Policy
Relevant
Science
Questions
Regarding
PM
Precursors
Prepared
by
Spyros
Pandis,
CMU;
David
Allen,
University
of
Texas
at
Austin;
Armistead
(
Ted)

Russell,
Georgia
Institute
of
Technology;
and
Paul
A.
Solomon,
US
EPA,
ORD
Today:
In
the
context
of
today's
atmosphere;
today's
PM
concentrations,
composition,
and
atmospheric
chemistry.

Based
on
measurements
and
observed
atmospheric
processes
around
the
country,
and
based
on
receptor
modeling
and
air
quality
modeling
analysis:

Atmospheric
processes
and
emission
sources:

I.
Do
VOC
and
NOx
emissions
contribute
directly
and
significantly
to
PM
formation
of
nitrates,
and
primary
and
secondary
organic
aerosols,
and
if
so
how?
­
On
what
scales,
regional
vs.
urban
area
(
transportation
planning
areas)?
­
Under
what
meteorological
conditions
(
time
of
year
and
typical
weather
conditions)?
­
How
does
this
contribution
differ
across
the
country?

Primary
aerosols
are
particles
emitted
directly
from
their
sources
(
or
are
condensed
quickly
to
the
particle
phase
upon
leaving
a
high­
temperature
source),
while
secondary
aerosols
are
formed
in
the
atmosphere
through
chemical
reactions.
Secondary
aerosols
may
be
stable,
such
as
ammonium
or
sodium
sulfate
and
remain
in
the
particle
phase
under
ambient
conditions,
or
they
may
be
semi­
volatile
and
have
mass
fractions
in
both
the
particle
and
gas
phases,
such
as
ammonium
nitrate
and
a
number
of
organic
compounds.
The
distribution
among
the
phases
is
based
on
thermodynamics
and
depends
on
variables
such
temperature
and
relative
humidity.

The
following
discussion
focuses
mostly
on
Pittsburgh
and
the
NE
US,
Atlanta,
representing
the
Southeast
and
Houston
as
examples
of
three
different
regions
of
the
country.
As
applicable,
additional
notes
are
included
about
western
states,
often
with
an
emphasis
on
California
where
two
Supersites
Projects
were
conducted.

Contribution
to
Aerosols
The
larger
VOCs
(
more
than
6
carbon
atoms)
are
the
precursors
of
secondary
organic
aerosols
and
NOx
is
the
precursor
of
nitrates.
A
fraction
of
the
NOx
is
converted
to
nitric
acid,
which
then
reacts
with
gas
phase
ammonia
to
form
particle
phase
ammonium
nitrate
or
reacts
with
other
particle
phase
bases,
such
as
metal
oxides.
In
PM2.5,
most
of
the
nitrate
is
ammonium
nitrate,
while
in
PMc
most
of
the
nitrate
is
associated
with
metals,
such
as
Ca(
NO3)
2.
A
number
of
organic
compounds
undergo
chemical
oxidation
to
form
less
volatile
species,
which
may
be
non­
volatile
or
semi­
volatile.
So,
there
is
a
direct
link
between
selected
VOCs
(
mainly
monoterpenes,
sesquiterpenes,
and
aromatics)
and
SOA
and
also
NOx
to
nitrate.
VOCs
do
not
contribute
to
primary
organic
aerosols.

