►
Description
The 27th Annual CESM Workshop will be a virtual event. Specifically, the Workshop will begin with a full-day schedule on 13 June 2022 with presentations on the state of the CESM; by the award recipients; and two presentations from our invited speakers in the morning, followed by order 15-minute highlight and progress presentations from each of the CESM Working Groups (WG) in the afternoon.
To learn more:
https://www.cesm.ucar.edu/events/workshops/2022/
A
All
right
I'll,
let
other
folks
trickle
in
over
the
next
minute
or
two
but
good
morning.
Everybody
welcome
to
the
27th
annual
csm
workshop
joint
session
between
chemistry,
climate
and
whole
atmosphere.
Working
groups.
My
name
is
nick
davis,
I'm
one
of
the
co-chairs
of
the
whole
atmosphere
working
group.
A
A
Third,
we'll
be
focused
on
science
and
model
evaluation
and,
like
I
said,
we
will
also
have
a
future
plans
and
discussion
at
the
end,
so
the
format
is
going
to
be
10
minute.
Talks
we'll
run
the
talks
together
in
each
little
mini
session
and
then
we'll
have
a
combined
q,
a
and
discussion
at
the
end,
where
you
can
ask
speakers
questions
or
have
a
discussion
either
on
a
specific
talk
or
even
on
the
main
thrust
of
the
session.
A
As
you
can
tell,
our
schedule
is
quite
tight
because
you
didn't
want
to
have
a
poster
session,
so
we
would
really
appreciate
it
if
you
you
keep
on
schedule,
so
that's
it
for
our
introduction.
I
can
turn
it
over
to
our
first
session
chair.
B
Okay,
thanks
nick
okay,
I
will
start
sharing
the
first
session
and
we'll
be
starting
right
at
8
35.
So
I
guess
we
could
start
a
couple
of
minutes
early.
Just
that
gives
us
a
couple
minutes
extra
yeah.
So
our
first
session
is
mostly
all
on
musica
version,
zero
developments
and
overviews,
and
the
first
speaker
is
luisa
emmons,
luisa,
do
you
wanna
start
sharing
and
I
think
we'll
just
start
right
away.
I
assume
that
everybody
is
already
in
the
session
who
wanted
to
join
this
okay,
go
ahead.
C
Great,
thank
you
yeah,
so
I'll
just
give
a
little
update
on
activities
related
to
musica
and
apologies
to
those
of
you
who
have
seen
this
before
probably
many
times,
but
just
to
start
out.
Our
vision
for
musica
is
to
really
develop
a
model
independent
framework
that
will
really
facilitate
swapping
different
chemical
schemes
and
photolysis
schemes
and
aerosol
schemes
coupled
to
an
earth
system
model
and
to
for
studies
of
the
troposphere
through
the
thermosphere,
and
it
were
but
doing
this
development
in
coordination
with
sema.
C
So
a
lot
of
the
activity
that
we
hope
will
really
lead
to
musica
is
still
in
development
and
design,
but
we
have
some
configurations
of
cesm
that
we're
calling
musica
and
are
are
getting
started
on
some
of
our
goals
for
this
infrastructure.
C
So
if
you
want
more
information,
we
have
a
website
on
the
in
the
acom
website
website
and
a
vision
paper
was
published
in
bands
a
couple
years
ago.
C
So
our
goals
for
musica
are
really
to
develop
this
collaboratively
with
the
community
and
to
do
that,
we're
we
have
a
few
ways
the
community
can
be
involved
at
the
moment.
So
we
have
some
released
model
configurations
that
I'll
tell
you
more
about
in
a
minute.
C
We've
also
run
musica
version
zero
and
are
providing
the
output
to
the
community,
so
you
can
just
use
some
existing
runs
and
we'd
also
like
to
encourage
you
to
participate
in
development,
and
we've
established
these
different
working
groups
to
address
different
topics
related
to
music
and
improving
the
model
representation
of
these
webpage
again.
C
So
what
we're
calling
musica
version?
Zero
is
really
a
configuration
of
cam,
chem
or
wacom
in
cesm,
using
the
spectral
element,
dynamical
core
with
the
re
regional
refinement,
variable
resolution
grid.
