
districts
may
have
been
reluctant
to
locate
source­
oriented
monitors
that
might
assist
in
understanding
source
impacts
because
such
monitors
might
cause
a
large
area
to
be
designated
in
nonattainment
of
NAAQS.

The
new
forms
for
these
standards
are
intended
to
provide
more
robust
measures
for
the
PM
indicator.
While
PM10
network
design
and
siting
criteria
are
unchanged,
new
PM2.5
monitoring
networks
to
determine
compliance
or
noncompliance
are
intended
to
best
represent
the
exposure
of
populations
that
might
be
affected
by
elevated
PM2.5
concentrations.
As
used
in
this
document,
the
word
compliance
means
attainment
of
a
NAAQS.
This
involves
new
concepts
of
spatial
averaging
and
the
operation
of
some
monitoring
sites
for
PM2.5
measurements
that
are
not
eligible
for
comparison
to
one
or
both
of
the
PM2.5
NAAQS."

It
is
clear
from
the
excerpt
above
that
EPA
recognized
the
need
for
a
fundamental
change
in
PM
network
design
and
the
use
of
collected
monitoring
data
for
the
new
PM2.5
standard.
Previous
use
of
source­
oriented
monitors
not
representative
of
overall
population
exposure
for
determining
NAAQS
compliance
was
judged
by
EPA
to
be
inappropriate
for
PM2.5,
given
that
potential
exposures
at
these
locations
are
unrelated
to
health
indicators
such
as
hospital
admissions
and
premature
death
which
provided
the
basis
for
the
new
PM2.5
standards.
EPA
reiterates
this
point
in
40CFR58,
Appendix
D
(
see
section
2.8.1.2.3
of
Doc.
D):

"
The
health­
effects
data
base
that
served
as
the
basis
for
selecting
the
new
PM2.5
standards
relied
on
a
spatial
average
approach
that
reflects
average
community
oriented
area­
wide
PM
exposure
levels."

In
the
previous
excerpt
from
the
network
design
guidance
(
Doc.
A,
section
1.0),
EPA
provides
a
more
detailed
discussion
of
the
distinction
between
PM10
and
PM2.5
monitoring
networks
under
the
new
standards.
PM10
design
and
siting
criteria
remain
unchanged,
retaining
their
focus
on
identifying
the
highest
concentrations
in
an
area,
regardless
of
the
potential
for
overall
population
exposure.
However,
EPA
states
that
new
PM2.5
networks
intended
for
determining
NAAQS
compliance
should
represent
the
exposure
of
populations
that
might
be
affected
by
elevated
PM2.5
levels.

EPA
notes
that
these
fundamentally
new
concepts
would
result
in
operation
of
some
PM2.5
monitoring
sites
that
are
not
eligible
for
comparison
to
one
or
both
of
the
PM2.5
NAAQS.
EPA
acknowledges
that,
in
the
past,
use
of
such
source­
oriented
sites
for
NAAQS
compliance
served
as
a
disincentive
for
air
quality
agencies
to
site
such
monitors
for
purposes
of
characterizing
source
contributions
in
an
area,
due
to
concerns
that
collected
data
might
lead
to
a
nonattainment
designation.

Based
on
the
discussion
in
previous
sections
of
this
technical
support
document
(
TSD),
there
is
little
doubt
that
the
Stiles
Street
monitor
is
a
source­
oriented
site.
The
site
is
located
in
a
heavily
industrialized/
commercialized
area,
far
removed
from
areas
of
general
population
exposure.
The
monitor
is
sited
within
the
CTDOT
right­
of­
way,
27
immediately
adjacent
to
(
and
sandwiched
between)
the
I­
95
freeway,
the
southbound
Stiles
Street
entrance
ramp
to
the
interstate,
and
Stiles
Street.
Traffic
volumes
on
that
portion
of
I­
95
exceed
100,000
vehicles
per
day,
while
the
freeway
ramp
and
Stiles
Street
serve
as
the
primary
I­
95
access
point
for
heavy
trucks
leaving
the
New
Haven
Terminal
and
many
other
nearby
businesses.
Both
the
entrance
ramp
and
the
adjacent
portion
of
I­
95
are
built
on
steep,
uphill
grades
due
to
their
immediate
proximity
to
a
major
river
crossing,
the
Q­
Bridge.
As
a
result,
trucks
passing
by
the
monitor
experience
high­
load
acceleration
to
achieve/
maintain
highway
speeds
as
they
merge
onto
the
interstate
from
the
ramp
and/
or
approach
the
Q­
Bridge
on
I­
95.