While
all
the
nitrates
can
be
attributed
to
NOx,
the
fraction
of
the
observed
organic
PM
that
started
its
atmospheric
lifetime
as
VOCs
is
the
subject
of
debate.
Estimates
from
Pittsburgh
are
that
on
an
annual
average
basis
around
20%
of
the
organic
aerosol
was
secondary.
For
summer
pollution
episodes1
around
60%
of
the
organic
aerosol
was
secondary.
Similar
results
were
found
for
Houston,
though
there
is
evidence
that
a
larger
fraction
may
be
secondary
in
Atlanta
during
the
summer.
There
appears
to
be
a
significant
fraction
of
SOA
in
the
winter
in
Atlanta
as
well,
due
to
the
evergreen
forests
and
continued
anthropogenic
emissions.
Approximately
half
of
the
SOA
in
Pittsburgh
was
due
to
natural
sources
of
VOCs
and
half
was
due
to
anthropogenic
sources.
For
Houston,
especially
at
rural
sites,
most
of
the
SOA
during
the
summer
has
been
attributed
to
biogenic
precursors,
and
modeling
studies
suggest
that
monoterpene
reactions
with
ozone
are
dominant
reaction
pathway
for
the
formation
of
biogenic
SOA.
Isotopic
studies
suggest
that
a
large
fraction
(~
80%
or
more)
of
the
SOA
in
Atlanta
is
biogenic.
These
findings
suggest
that
ozone
mitigation
strategies
may
have
some
effectiveness
in
reducing
SOA
formation.
However,
as
with
the
formation
of
nitrate
(
i.
e.,
oxidation
of
NOx
to
nitric
acid),
secondary
organic
aerosol
formation
may
occur
at
ozone
levels
well
below
the
NAAQS
for
ozone.
For
nitrate
this
is
evidenced
by
the
large
amount
of
nitrate
observed
in
Central
California
in
the
winter
during
extended
periods
of
fog.

Sulfate,
nitrate,
and
organic
carbon
are
all
observed
in
the
PMc
fraction
of
the
aerosol.
Sulfate
and
nitrate
are
often
associated
with
metals,
primarily
crustal
related
elements
or
from
sea
salt.
A
fraction
is
from
primary
emissions,
such
as
sulfate
in
sea­
salt,
however,
gas
phase
sulfur
and
nitrogen
acids
also
react
with
the
basic
metal
oxides
in
soil
dust
to
form
metal
sulfate
or
nitrate
species,
which
can
be
considered
as
secondary
aerosols
for
sulfate
and
nitrate.
The
dust
is
blown
into
the
air
by
wind
or
other
processes,
e.
g.,
motor
vehicle
movement.
Organic
aerosols
in
PMc
are
mainly
due
to
biological
materials,
such
as
pollen
and
plant
debris,
but
also
include
some
SOA.

Regional
vs.
Urban
Component
Pittsburgh,
Atlanta
and
Houston
results
suggest
that
there
is
a
strong
regional
component
for
secondary
organic
aerosol
formation.
VOCs
react
over
several
days
under
stagnant
conditions
and
can
react
to
form
aerosols
during
transport.
As
well,
SOA
are
predominately
found
in
the
fine
fraction,
which
can
be
transported
a
1000
km
or
more.
During
the
winter,
SOA
formation
is
much
lower,
so
in
urban
areas
primary
organic
aerosols
dominate.
In
the
northeast
US
during
the
summer,
PM
nitrate
has
both
local
and
regional
contributions,
although
nitrate
contributions
to
fine
PM
are
much
lower
in
the
summer.
During
the
winter
the
regional
component
of
the
nitrates
dominates
and
nitrate
concentrations
are
much
higher
since
ammonium
nitrate
particle
formation
is
favored
in
1
Episodes
are
one
or
multiple
days
with
periods
of
high
pollution,
in
the
east
usually
associated
with
stagnant
air
masses
and
high
temperatures
and
photochemical
processing
largely
drives
the
pollution.
the
winter.
For
southeast
Texas,
nitrate
concentrations
were
relatively
minor
contributors
to
PM
mass,
typically
accounting
for
~
5%
of
total
mass
loading
on
an
annual
average
basis
(
approximately
10%
of
total
mass
loading
during
winter
months).
In
Atlanta,
sulfate
is
largely
regional,
and
dominates
summertime
PM.
Much
of
the
OC
is
also
regional,
though
there
is
a
significant
elevation
of
OC,
and
even
more
so
EC,
in
urban
areas.
There
appears
to
be
more
of
a
pronounced
increase
in
EC
in
the
urban
areas
of
the
Southeast
than
in
Pittsburgh,
where
EC
had
more
regional
characteristics.