So
one
of
the
grids
that
we
have
available
in
csm
2.2
is
this
grid
of
14
kilometers
over
the
contiguous
u.s
then
gradually
increasing
to
100
kilometer
resolution
over
most
of
the
globe.
C
C
There
are
also
a
lot
of
tools,
some
tools
available,
so
you
can
create
your
own
variable,
mesh
and
I'll
go
into
that
in
a
minute
as
well.
So
the
simulation
that
we've
run
and
have
provided
output
for
is
using
this
conus
grid
and
it's
available
on
the
end
card.
Gdex
used
to
be
called
the
dash
repository,
but
it's
available
for
anyone
to
download
and
we've
currently
run
2012
and
2013
and
if
you'd
like
additional
years,
we're
happy
to
help.
C
Do
that
and
you
get
welcome
to
get
in
touch
with
us
and
there
are
additional
details
on
the
music
of
wiki
about
the
simulation,
exactly
the
emissions
and
everything
and
what
what's
in
the
output
and
if
you
have
any
questions,
feel
free
to
contact
us.
C
This
is
one
example
sergio
abarra
who's
now
at
noaa,
gml
has
used
the
this
grid
over
south
america
that
was
packed
developed
by
patrick
callahan,
for
studying,
for
comparing
to
wharf,
essentially
looking
more
at
the
meteorology
and
sergio,
has
now
added
chemistry
or
run
this
resolution
in
cam
chem,
and
so
we
have
full
chemistry
simulations
for
this
grid,
starting
out
he's
just
using
the
half
degree
version
of
this
grid
any
30
by
two
but
plans
to
look
at
the
finer
resolution
and
really
to
look
at
the
impact
of
biomass
burning
on
air
quality.
C
Another
grid
that
has
been
developed
is
doosan.
Joe
has
created
this
grid
over
east
asia
that
goes
down
to
seven
kilometers
over
the
korean
peninsula.
With
you
know,
four
levels
of
refinement
gradually
reaching
that
and
has
been
looking
at
how
these
different
resolutions
differently:
simulate
ozone
and
gas-phase
species
and
through
comparison
with
the
chorus
aq
aircraft,
observations
as
well
as
surface
observations,
and
so
the
black
lines
here
is
the
observations.
C
The
black
boxes
are
the
aircraft
observations
and
the
four
colors
are
the
model
resolutions
at
different
model
at
different
resolutions
and
show
that
you
can
get
rather
different
answers
for
at
these
different
resolutions
and
while
some
of
the
courser
resolutions
sometimes
look
better,
that's
probably
not
for
the
right
reason.
C
So
there's
a
lot
to
investigate
there
to
really
understand
the
impacts
of
changing
resolution,
so
we're
trying
to
put
together
a
library
of
all
the
grids
that
people
have
developed
and
that
are
available
and
not
all
of
these
have
been
tested
with
chemistry
but
are
or
have
some
like.
This
arctic
grid
have
been
used
for
cam
studies,
but
we
think
could
certainly
be
used
for
chemistry.
C
There's
a
grid
for
europe,
east
asia,
africa,
australia,
and
so
we
started
this
page
on
our
wiki
site.
That
gives
details
of
what
the
resolution
is
called,
where
you
can
get
it
who
to
contact
whether
it's
been
published,
and
so,
if
you're
working
on
making
a
grid.
We'd
love
to
have
you
share
it
with
us,
and
we
can
include
it
on
the
wiki
page
and
if
you're
interested
in
using
any
of
these
grids,
please
have
a
look
and
contact.
The
people
mentioned
for
more
information.
C
One
other
topic
we've
been
working
on
is
a
model
evaluation
tool
called
melodies
monet,
and
this
is
a
python
based
framework
that
will
facilitate
comparing
model
output
with
atmospheric
chemistry.
C
So
it's
python
based
with
standard
libraries,
and
we
have
a
lot
of
python
notebook
examples,
and
then
it
spits
out
plots
like
this,
which
is
the
model
results
with
the
air.
Now,
observations
overlaid
them
and
something
like
a
box
plot
of
the
statistics
of
the
observations
in
the
model
and
lots
of
other
plots,
and
we
are
definitely
also
would
love
to
have
community
assistance
in
further
developing
this
and
reading
different
models
and
observations
and
and
producing
output.
