1.
The
level
of
documentation
that
should
be
required
to
substantiate
an
exceptional
or
natural
event
claim.

The
level
of
documentation
should
be
commensurate
with
the
scale
associated
with
the
event.
A
regional
scale
ozone
or
PM2.5
event
will
involve
multiple
monitors
and
may
even
be
visible
by
satellite;
some
states
may
have
corroborating
speciated
or
additional
real
time
data.
A
local
scale
event
may
be
generated
by
locally
high
winds,
and
there
may
be
only
one
met
sensor
available
in
the
area
to
generate
the
data.
The
corroborating
data
should
be
regionally
consistent
for
regional
scale
exceptional
or
natural
events.
All
corroborating
data
should
be
definitive
and
of
high
quality.

2.
The
flagging
of
data
below
the
standard
that
may
affect
the
annual
average
for
an
area.
Should
there
be
an
air
quality
threshold
for
consideration
of
an
exceptional
event
or
natural
event?
Such
as
one
or
two
standard
deviations
from
the
mean
concentration
reported
a
particular
monitoring
site.

No,
this
type
of
criterion
is
arbitrary
and
should
not
be
incorporated
in
the
policy.

If
this
type
of
approach
is
chosen,
should
it
be
used
to
set
Significant
Harm
Levels
for
Emergency
Episode
Plans?

No.
The
Significant
Harm
Levels
should
correspond
to
human
health
effects
and
not
to
whether
monitored
values
are
typical
for
a
given
location.

Should
Natural
Event
Action
Plans
and
Exceptional
Event
Mitigation
Action
Plans
be
made
a
part
of
the
Emergency
Episode
Plans?

No,
the
two
plans
are
designed
for
entirely
different
situations,
and
provide
the
department
with
different
types
of
authority
to
mitigate
the
high
pollutant
concentrations.

3.
The
time
frame
for
State,
local,
and
Tribal
air
quality
agencies
to
flag
and
submit
documentation
to
the
Regional
Office
related
to
an
exceptional
or
a
natural
event,
and
the
time
period
that
Regional
Offices
should
take
action
in
order
to
concur
or
a
non
concur
on
flagged
data
in
the
AQS
database.

States
should
have
at
least
two
years
from
the
natural
event
for
the
submittal
of
these
flags.
Turn
around
times
for
validated
data
from
national
laboratories
can
be
long;
for
example,
at
present,
(
8/
30/
2005)
the
IMPROVE
data
for
2004
is
still
not
finalized
and
in
the
VIEWS
system.
This
data
would
be
critical
in
establishing
a
natural
event
for
PM2.5.
Historically,
toxics
data
validation
may
also
take
a
year
or
more.
There
must
be
time
in
the
process
for
the
event
to
be
identified,
and
for
the
relevant
data
to
be
validated
and
entered
into
a
database,
analyzed,
and
interpreted.
Conclusions
about
causation
should
peer
reviewed
by
data
analysts,
and
disseminated
to
affected
parties,
and
all
of
these
steps
add
time
to
the
process
of
flagging
data
effected
by
natural
events.
4.
Should
there
be
a
bright
line
test
associated
with
high
wind
events,
or
should
we
continue
to
make
decisions
concerning
these
types
of
events
by
requiring
States
to
submit
documentation
to
show
that
high
wind
events
should
be
considered
as
unusual
and
not
typical
for
an
area.

The
current
system
is
preferable
to
a
bright
line
test.
The
state
incorporates
typical
wind
fields
into
its
air
dispersion
modeling
analysis.
When
a
question
of
whether
or
not
a
high
wind
event
is
reasonably
controllable
under
the
SIP
arises,
a
comparison
of
the
met
data
during
the
event
to
the
climatological
met
fields
used
for
modeling
should
suffice
to
determine
whether
the
meteorology
during
an
event
is
atypical
for
an
area.