In
Central
and
Southern
California,
PM2.5
nitrate
also
peaks
in
the
winter
and
in
the
early
morning
summer
and
winter.
In
the
summer,
much
of
the
nitrate
is
forced
into
the
gas
phase
due
to
the
higher
temperatures
and
lower
relative
humidity.
Transport
of
precursors
in
Southern
California
from
west
to
east
with
nitric
acid
formation
along
the
way
also
results
in
very
high
nitrate
levels
in
the
eastern
end
of
the
valley,
which
are
exacerbated
due
to
very
high
ammonia
levels
in
the
eastern
half
of
the
basis.
In
Central
California,
urban
concentrations
of
nitrate
are
higher
than
surrounding
regions
during
pollution
episodes
in
the
winter;
however,
the
entire
valley
typically
is
driven
by
meteorology
with
periods
of
dispersion
resulting
in
lower
concentrations
and
periods
of
stagnation
in
higher
concentrations.

Meteorological
Conditions
and
Nationwide
Variation
The
SOA
production
is
more
significant
during
pollution
episodes
and
in
general
during
sunny
warm
days.
For
Pittsburgh
and
Houston
it
was
estimated
that
SOA
was
30%
and
50%,
respectively,
of
the
organic
PM
during
the
summer
and
only
10%
during
the
winter.
As
mentioned
above,
SOA
in
Pittsburgh
peaks
during
summertime
episodes
at
60%
with
annual
average
values
around
20%.
The
nitrates
on
the
other
hand
have
higher
concentrations
in
the
NE
US
during
the
winter.
During
the
summer,
their
concentrations
peak
during
the
night
since
ammonium
nitrate
evaporates
during
the
day.
In
general,
since
ammonium
nitrate
formation
is
driven
thermodynamically,
assuming
sufficient
nitric
acid
and
ammonia
are
present,
ammonium
nitrate
peaks
before
sunrise,
during
cooler
temperatures
and
higher
humidity
with
evaporation
from
the
particle
phase
during
the
day.
In
California
and
several
other
western
cities,
ammonium
nitrate
is
a
significant
fraction
of
the
aerosol
all
year
long,
although
it
peaks
in
the
winter
and
early
morning
hours.
Higher
nitrate
in
the
west
is
due
to
lower
levels
of
SO2,
large
amounts
of
NOx,
and
relatively
high
levels
of
ammonia.

II.
How
does
the
contribution
of
SO2
emission
to
secondary
PM
formation
differ?
­
By
scale,
regional
to
urban?
­
By
time
of
year
and
weather
pattern?
­
By
region
of
the
country?

This
is
a
relatively
straightforward
process
as
most
of
the
existing
sulfate
is
the
result
of
the
reactions
of
emitted
SO2.
There
are
differences
in
the
details
of
the
formation
(
e.
g.,
formation
in
clouds
versus
formation
in
cloud
free
air)
but
the
overall
path
from
emitted
SO2
to
sulfate
does
not
change
much.
The
details
determine
the
response
of
the
sulfate
to
changes
in
SO2
emissions
(
proportionality
or
lack
there­
of).
In
general,
the
relationship
deviates
from
proportionality
in
days
with
significant
cloud
cover
and
close
to
the
SO2
sources.
It
becomes
close
to
proportional
in
cloud­
free
environments
and
far
from
the
SO2
sources.

In
the
eastern
US
during
the
summer,
sulfate
followed
by
organic
carbon
are
usually
the
major
components
of
PM2.5.
A
considerable
fraction
of
the
sulfate
found
in
Pittsburgh
and
other
NE
cities
is
regional
in
nature.
In
the
winter,
sulfate
concentrations
are
lower,
about
equal
to
nitrate
and
organic
carbon.
In
the
western
US,
sulfate
is
usually
a
smaller
fraction
of
the
PM2.5
mass
usually
with
concentrations
well
below
nitrate
and
organic
carbon.