C
And
so
then,
finally,
over
the
past
year,
we've
ran
a
music
tutorial
series
and
this
is
the
webpage
for
it.
And
so
we
had
a
session
about
each
month
and
showing
how
to
look
at
musica
output
and
running
the
model
and
all
the
videos
and
slides
from
each
session
is
available
online.
C
And
we
encourage
you
to
have
a
look
at
those
if
you'd
like
to
get
started
with
any
of
this
and
provide
us
feedback
if
you'd
like
additional
sessions-
and
we
also
have
a
musica
mailing
list
and
which
you
can
sign
up
on
the
musical
webpage.
If
you'd
like
more
information
and
I'll
stop
there.
If
I
started
two
minutes
early,
but
thanks
thanks.
B
As
nick
has
explained,
we
have
questions
at
the
end
of
the
session,
so
you
can
put
them
in
the
chat.
If
you
have
direct
question
and
luisa
can
answer
them
as
well,
and
while
I'm
introducing
nori
nori,
please
start
sharing
your
screen.
We,
the
next
speaker,
is
nori
marsiskala
and
she's
visiting
inca
right
now.
She
will
talk
about
evaluation
of
model
stimulated
ozone
and
its
precursors
in
musical
version
zero
against
in-situ
airborne
measurements
over
the
continental
yeah
continental.
Us
please
go
ahead.
Nori
thank.
D
You
awesome
thank
you
so
much
simone
and
luisa.
Good
morning.
Everyone,
like
someone
mentioned
my
name,
is
nori
beth
merskill
and
I'm
a
grad
student
visitor
here
at
ncar,
with
the
asp
gvp
program
hosted
by
dr
emmons.
D
The
main
objectives
of
this
project
are
to
understand
the
causes
of
ozone
production
in
areas
of
non-attainment
and
their
impact
on
ambient
air
quality
using
high
resolution
chemistry,
climate
modeling.
Additionally,
this
project
also
aims
to
investigate
the
sensitivity
of
model,
horizontal
grid
resolutions
to
atmospheric
composition
and
chemistry.
D
So,
as
louisa
mentioned
earlier,
the
multi-scale
infrastructure
for
chemistry
and
aerosols
version,
0
or
simply
musica
version
0
is
a
configuration
of
cam
of
the
chemcam
global
chemistry
climate
model
using
spectral
element
dicor,
which
is
that
unstructured
grid
mesh
based
on
a
cubesphere
allowing
for
regional
refinement.
D
The
frappe
and
discover
aq
colorado
campaigns
were
held
jointly
by
ncar
and
nasa
respectively,
in
the
summer
of
2014,
with
the
goals
of
identifying
main
drivers
of
summertime
ozone
in
the
northern
front
range,
metropolitan
area
and
understanding
sources,
transport,
chemical
transformations
of
air
pollutants
and
how
they
relate
to
satellite
observations.
D
D
So
the
early
musica
version,
zero
model
simulations
for
the
summer
of
2014
demonstrated
that
biogenic
leoses
isoprene
amount
of
terpenes
specifically
were
being
overestimated
in
the
model
compared
to
aircraft
observations
from
frappe
with
normalized
mean
biases
of
263
and
46
respectively
average.
You
know
across
the
all
the
altitudes
here
on
the
figure
to
the
left,
isoprene
and
monoterpene
emissions
in
the
northern
front
range
metro
area,
make
up
13
and
21
of
total
colorado
emissions
without
taking
into
consideration
the
emissions
from
the
mountain
region
to
the
left
of
the
nfrma
region
in
colorado.
D
In
response
to
this,
we
decided
to
constrain
the
biogenic
vocs,
isoprene
and
monoterpene
by
reducing
emissions
in
musica
version
zero
by
75
percent,
50
and
25
for
the
summer
of
2014.
In
order
to
quantify
their
impact
on
ozone
production.
The
figures
to
the
right
of
the
slide
show
that
reduction
for
isoprene
and
mono
terpene,
where
figure
a
shows
a
hundred
percent
b
is
75
c
is
50
and
d
is
25
of
biogenics
present
in
the
model.