Consistent
with
EPA's
guidance,
CTDEP
views
the
Stiles
Street
PM2.5
monitor
as
a
source­
oriented
site,
appropriately
used
for
characterization
of
source
contributions,
but
not
for
determining
compliance
with
the
PM2.5
NAAQS.
As
evidenced
elsewhere
in
this
TSD,
CTDEP
is
in
the
process
of
analyzing
data
from
this
(
and
other)
monitor(
s)
to
gain
a
better
understanding
of
source
contributions
in
the
area.
Based
on
initial
results,
the
Department
is
already
developing
control
strategies
to
reduce
emissions
in
New
Haven
and
elsewhere
in
the
state.

B.
Appropriate
Classification
Scale
for
the
Stiles
Street
Monitor
EPA's
fundamental
change
in
network
design
for
the
new
PM2.5
standard
is
reflected
in
several
new
concepts
introduced
in
the
network
design
guidance
document,
including
what
EPA
calls
the
"
Receptor
Site
Zone
of
Representation"
and
"
Community­
Oriented
Monitoring"
(
see
Sections
2.2.2
and
2.2.3
of
Doc.
A).

"
2.2.2
Receptor
Site
Zone
of
Representation
PM10
and
PM2.5
concentrations
measured
at
any
receptor
result
from
contributions
of
emissions
from
nearby
and
distant
sources
and
the
zone
of
representation
of
a
monitoring
site
depends
on
the
relative
amounts
contributed
by
sources
on
different
spatial
scales.
The
dimensions
given
below
are
nominal
rather
than
exact
and
are
presented
as
defined
in
40
CFR
part
58.
They
indicate
the
diameter
of
a
circle,
or
the
length
and
width
of
a
grid
square,
with
a
monitor
at
its
center.
 
 
 

 
Microscale
(
10
to
100
m):
Microscale
monitors
show
significant
differences
between
PM2.5
monitors
separated
by
10
to
50
m.
This
often
occurs
when
monitors
are
located
right
next
to
a
low­
level
emissions
source,
such
as
a
busy
roadway,
construction
site,
wood
stove
chimney,
or
short
stack.
Compliance
monitoring
site
exposure
criteria
intend
to
avoid
microscale
influences
even
for
sourceoriented
monitoring
sites.
A
microscale
zone
of
representation
is
primarily
useful
for
studying
emissions
rates
and
zones
of
influence,
as
illustrated
in
Figure
2.1.7.

 
Middle
Scale
(
100
to
500
m):
Middle­
scale
monitors
show
significant
differences
between
locations
that
are
~
0.1
to
0.5
km
apart.
These
differences
28
may
occur
near
large
industrial
areas
with
many
different
operations
or
near
large
construction
sites.
Monitors
with
middle­
scale
zones
of
representation
are
often
source­
oriented,
used
to
determine
the
contributions
from
emitting
activities
with
multiple,
individual
sources
to
nearby
community
exposure
monitors.