In
Houston,
the
fraction
of
PM
mass
due
to
sulfates
(~
40%
including
the
mass
of
ammonium
ion)
was
relatively
constant
throughout
the
year,
with
both
regional
and
local
components.
In
many
western
states,
such
as
California,
sulfate
is
considerably
lower
than
nitrate
and
organic
carbon,
even
during
the
summer.
This
is
due
to
considerably
lower
emissions
levels
for
SO2
in
western
states.

In
the
Southeast,
sulfate
is
much
higher
in
the
summer
than
in
the
winter
such
that
in
the
summer
it
is
usually
the
largest
component
of
PM2.5,
but
on
an
annual
basis,
OC
is
greater
because
sulfate
goes
down
markedly
in
the
fall­
spring
seasons.
Nitrate
definitely
increases
in
the
winter,
but
is
seldom
a
major
contributor.

III.
Do
VOC
and
NOx
emissions
contribute
indirectly
and
significantly
to
secondary
sulfate
formation,
and
if
so
how?

Secondary
sulfate
formation
requires
oxidants,
mainly
OH
for
the
cloud­
free
formation
and
hydrogen
peroxide
for
the
in­
cloud
formation.
The
VOCs
and
NOx
determine
the
intensity
of
the
photochemistry
in
the
atmosphere
so
they
affect
the
levels
of
OH
and
H2O2.
Note
that
this
effect
of
VOC/
NOx
is
much
weaker
than
the
effect
that
these
pollutants
have
on
ozone.
The
effect
is
estimated
to
be
relatively
small
as
shown
in
the
following
table
from
the
NARSTO
PM
Science
Assessment).
Table
1.
Typical
Pollutant/
Air
Quality
Problem
Relationships
IV.
What
role
do
NH3
emissions
play
in
secondary
PM
formation?
Are
anthropogenic
contributions
significant?

In
the
US
according
to
all
the
existing
inventories
most
of
the
NH3
is
due
to
animal
feeding
operations
and
agricultural
activities.
Adding
the
contributions
of
industrial
sources,
the
anthropogenic
contributions
dominate.

Ammonia
is
the
predominant
gas
phase
base
in
the
atmosphere
and
provides
much
of
the
neutralizing
potential
for
acid
gases.
While
sulfuric
acid
moves
directly
to
the
particulate
phase
after
its
formation
(
even
in
the
absence
of
ammonia),
ammonium
sulfate,
ammonium
bisulfate,
and
ammonium
nitrate
require
ammonia
to
form
from
their
gas
phase
precursors,
sulfuric
acid
and
nitric
acid,
as
ammonia
is
one
of
the
two
reactants
required
in
each
case
to
form
the
salt.
However,
sulfuric
acid
reacts
more
readily
with
ammonia
to
form
the
ammonium
sulfate
salt
than
does
nitric
acid.
Thus,
ammonium
nitrate
formation
requires
sufficient
ammonia
to
first
neutralize
the
sulfuric
acid.
Therefore,
SO2
reductions
in
the
eastern
US
may
result
in
lower
than
expected
reductions
of
PM2.5
mass
as
some
of
the
sulfate
not
formed
may
be
replaced
by
additional
nitrate
since
more
ammonia
will
be
available
for
reactions
with
nitric
acid.
This
result
has
been
confirmed
by
recent
modeling
studies
in
Pittsburgh.
For
SO2
oxidation,
ammonia
affects
only
indirectly
the
formation
of
sulfates
by
increasing
the
pH
of
clouds
in
some
areas
and
accelerating
the
aqueous
phase
reactions
that
occur
in
clouds.
Overall
this
effect
is
small
(
Table
1)
for
the
Eastern
US.

V.
Are
anthropogenic
VOC
emissions
a
substantial
contributor
to
primary
and
secondary
PM
formation?

The
VOC
emissions
are
not
expected
to
contribute
significantly
to
primary
PM
formation,
since
VOC
are
by
definition
gas
phase
species.
Their
contribution
to
the
SOA
is
still
an
issue
of
active
debate
as
described
above
in
the
response
to
the
first
question.
However,
VOC
controls
may
impact
indirectly
primary
PM
emissions
(
e.
g.,
primary
OC
emissions
may
also
change
due
to
VOC­
motivated
combustion
controls).