D
So
once
the
sensitivity
simulations
were
ran,
I
analyzed
the
differences
of
surface
ozone
concentrations
between
the
control
simulation
and
the
sensitivity
simulations
with
that
focus
on
colorado.
D
Based
on
these
plots,
it
can
be
seen
that,
in
comparison
to
the
100
percent
of
biogenics
present,
the
further
reduction
of
biogenic
vocs
in
the
model
leads
to
non-linear
changes
averaged
over
colorado,
with
maximum
changes
of
almost
three
parts
per
billion,
and
I
mean
the
reason
that
I'm
looking
at
this
is
because
of
the
rocky
mountains
and
they're
focused
on
the
isoprene
by
monoterpene
emissions.
D
So
in
order
to
quantify
the
impact
of
biogenic
vocs
on
ozone
production,
I
began
by
evaluating
the
vertical
profiles
of
various
sensitivity,
simulations
against
the
frappe
c-130
aircraft
measurement
and
found
that
isoprene
and
ozone
agreed
best
with
observations
when
there
was
50
presence
of
viogenic
theoc's
monoterpenes,
when
there
was
a
75
percent
present
and
formaldehyde
when
there
was
a
25
presence.
Overall,
isoprene
formaldehyde
and
ozone
concentrations
showed
better
agreement
with
frappe
when
presenting
that
reduction
of
biogenic
emissions,
while
monoterpenes
remain
underrepresented,
particularly
in
the
higher
altitudes
here.
D
In
addition
to
evaluating
the
model
against
aircraft
observations,
I
also
was
looking
at
the
voc
measurements
from
the
uc
irvine
hole.
Air
sampler
data
sets
available
on
the
nasa
and
campaign
website
and
divided
the
data
by
region
in
the
red.
We
have
designated
that
as
the
rocky
mountain
area,
the
blue
square
is
the
foothills.
So
it's
that
region.
That
is
not
quite
the
mountains,
but
not
quite
the
urban
area.
D
D
I
found
that
isoprene,
which
is
the
top
figure
agreed
best
with
observations
when
there
was
typically
between
25
to
50
percent
of
biogenics
present
in
the
model,
when
looking
both
at
the
mean
and
median
biases
for
each
region
for
monoterpenes,
which
is
the
bottom
figure
in
terms
of
the
mean
bias.
D
The
model
agreed
best
when
there
was
a
75
percent
biogenic
vocs
present
in
the
model,
with
differences
up
to
approximately
18
parts
per
trillion
and
for
median
biases
it
was
between
25
and
50
percent,
with
changes
up
to
14
parts
per
trillion.
D
Although
no
concrete
conclusions
can
be
made
at
this
time
about
which
sensitivity
simulation
agrees
best
with
observations,
it
can
be
noted
that
the
bices
for
isoprene
monoterpenes
are
reduced
when
the
biogenic
emissions
in
the
model
are
reduced.
D
So,
additionally,
in
order
to
look
at
the
sensitivity
of
ozone
to
model
grid
resolution,
I
am
refining
a
grid
over
michigan
at
about
seven
kilometers.
I
plan
to
run
music
of
version
zero
simulations
using
this
refined
grid
to
look
at
particularly
the
ozone
atmospheric
chemistry
over
michigan
and
compared
to
the
moose
campaign
measurements
so
moose
or
the
michigan
ontario
ozone
source
experiment
is
led
by
the
michigan
department
of
environment,
great
lakes
and
energy,
with
various
participants
from
universities
to
federal
agencies
and
environment,
climate
change,
canada.
D
This
campaign
seeks
to
define
potential
attainment
strategies
in
southeast
michigan
and
better
understand
what
contributes
to
the
ozone
exceedances
in
this
area.
So
moose
phase
one
began
last
year
from
may
24th
to
june
30th
of
2021
and
moose
phase.
Two
is
continuing
on
this
summer
so
and
the
images
here
just
kind
of
show
the
monitoring
sites
where
some
of
the
sensors
have
been
located.
D
So
in
summary,
of
the
preliminary
findings
of
the
musica
version,
zero
simulation
output
evaluated
against
the
frappe
c-130
flight
track
data
set
determined
that
isoprene
amount
of
terpenes
were
being
overestimated
in
the
model.