 
Neighborhood
Scale
(
500
m
to
4
km):
Neighborhood­
scale
monitors
do
not
show
significant
differences
in
particulate
concentrations
with
spacing
of
a
few
kilometers.
This
dimension
is
often
the
size
of
emissions
and
modeling
grids
used
in
large
urban
areas
for
PM
source
assessment,
so
this
zone
of
representation
of
a
monitor
is
the
only
one
that
should
be
used
to
evaluate
such
models.
Sources
affecting
neighborhood­
scale
sites
typically
consist
of
small
individual
emitters,
such
as
clean,
paved,
curbed
roads,
uncongested
traffic
flow
without
a
significant
fraction
of
heavy­
duty
vehicles,
or
neighborhood
use
of
residential
heating
devices
such
as
fireplaces
and
wood
stoves.
 
 
 

2.2.3
Community­
Oriented
Monitoring
Community­
oriented
(
core)
monitoring
sites
are
beyond
the
zone
of
influence
of
a
single
source,
and
should
have
neighborhood­
to
urban­
scale
zones
of
representation.
The
principal
purpose
of
community­
oriented
monitoring
sites
is
to
approximate
the
short­
term
and
long­
term
exposures
of
large
numbers
of
people
where
they
live,
work,
and
play.
A
monitor
placed
at
the
fence
line
of
an
emissions
source
would
not
be
considered
to
represent
community
exposures,
even
though
there
might
be
residences
abutting
that
fence
line.
A
monitor
placed
in
the
middle
of
an
area
adjacent
to
a
source
would,
however,
be
deemed
a
community
exposure
monitor
for
that
neighborhood
provided
that
the
location
represented
a
zone
of
at
least
0.5
km
in
diameter.
The
fence
line
monitor
might
still
be
operated
because
it
provides
information
on
how
much
the
nearby
source
contributes
to
the
community­
oriented
site.
The
data
from
the
fence
line
monitor
would
not
be
used
to
determine
annual
NAAQS
compliance,
though
it
might
be
used
to
make
comparisons
to
the
24­
hour
standard
or
to
design
control
strategies
to
bring
the
area
into
compliance
with
the
annual
NAAQS."

The
Stiles
Street
monitor's
location
immediately
adjacent
to
both
extremely
high
volumes
of
traffic
on
I­
95
and
significant
amounts
of
accelerating
heavy
truck
traffic
on
Stiles
Street
and
the
highway
entrance
ramp
is
consistent
with
EPA's
above
description
of
a
microscale
zone
of
representation.

The
localized
effect
of
trucks
traveling
on
Stiles
Street
and
the
on­
ramp,
when
combined
with
the
extreme
accelerations
required
to
achieve
highway
speeds
due
to
the
steep,
uphill
grades
of
the
ramp
and
I­
95
approach
to
the
Q­
Bridge,
appear
to
create
a
situation
that
is
unique
from
other
areas
in
New
Haven.
This
is
evidenced
by
available
ambient
measurements
from
three
other
monitoring
sites
in
the
area
(
i.
e.,
195
Oleander
Avenue
at
the
previous
West
Haven
toll
plaza,
Woodward
Fire
House,
and
State
Street;
see
Figure
8)
that
are
also
located
in
the
immediate
vicinity
of
high
traffic
interstate
highways.
For
29
the
6­
month
period
from
April
through
September
2003,
measured
values
at
those
sites
ranged
from
1.5
to
3.8
ug/
m3
less
than
the
15.7
ug/
m3
average
recorded
at
Stiles
Street
(
see
Figure
8
of
this
TSD).
The
unique,
localized
influences
in
the
Stiles
Street
area
are
the
likely
cause
of
this
2
to
4
ug/
m3
increment
that
results
in
PM2.5
values
at
Stiles
Street
exceeding
the
annual
standard.
When
these
microscale
influences
are
considered
together
with
the
fact
that
the
Stiles
Street
monitor
does
not
meet
EPA's
description
of
a
community­
oriented
monitoring
site
(
i.
e.,
beyond
the
influence
of
a
single
source,
with
a
neighborhood
to
urban
scale
of
representation),
CTDEP
concludes
that
data
from
this
site
should
not
be
used
to
determine
compliance
with
the
annual
PM2.5
NAAQS.