For
the
NE
US,
the
existing
estimates
for
SOA
are
that
less
than
30%
of
the
organic
PM
on
annual
basis
(~
20%
described
above
for
Pittsburgh)
is
anthropogenic.
Even
if
all
the
SOA
were
anthropogenic
(
unlikely)
the
VOC
emissions
would
be
responsible
for
less
than
30%
of
the
organic
PM.
For
specific
air
pollution
episodes
the
contributions
of
anthropogenic
SOA
could
be
higher
and
up
to
50%
or
so.
Chemical
Transport
Models
like
PMCAMx
are
suggesting
that
roughly
half
of
this
SOA
is
anthropogenic
in
the
NE
US
but
this
estimate
is
quite
uncertain
(
estimates
range
from
10
to
70%)
.
In
Texas,
during
the
summer,
these
models
suggest
that
the
majority
of
the
SOA
is
due
to
the
reactions
of
biogenic
monoterpenes
with
ozone,
but
during
the
winter,
this
pathway
will
be
negligible.
Likewise,
in
the
Southeast,
isotopic
analysis
along
with
CMAQ,
PMCAMx
and
URM
modeling
suggest
that
biogenic
VOC
emissions
contribute
the
majority
of
SOA.

VI.
What
are
the
relative
benefits
of
reducing
SO2
emissions,
NH3
emissions,
VOC
emissions,
and/
or
NOx
emission
in
reducing
PM
mass?

The
general
responses
of
the
concentrations
to
changes
in
pollutant
emissions
are
shown
in
Table
1.

The
magnitudes
of
the
effects
(
and
sometimes
the
directions
of
the
changes)
are
expected
to
be
rather
different
depending
on
the
area
of
the
US
(
East
versus
West,
etc.)
and
the
season.
For
example,
the
analysis
of
the
results
in
Pittsburgh
suggested
that
in
that
area
the
wintertime
nitrate
can
be
reduced
either
with
reductions
in
NOx
or
NH3,
while
the
summertime
is
limited
by
ammonia
availability
(
there
is
plenty
of
gas­
phase
nitric
acid
available),
thus,
reductions
in
ammonia
would
be
most
beneficial.

With
regards
to
SO2
reductions
to
PM
mass,
the
full
benefit
may
not
be
realized
since
SO2
reductions
will
reduce
ammonium
sulfate
levels,
but
the
associated
ammonia
will
be
available
for
reaction
with
nitric
acid
to
from
ammonium
nitrate,
thus
nitrate
levels
will
likely
increase.

Analysis
of
SOA
formation
in
Houston,
Los
Angeles,
and
Pittsburgh
suggests
that
strategies
designed
to
reduce
ozone
concentrations
will
result
in
relatively
modest
levels
of
SOA
reduction.

In
the
Southeast,
reducing
SO2
emissions
will
have
a
major
impact
on
sulfate
levels,
though
less
than
proportional
to
the
reductions.
However,
one
gets
the
biggest
benefits
on
the
days
with
the
highest
sulfate
levels,
so
SO2
controls
will
have
the
largest
benefits
on
high
PM
days.
Modeling
results
in
the
Southeast
find
some,
but
lesser
benefits,
from
reducing
NH3,
NOx
or
VOC
emissions.
Reducing
NH3
will
lower
the
amount
of
ammonium
somewhat
less
than
proportionally,
result
in
a
small
decrease
in
sulfate,
and
while
a
relatively
larger
fraction
of
nitrate
will
be
removed,
nitrate
is
a
more
minor
constituent.
Reducing
NOx
emissions,
like
in
the
Northeast,
does
reduce
nitrate,
but
much
less
than
proportionally,
with
small
impacts
on
ammonium
and
sulfate,
and
has
a
minimal
impact
on
OC.