So,
with
this
being
said,
we
decided
to
reduce
the
biogenic
emissions
in
musica
by
2550
and
75
percent.
D
The
initial
analyses
with
the
aircraft
and
stationary
data
sets
from
frappe
and
discover
aq
were
also
conducted,
and
then,
although
no
conclusion
can
be
made
at
this
time,
the
initial
model
evaluation
with
aircraft
and
whole
air
sampler
measurement
does
point
to
that.
The
reduction
in
biogenics
in
the
model
will
help
reduce
the
discrepancies
between
the
model
and
the
observations.
D
Further
analysis
is
neater
needed
to
better
evaluate
the
model
and
determine
optimal
biogenic
voc
content.
One
suggestion
has
been
to
potentially
regret
the
actual
vegetation
maps
used
in
the
model.
So
that's
something
that
we're
currently
looking
at
and
then
I
also
plan
to
continue
the
michigan
grid
this
summer
with
end
car.
So
I'm
excited
about
that
and
then
I
just
like
to
acknowledge.
Encar
the
national
science
foundation
and
nasa's
michigan
space
grant
consortium
for
supporting
this
project.
D
B
Thank
you
very
much
very
nice
talk
and
again,
please
put
your
questions
in
the
chat
for
now
and
keep
or
keep
them
for
the
later
discussion.
And
while
I
introduce
winfo,
please
start
sharing
your
screen.
So
the
next
speaker
is
when
futon
from
aecom
and
she
will
talk
about
comparing
the
regional
scale
model.
Variability
of
musica
with
a
regional
model
warframe
over
cloners
go
ahead.
Wendy.
E
Thank
you,
simon.
Can
you
see
my
screen?
E
E
So
the
short
answer
is
yes,
and
the
musica
version
zero's
capability
of
representing
the
original
scale
features,
is
comparable
to
that
of
wolfcam,
which
is
a
widely
used
original
model,
and
to
demonstrate
that
here
we
compare
music,
high
version
0
with
two
wolf
cam
simulations.
E
So,
let's
look
at
the
musical
version,
zero
simulation
on
configuration
first,
so
we
run
musical
version
zero
for
august
and
september
2020
with
a
standard
grid
and
which
has
the
original
refinement
over
corners.
As
shown
here
and
the
model
configuration
is
listed
on
the
right.
So
basically,
we
use
the
thin
version
2.5
for
fire
emissions
epa,
2017
nei
for
emission
for
anthropogenic
emissions
within
corners
and
attempts
volume
5.1
for
anthropogenic
emissions
outside
and
we
use
without
ts-1
and
mem4
for
gas
phase,
chemistry
and
aerosols
beams,
and
also
wind
and
temperature
are
nudged
to
merit2.
E
E
Therefore,
the
comparisons
of
the
absolute
amount
are
not
representative
here
and,
in
addition,
the
goal
of
this
analysis
is
to
understand.
If
music
can
represent
the
original
scale
features.
Therefore,
the
following
results
will
focus
on
the
spatial
and
the
temporal
variability
from
both
model
at
regional
scale,
instead
of
their
absolute
values.
E
So
I
will
only
show
a
few
comparison.
Comparisons
here
today,
for
the
sake
of
time,
and
here
is
precipitation,
I'm
showing
the
distribution
of
the
temporal
variability
in
precipitation
from
musica
on
the
left
and
the
wolf
cam
on
the
right
and
besides
this
west
coast
region,
the
temporal
variability
in
the
two
models
are
similar
and
in
addition,
this
figure
also
shows
that
there
are
issues
on
the
wolfcamp
domain
boundary
because,
as
we
can
see,
the
temporal
variability
near
the
four
edges
are
very
high.
E
So
each
of
each
panel,
a
different
spatial
scale
is
shown
and,
as
we
can
see,
the
spatial
variability
from
both
models
are
comparable.
At
these
scales
and,
interestingly
at
quarter
degree.
Spatial
scale
musica,
has
even
higher
spatial
viability
in
precipitation,
and
here
I'm
showing
the
spatial
variability
of
ozone
in
music
and
the
two
true
wolf
cam
simulations.