EPA
provides
additional
elaboration
on
classification
scales
later
in
the
same
network
design
guidance
document
(
see
section
2.3.3
and
2.3.4
of
Doc.
A).

"
2.3.3
Monitoring
Networks
PM2.5
monitoring
networks
may
be
new
networks
or
part
of
existing
networks.
Additional
sites
may
be
added
to
existing
networks
according
to
this
guidance.

 
State
and
Local
Air
Monitoring
Stations
(
SLAMS):
SLAMS
are
designed
and
operated
by
local
air
pollution
control
districts
to
determine:
1)
the
highest
concentrations
expected
to
occur
in
each
MPA;
2)
representative
concentrations
in
areas
of
high
population
density;
3)
the
impact
on
ambient
pollution
levels
of
significant
sources
or
source
categories;
4)
general
background
concentration
levels;
5)
the
extent
of
regional
pollutant
transport
among
populated
areas,
and
6)
welfare­
related
impacts
in
rural
and
remote
areas
(
i.
e.,
visibility
impairment
and
effects
on
vegetation).
Only
population­
oriented
SLAMS
acquire
data
for
determining
compliance
with
PM2.5
standards,
and
community­
oriented
(
core)
SLAMS
acquire
data
for
compliance
with
the
annual
PM2.5
standard."

CTDEP
records
indicate
that
the
Stiles
Street
PM2.5
monitor
has
been
classified
as
a
"
peak
concentration"
SLAMS
site,
not
as
a
"
population­
oriented"
site.
Therefore,
consistent
with
the
above
language
from
EPA's
guidance,
the
site
is
not
appropriate
for
determining
compliance
with
the
PM2.5
NAAQS.

Similarly,
Appendix
D
to
40CFR58
(
see
section
2.8.1.2.3
of
Doc.
C)
states:

" 
PM2.5
data
collected
from
SLAMS
and
special
purpose
monitors
that
are
representative,
not
of
area­
wide
but
rather,
of
relatively
unique
populationoriented
microscale,
or
localized
hot
spot,
or
unique
population­
oriented
middlescale
impact
sites
are
only
eligible
for
comparison
only
to
the
24­
hour
PM2.5
NAAQS.
However,
in
instances
where
certain
population­
oriented
micro­
or
middle­
scale
PM2.5
monitoring
sites
are
determined
by
the
EPA
Regional
Administrator
to
collectively
identify
a
larger
region
of
localized
high
ambient
PM2.5
concentrations,
data
from
these
population­
oriented
sites
would
be
eligible
for
comparison
to
the
annual
NAAQS."
30
As
a
"
peak
concentration"
SLAMS
monitor
(
i.
e.,
not
population­
oriented")
sited
at
a
localized
hot­
spot,
this
CFR
excerpt
corroborates
that
the
Stiles
Street
monitor
should
not
be
used
for
24­
hour
NAAQS
compliance
determination.

Section
2.3.4
later
adds
to
the
above,
providing
a
definition
of
community­
oriented
(
core)
sites
that
does
not
encompass
the
Stiles
Street
monitor
due
to
the
fact
that
it
is
clearly
"
located
within
the
microscale
or
middle­
scale
zone
of
influence
of
a
specific,
nearby
particle
emitter";
therefore,
the
monitor
should
not
be
used
to
determine
annual
PM2.5
NAAQS
compliance:

"
2.3.4
Site
Types
Several
types
of
sampling
sites,
not
all
of
which
are
designated
for
determining
compliance
with
NAAQS,
will
be
part
of
the
PM2.5
measurement
networks.

 
Community­
Oriented
(
Core)
Sites:
Community­
oriented
sites
are
located
where
people
live,
work,
and
play
rather
than
at
the
expected
maximum
impact
point
for
specific
source
emissions.
These
sites
are
not
located
within
the
microscale
or
middle­
scale
zone
of
influence
of
a
specific,
nearby
particle
emitter.
Communityoriented
sites
may
be
located
in
industrial
areas
as
well
as
and
in
residential,
commercial,
recreational,
and
other
areas
where
a
substantial
number
of
people
may
spend
a
significant
fraction
of
their
day."