VII.
Expression
of
levels
of
confidence
in
understanding,
and
the
availability/
reliability
of
policy
analysis
tools.
­
How
well
do
we
understand
these
processes?
­
How
good
are
our
emissions
estimates?
­
How
well
can
we
model
and
project
the
impacts
of
emissions
reductions?

According
to
the
recent
NARSTO
PM
Science
Assessment
"
several
strategy
development
tools
are
available
utilizing
analysis
(
e.
g.,
receptor
modeling)
and
simulation
(
e.
g.,
chemical
transport
modeling).
Receptor
models
and
chemical
transport
models
can
be
used
in
a
complementary
fashion
to
develop
advice
for
policy
makers,
as
part
of
a
corroborative
approach
to
providing
guidance
based
on
the
best
scientific
understanding
available.
Receptor
models
are
useful
in
selecting
scenarios
and
identifying
contributing
sources
and/
or
source
types.
Current
chemical
transport
models
are
one
useful
tool
for
guiding
policy
as
part
of
the
collective
scientific
analysis,
being
most
informative
regarding
the
inorganic
fraction
(
sulfate,
nitrate
and
ammonium)
on
regional
and
episodic
(
days
to
weeks)
scales."

The
formation
of
the
major
inorganic
aerosol
components
(
sulfates,
nitrates,
ammonium)
is
relatively
well
understood.
On
the
other
end,
there
are
a
lot
of
remaining
questions
about
the
formation
of
SOA
from
both
anthropogenic
and
biogenic
VOCs.
Our
confidence
in
the
emission
estimates
decreases
from
SO2
(
highest
confidence)
to
NOx,
VOC,
NH3,
primary
carbonaceous
aerosol
(
lowest
confidence).
As
a
result
of
these,
we
are
more
confident
about
the
responses
of
the
inorganic
aerosol
components
to
emission
controls
compared
to
the
effect
of
organic
aerosol
control
strategies.

During
the
eighteen
months
since
the
completion
of
the
NARSTO
assessment
(
see
for
example
Table
1)
our
confidence
and
understanding
have
increased
a
little.
The
most
important
step
for
the
answers
to
all
the
above
questions
will
be
the
evaluation
and
application
of
the
Chemical
Transport
Models.
These
models
integrate
everything
that
we
know
about
atmospheric
aerosols
and
comparisons
with
the
available
wide
range
of
measurements
collected
during
the
last
few
years
due
to
EPA
programs
(
STN,
IMPROVE,
Supersites
Program,
PM
Centers)
will
provide
quantitative
information
about
how
far
we
have
advanced
in
our
effort
to
understand
PM.

VIII.
Are
tools
available
now
for
regional,
State,
and
local
planners
to
make
reasonably
informed
decisions
about
emission
reduction
options
and
their
impacts
on
air
quality?

Yes
(
see
answer
to
the
previous
question.)
CMAQ
and
similar
models
(
e.
g.,
PMCAMX,
URM,
etc.)
can,
and
are,
being
used
to
help
inform
policy
decisions.
While
there
are
still
uncertainties
associated
with
the
use
of
such
models,
at
the
least,
they
should
be
directionally
correct,
and
if
the
model
is
able
to
accurately
reproduce
the
observed
levels
of
individual
compounds,
should
provide
reasonable
quantitative
results
as
well.
Further,
they
provide
an
estimate
of
what
level
of
PM
is
due
to
biogenic
sources.
For
those
species
that
are
accurately
simulated
by
the
model,
their
ability
to
correctly
provide
the
response
to
controls
should
be
similar
to
what
we
now
enjoy
for
ozone.

Receptor
models,
particularly
those
using
molecular
markers,
appear
to
provide
solid
information
as
to
the
sources
of
primary
PM,
though
again
with
some
uncertainties.
They
have
difficulty
distinguishing
between
PM
due
to
diesels
vs.
gasoline­
fueled
engines,
which
is
important
in
a
policy
setting.
Insufficient
analysis
has
been
conducted
to
establish
the
uncertainties
involved
in
receptor
modeling.