E
So
the
top
row
is
for
the
free
troublesphere
and
the
bottom
row
is
for
the
pbr
average
and
similar
as
before.
Each
panel
is
for
the
spatial
variability
at
a
different
spatial
scale,
and
the
first
first
box
is
newly
kind
of
the
rest
on
two
boxes
are
webcam,
so,
as
we
can
see,
musical
kind
of
represents
the
original
scale
variability
in
ozone
reasonably
well
compared
to
the
wolfcam.
E
And
here
is
the
same
figure,
but
for
pm
2.5,
and
the
conclusion
is
also
valid
here
for
pm
2.5,
and
here
I'm
showing
the
distribution
of
temporal
variability
of
ozone.
The
top
row
is
worth
camp.
The
bottom
row
is
mudeka
and
on
the
left
is
free.
Troposphere
on
the
right
is
pbl
average
and
overall,
the
spatial
distribution.
E
The
magnitude
of
the
temporal
variability
from
both
models
are
close,
and
the
difference
here
in
this
region
is
due
to
that.
Musica
has
a
better
representation
of
stratosphere
conditions,
which
is
important
for
stratospheric
intrusion,
and
here
is
the
same
figure,
but
for
pm
2.5
and
the
conclusion
is
overall
valid
here,
but
the
difference
in
this
pacific
northwest
region
is
likely
due
to
the
difference
in
the
fire
emissions,
and
I've
only
showed
a
few
cases
in
the
wolfcam
and
musical
comparison,
and
I
hope
I
convince
you.
E
The
conclusion
is
that
the
music
have
urgency,
or
those
capability
of
representing
the
original
scale
features
is
comparable
to
that
of
wolfcam,
and
I
would
also
like
to
take
this
opportunity
to
share
another
study
that
I've
been
working
on,
which
is
develop,
evaluate
and
also
apply
music
over
africa,
and
here
is
a
model
grid
that
I
created
for
africa,
and
I
it
has
the
original
refinement
of
with
a
quarter
degree
over
the
continent,
and
I've
successfully
run
the
model
for
2017
and
I'm
currently
evaluating
the
the
model
over
africa
with
institute
observations
and
also
satellite
products,
and
here
I'm
also
using
wolfcam
as
a
reference
here,
and
here
are
two
examples
showing
the
comparisons.
E
Long
story
short
to
the
so
far.
The
performance
of
music
over
the
region
is
comparable
to
wolf,
cam
and
I'm
doing
more
more
evaluation
here
and
if
you
are
interested
in
this
study,
please
feel
free
to
contact
me
and
I'll
stop
here
for
the
with
the
take
home
message
and
thank
you
for
listening.
B
Thank
you
very
much,
renfrew
very
nice.
Again.
If
you
have
questions,
please
place
them
in
the
chat.
There's
already
one
for
you
to
answer,
and
then
there
will
be
more
discussion
afterwards
and
while
I
introduce
doug
please
can
you
share
your
screen,
so
the
next
talk
is
by
doc,
innocent
and
also
june
jung
with
the
title
influence
of
asian
summer
monsoon
eastward
shedding
events
on
upper
troposphere,
chemical
composition.
F
Yeah
this
talk
10
minute
talk,
is
going
to
give
a
summary
of
a
study,
we're
working
on
looking
at
the
asm
influence
on
chemical
composition
of
the
utls,
using
a
specified,
dynamic,
ccm
or
musica,
and
it's
in
support
of
the
a-clip
mission,
which
is
a
mis
aircraft
mission.
We're
going
to
have
this
summer
in
august,
based
out
of
korea
called
a-clip,
it's
the
asian
summer,
monsoon
chemical
and
climate
impact
project.
It's
led
by
laura
pan
and
paul
newman,
and
it's
a
two
aircraft
project,
the
g5
nsf
g5
and
the
nasa
wb-57.
F
Oh
there
we
go
again.
The
model
we're
going
to
use
is
musica
and
we've
been
talking
about
that.
This
morning
we
have
two
versions
of
it:
the
one
degree
and
the
quarter
degree
regionally
refined.