Later
in
Section
2.3.4,
EPA
defines
"
daily
compliance
sites":

"
 
Daily
Compliance
Sites:
Daily
compliance
sites
are
used
to
determine
NAAQS
compliance
for
the
24­
hour
(
daily)
PM2.5
standard,
but
not
for
the
annual
standard.
Because
a
daily
compliance
site
does
not
necessarily
represent
community­
oriented
monitoring,
it
may
be
located
near
an
emitter
with
a
microscale
or
middle­
scale
zone
of
influence.

The
PM
monitoring
regulations
state
that
any
population­
oriented
site
is
eligible
for
comparison
to
the
24­
hour
PM2.5
standard.
If
the
monitoring
site
is
also
representative
of
community­
wide
air
quality,
it
is
eligible
for
comparison
to
the
annual
PM2.5
NAAQS.
With
a
few
anticipated
exceptions,
almost
all
sites
in
the
new
network
will
be
population­
oriented.
A
site
may
be
population­
oriented
and
at
the
same
time
be
source
oriented
or
reflective
of
maximum
concentration.
The
same
is
true
for
the
existing
PM10
network.

Population­
oriented
sites
may
be
located
in
hot
spot
locations
and
other
portions
of
the
above
areas
which
are
likely
to
invoke
exposure
to
fine
particles
for
at
least
part
of
a
24­
hour
sampling
period.
Hot
spot
locations
have
a
micro
or
middle
measurement
scale
of
representativeness.
Microscale
means
that
the
24­
hour
measurements
should
vary
by
no
more
than
±
10%
within
a
circle
of
diameter
100
meters.
Middle
scale
means
that
the
24­
hour
measurements
should
vary
no
more
than
±
10%
within
a
circle
of
diameter
100­
500
meters.
These
distances
are
the
area
around
the
monitor
which
may
be
different
than
the
distance
to
the
nearest
major
influencing
source."
31
Although
CTDEP
wouldn't
agree,
it
is
possible
that
the
above
description
of
daily
compliance
sites
could
be
interpreted
to
include
the
Stiles
Street
monitor.
If
this
argument
is
conceded,
data
from
this
site
could
be
used
for
24­
hour
PM2.5
NAAQS
compliance
determinations.
In
any
case,
though,
EPA
guidance
is
clear
that
data
Stiles
Street
data
should
not
be
used
to
determine
annual
PM2.5
NAAQS
compliance
due
to
its
peak
concentration
objective
and
siting
in
an
area
not
representative
of
community
exposure.

Appropriate
scales
of
representativeness
for
PM2.5
compliance
monitoring
are
also
addressed
in
Appendices
D
and
E
of
40CFR58
(
Doc.
C
and
Doc.
D,
respectively,
as
defined
at
the
beginning
of
Section
5
of
this
TSD).
The
EPA
guidance
discussed
earlier
is
clear
in
stating
that
monitors
intended
for
determining
annual
NAAQS
compliance
should
represent
the
neighborhood
scale
exposure
of
populations
and
not
the
highest
concentrations
in
a
non­
populated
area
measured
by
a
source­
oriented
monitor
such
as
Stiles
Street.
This
point
is
echoed
in
Appendix
D
of
40CFR58
(
see
Section
2.8.1.2.2
of
Doc.
C),
which
states:

"
2.8.1.2.2
Comparisons
to
the
PM2.5
NAAQS
may
be
based
on
data
from
SPMs
in
addition
to
SLAMS
(
including
NAMS,
core
SLAMS
and
collocated
PM2.5
sites
at
PAMS),
that
meet
the
requirements
of
§
58.13
and
Appendices
A,
C
and
E
of
this
part,
that
are
included
in
the
PM
monitoring
network
description.
For
comparison
to
the
annual
NAAQS,
the
monitors
should
be
neighborhood
scale
community­
oriented
locations."