At
present,
using
both
receptor
modeling
and
emissions­
based
modeling,
together,
provide
a
strong
foundation
upon
which
to
base
control
strategy
decisions,
particularly
if
the
two
provide
good
results
(
i.
e.,
they
perform
well
against
observations)
and
can
be
reconciled
in
a
meaningful
way.

IX.
How
much
will
our
confidence
and
the
availability/
reliability
of
our
tools
likely
to
improve
over
the
next
3­
5
years?

Significantly
because
of
the
evaluation
of
these
tools
against
the
measurements
collected
by
the
different
monitoring
programs
(
speciation
network,
Supersites
Program,
other
special
studies).
This
process
has
started
and
is
moving
rather
rapidly.
Within
the
next
five
years,
the
various
model
applications
will
lead
us
to
identify
the
current
problems,
and
develop
evaluation
criteria.
Since
the
NARSTO
Assessment,
problems
have
been
identified
and
rectified
in
CMAQ
that
have
significantly
improved
its
ability
to
simulate
species
such
as
nitrate
(
errors
being
reduced
by
an
order
of
magnitude).
Projects
are
underway
to
quantify
uncertainties.
Our
experience,
and
more
broadly
the
experience
by
individuals
applying
the
models
in
a
regulatory
setting,
will
expose
when
the
models
work
well
and
not,
and
for
what
species.
Comparing
this
to
ozone
modeling,
the
community
understands
that
the
models
are
not
perfect,
and
there
are
times
when
the
models
do
not
provide
reasonable
results.
The
community
has
developed
ways
to
deal
with
these
uncertainties
and
problems.
Also,
over
the
next
five
years,
the
natural
experiment,
where
SO2
emissions
should
have
come
down
even
further,
we
will
be
able
to
examine
that
aspect
of
the
model
directly:
do
we
get
the
correct
response
for
sulfate,
nitrate
and
ammonium.
A
problem
with
saying
how
much
better,
is
that
we
do
not,
at
present,
have
a
quantitative
estimate
of
the
current
model
uncertainties.
Within
the
next
five
years,
one
can
expect
that
the
uncertainty
in
simulating
sulfate
should
be
reduced
by
about
half,
primarily
from
conducting
enough
applications
to
establish
confidence
in
that
aspect
of
the
model.
The
uncertainty
in
our
ability
to
simulate
nitrate
will
remain
location
and
time
specific.
In
areas
that
are
ammonium
nitrate
rich,
and
the
emissions
of
ammonium
and
NOx
relatively
well
known,
we
are
already
in
reasonable
shape.
In
other
areas,
one
can
expect
the
models
to
go
from
being
somewhat
unreliable
when
used
in
a
non­
research
application,
to
being
considered
to
have
uncertainties
on
the
order
of
maybe
50%,
possibly
better.
Our
ability
to
simulate
ammonium
follows
from
those
two
components.
SOA
is
a
bigger
issue.
We
have
two
very
different
needs
in
this
case.
First
is
a
better
understanding
of
the
science
of
SOA
formation.
Second
is
a
better
understanding
of
the
emissions,
primarily
biogenic.
We
recently
have
made,
and
are
continuing
to
make,
large
strides
in
understanding
the
science
of
SOA
formation.
In
the
next
five
years
this
will
pose
a
relatively
smaller
part
of
the
uncertainty.
We
continue
to
be
frustrated
in
our
ability
to
accurately
estimate
biogenic
emissions,
and
we
will
need
to
do
so
for
an
increasingly
larger
set
of
compounds,
some
of
which
are
difficult
to
measure.
This
will
continue
to
inhibit
our
ability
to
accurately
simulate
OC
levels.