I'm
only
going
to
show
results
from
the
one
degree
today
and
the
version
we're
going
to
use
is
the
58
level,
which
has
500
vertical
res
500,
meter
vertical
resolution
up
to
the
upper
troposphere
and
it's
nudged
by
nasa
ameritu,
and
this
model
course
has
detailed,
interactive,
tropospheric
and
stratospheric
chemistry.
F
So
a
really
quick
background
on
the
asian
summer,
monsoon
transport,
the
anti-cyclonic
circulation,
is
set
up
every
year
in
the
northern
hemisphere
summer.
Within
the
asm
surface,
emissions
are
influenced
by
convective
uplifting
to
the
upper
troposphere
and
then
dispersed
into
the
northern
hemisphere,
eastward
eddy,
shedding
by
eastwood
heading
shedding
events,
and
so
we
have
two
simple
questions
here.
We
want
to
know
what
the
magnitude
of
the
eastward
shedding
events
are
on
chemical
species
into
the
free
troposphere
and
two.
F
F
We've
looked
at
about
10
species
so
far,
I'll
show
just
co
and
pan
results
today
and
in
the
next
slide,
I'm
going
to
highlight
some
detail:
some
chemistry
on
pan
and
how
it's
formed
by
animal
emissions
so
quickly,
pan
in
the
in
the
lower
troposphere,
is
thermally
decomposed
and
until
you
get
up
into
a
region
where
the
temperatures
get
colder
and
the
slows
this
process
slows
down.
F
And
then
eventually
it
is
photolyzed
when
you
get
into
the
upper
troposphere,
lower
stratosphere,
but
there's
a
there's,
a
sweet
spot
region
here
where
pan
can
form
and
once
it
forms
it
has
a
lifetime
of
about
one
month.
So
it's
it's
a
good
tracer
of
air
coming
out
of
the
asm,
and
also
it
will
it'll
be
transported
into
the
free
troposphere
where
it
can
release
its
nox
and
eventually
form
ozone,
and
this
has
been
studied
by
other
other
groups
and
the
references
are
below.
F
So
what
are
what
are
these
eastward
shedding?
Events?
Look
like
these
are
two
examples
on
the
left
is
co
and
on
the
right
is
pan.
This
is
a
year
2020,
which
we
had
a
large
number
of
eastward
shedding
events.
F
These
are
anomalous
figures,
anomaly
figures
such
that
you
can
see
that
the
the
colors
here
are
carbon
monoxide
and
or
pan,
and
there
are
events
that
are
going
from
the
bottom,
which
is
middle
of
july
to
the
top
of
the
figure
which
is
the
end
of
august,
and
there
are
one
two
three,
maybe
four
or
five
large
events
in
2020
and
if
you
look
on
the
right,
you
see
the
same
signature
in
pan,
which
is
exciting.
Now
we
know
that
panicking
is
formed
from
animal
emissions,
lightning
and
or
surface
emissions.
F
These
dash
vertical
lines-
and
you
can
see
in
2016,
especially
if
you're
in
august
you're,
not
going
to
see
a
lot
of
large
eastward
shedding
events.
There
are
small
numbers
of
them,
but
it'll
it'll,
just
look
different
you're,
not
gonna,
get
the
same
impact,
so
we're
we're
somewhat
hoping
for
a
2020
like
experience
this
summer,
and
you
see
both
the
low
amount
of
eastward
shedding
co
and
the
low
amount
of
pan
for
this
year.
F
This
is
a
figure
that
ran
smith
met
made,
and
this
is
for
pan,
and
you
can
see
several
things
on
here.
One
the
white
line
is:
is
a
gph
value
that
kind
of
outlines
the
anti-cyclone
of
the
asian
summer
monsoon.
F
The
black
line
is
the
tropopause
at
this
pressure,
altitude
of
150,
hectopascals
and
so
above
the
and
on
the
right
here
are
pdf
values:
don't
worry
about
trying
to
look
at
the
individual
pds,
but
just
look
at
the
mean,
which
are
the
dashed
lines
here.
So
if
you
look
at
the
light
blue
color
here,
this
is
the
stratospheric
side
of
the
asm
anticyclone.
So
it's
on
the
other
side
of
the
tropopause
there's
very
little
amount
of
pan
in
this
region
of
the
atmosphere.