Later
in
the
same
reference
(
see
2.8.1.3.7
of
Doc.
C)
EPA
elaborates,
stating:

"
2.8.1.3.7
Core
monitoring
sites
shall
represent
neighborhood
or
larger
spatial
scales.
A
monitor
that
is
established
in
the
ambient
air
that
is
in
or
near
a
populated
area,
and
meets
appropriate
40
CFR
part
58
criteria
(
i.
e.,
meets
the
requirements
of
§
58.13
and
§
58.14,
Appendices
A,
C,
and
E
of
this
part)
can
be
presumed
to
be
representative
of
at
least
a
neighborhood
scale,
is
eligible
to
be
called
a
core
site
and
shall
produce
data
that
are
eligible
for
comparison
to
both
the
24­
hour
and
annual
PM2.5
NAAQS.
If
the
site
is
adjacent
to
a
dominating
local
source
or
can
be
shown
to
have
average
24­
hour
concentrations
representative
of
a
smaller
spatial
scale,
then
the
site
would
only
be
compared
to
the
24­
hour
PM2.5
NAAQS."

This
excerpt
again
supports
the
conclusion
that
the
Stiles
Street
monitor,
based
on
its
location
immediately
adjacent
to
I­
95
and
the
Stiles
Street
ramp,
should
not
be
used
to
determine
annual
PM2.5
NAAQS
compliance.

Appendix
E
to
40CFR58
(
see
Section
8.3
of
Doc.
D)
provides
information
on
scales
of
representativeness
related
to
spacing
of
PM
samplers
from
nearby
roadways.
Figure
2
in
Appendix
E
(
reproduced
here
on
the
last
page)
depicts
acceptable
distances
for
micro,
middle,
neighborhood,
and
urban
scale
PM10
monitoring,
based
on
the
measured
distance
from
the
edge
of
the
nearest
traffic
lane
presumed
to
have
the
most
influence
on
the
site.
32
EPA
notes
that
this
presumption
is
an
oversimplification
of
usual
urban
settings,
which
normally
have
several
streets
impacting
a
given
site.

It
should
also
be
noted
that
this
section
of
the
CFR
appears
to
only
apply
to
NAMS
monitors
(
the
Stiles
Street
PM2.5
monitor
is
a
SLAMS,
but
not
a
NAMS).
Additionally,
the
associated
figure
was
developed
for
PM10,
and
may
not
be
applicable
for
PM2.5.
(
PM10
has
greater
potential
for
particle
settling,
so
the
15
meter
microscale
cutoff
in
Figure
2
may
be
underestimated
for
PM2.5.)

Notwithstanding
these
caveats,
applying
Figure
2
to
the
Stiles
Street
site,
and
assuming
I­
95
(
average
daily
traffic
in
excess
of
100,000
vehicles
per
day)
is
the
most
influencing
roadway,
CTDEP's
monitor
would
be
representative
of
a
"
middle
scale"
because
it
is
located
somewhat
further
away
from
the
road
than
15
meters.
Applying
the
same
figure
to
the
Stiles
Street
on­
ramp,
assuming
that
accelerating
trucks
and
cars
make
it
the
most
influencing
roadway
despite
lower
traffic
volumes,
would
classify
the
CTDEP
monitor
as
representative
of
a
"
microscale"
due
to
the
less
than
15­
meter
separation
between
the
probe
inlet
and
the
edge
of
the
roadway.
However,
regardless
of
whether
the
Stiles
Street
monitor
is
microscale
or
middle­
scale,
it
is
clear
from
the
figure
that
it
does
not
qualify
for
a
neighborhood
scale
classification.
This
is
crucial
because
of
the
many
excerpts
cited
above
from
EPA
guidance
and
regulations
that
indicate
annual
PM2.5
NAAQS
compliance
should
be
determined
based
on
data
from
monitors
sited
at
neighborhood
scale,
community­
oriented,
locations.
33