A
second
question
is
how
well
the
models
will
be
able
to
estimate
the
response
to
controls.
If
the
models
are
able
to
reasonably
reproduce
the
observations
of
the
component
and
related
species
(
e.
g.,
precursors),
the
uncertainty
in
the
response
to
controls
is
smaller.
This
is
important
when
considering
SOA.
Much
of
our
uncertainty
in
simulating
SOA
will
be
from
biogenic
emissions.
However,
this
clouds
our
ability
to
identify
how
well
we
are
simulating
anthropogenic
SOA,
and
hence
our
ability
to
assess
how
SOA
will
respond
to
controls.
On
the
other
hand,
we
will
better
understand
the
science
of
SOA
formation,
and
we
have
a
better
feel
for
anthropogenic
VOC
emissions,
so,
in
five
years,
we
should
be
able
to
accurately
simulate
the
impact
of
VOC
controls
on
anthropogenic
SOA.
What
is
required
in
the
mean
time
are
well
designed
field
experiments
to
identify
the
anthropogenic
component
of
SOA
to
better
test
the
models.
Tomorrow:
In
the
context
of
the
atmosphere
of
2010
(
post
mobile
sources
rule
and
interstate
air
quality
transport
rule
reductions
of
44%
of
2000
SO2
emission
levels,
and
37%
of
2000
NOx
levels
 
see
EPA
2003­
2008
Strategic
Plan)

Reducing
SO2
and
NOx
significantly
will
lower
the
acidity
of
the
PM,
though
it
is
often
neutralized
at
this
point
as
it
is.
Nitrate
formation
is
NOx
sensitive
much
of
the
time
(
at
least
in
much
of
the
eastern
US),
particularly
during
the
winter,
and
will
become
more
so
as
SO2
is
reduced.
A
44%
reduction
in
SO2
will
lead
to
the
nitrate
formation
system
becoming
NOx
sensitive
during
much,
if
not
all,
of
the
year
(
this
may
even
be
the
case
in
CA).
With
such
large
reductions
in
SO2
and
NOx
emissions,
the
system
would
not
be
as
sensitive
to
ammonia
emissions.

Such
a
large
drop
in
SO2
and
NOx
will
lead
areas
like
the
Southeast
to
be
dominated
by
OC,
and
from
the
current
studies,
much
of
that
appears
to
be
biogenic
(
though
this
includes
man­
made
fires).
This
should
also
lead
to
most
of
the
areas
in
the
Southeast
reaching
the
15
ug/
m3
standard.

Atmospheric
processes
and
emission
sources
What
are
the
projected
changes
in
PM
concentration
and
composition?
By
region
and
time
of
year?
In
relation
to
current
standards
(
15

g/
m3
annual
and
65

g/
m3
daily),
and
in
relation
to
projected
lowered
standards
(
12

g/
m3
annual
and
35

g/
m3
daily
 
see
current
Draft
Staff
Paper)

Is
SO2
expected
to
continue
being
a
significant
precursor
of
PM,
and
what
are
some
likely
changes
to
current
sulfate
levels
around
the
country?

Yes 
without
a
doubt
in
the
SE
and
likely
the
NE.

Are
VOC
and
NOx
expected
to
be
significant
PM
precursors,
and
how
will
this
change
from
current
conditions
for
different
regions
of
the
country?

NOx
will
likely
gain
in
importance
as
a
precursor
in
the
east.
Anthropogenic
VOC
will
continue
to
play
what
appears
to
be
a
somewhat
small
role,
though
more
definitive
measurements
and
modeling
studies
are
required
to
verify
this.

Is
the
OC/
EC
fraction
projected
to
become
the
dominant
component
of
PM?
What
is
the
VOC
and
NOx
contribution
likely
to
be
for
this
fraction?
How
much
of
this
fraction
will
be
anthropogenic
in
origin?

OC/
EC
will
be
the
major
contributor
in
the
SE,
but
not
dominant,
as
there
will
still
be
substantial
amounts
of
sulfate.
Anthropogenic
VOC
will
lead
to
a
small
fraction
of
the
total
OC
(
probably
less
than
15%).