F
If
you
look
in
where
this
dark
blue
mean
is
here,
you
can
see
that
there
is
some
shedding
that
has
spread
down
into
the
tropic
subtropics,
but
but
in
general,
is
pretty
low.
So
this
is
below
it's
not
included
in
the
white
line
of
the
gph
and
it's
below
the
tropopause
and
and
that's
what
the
blue
line
is.
F
Now,
if
you
look
inside
the
asm
anticyclone,
it's
everything
inside
the
white
line
here
the
mean
is
somewhere
around
500
parts
per
trillion
and
if
you
just
want
to
focus
on
the
shedding
region,
which
is
this
region
right
here,
which
is
between
125
east
175
east
inside
the
white
line,
it's
its
mean
is
also
very
high.
In
fact,
this
time
it's
actually
greater
than
the
mean
of
the
whole
asm.
F
So
this
is
the
co
distribution,
histogram
and
again,
the
light
blue
is
outside
the
asm
in
the
stratosphere.
F
The
dark
blue
is
outside
the
esm
and
the
troposphere
the
gold
is
what's
inside
the
asm
everywhere
and
then
the
red
is
only
in
the
shedding
region,
and
there
are
the
horizontal
axis
shows
every
year
that
we
looked
at.
The
red
line
is
showing
the
minimum
value
in
the
shedding
region.
So
right
off
the
bat
you
can
see
that
within
the
inside,
the
asm
is
similar
across
all
years
for
co,
there's
a
little
bit
of
variability,
but
not
not
that
much.
F
If
you
look
at,
however,
what's
in
the
shedding
region
itself,
the
red
bars,
what
you
see
is
the
maximum
years
2017.
F
F
The
difference
between
the
max
and
the
men
is
about
30
percent.
So,
depending
on
which
year
you
pick
you're
going
to
have
about
30
percent,
more
or
less
influence
of
the
asm
eastward
shedding
events
in
the
region
where
we're
actually
going
to
be
measuring
co.
F
F
When
you
look
just
in
the
shedding
region,
which
is
the
red,
the
max
and
min
years,
there's
a
roughly
61
percent
more.
No,
that
can
be
produced
depending
on
which
year
you
pick
and
then
finally,
the
influence
on
pan
same
kind
of
figure.
In
this
case,
the
the
shedding
region
has
up
to
73
percent
more
pan
in
a
year,
that's
very
active
with
eastwood
shedding
events
versus
a
minimum
year
like
2016
or
2018..
F
So
going
into
this,
I
wasn't
really
sure
what
these
values
would
be,
but
you
can
see
that
it
does
matter
and
that
the
dynamics
of
the
year
does
really
set
up
either
a
lot
of
pan
a
lot
of
no
or
nco
or
not.
F
So,
in
summary,
we've
taken
a
look
at
eastwood
shedding
events
and
we've
seen
that
there
they
do
have
a
large
inner
annual
variability.
This
variability
will
impact
the
airborne
missions
like
a
clip.
F
We've
examined
six
years
and
found
significant
variability
in
several
species.
F
Anno
and
pans
show
more
variability
than
what
you
see
in
co
and
we
still
need
to
quantify
the
different
emission
sources
of
nox,
for
example
the
surface
emissions
versus
lightning,
and
one
way
to
do
that
is
we
have
tagged
mechanisms
that
louise
emmons
and
others
have
worked
on,
that
we
could
apply
to
this
problem
and
really
follow
where
the
knox
is
coming
from
from
its
process.
F
The
exporter
pan
will
impact
tropospheric
ozone
production.
This
has
been
discussed
in
multiple
studies
and
we'd
like
to
follow
that
ozone
production
and
quantify
that
in
the
future,
and
we
can
do
some
tagging
on
that
also.
So,
in
summary,
we
have
a
little
bit
more
work
to
do
on
this
project,
but
we
are
looking
forward
to
a
very
exciting
airborne
mission
this
summer.
Thank
you
for
your
attention.
B
Thank
you
very
much.
Doug
very
interesting
talk
again
put
your
questions
in
the
chat
or
wait
for
the
final
discussion.
We
have
one
more
talk
in
this
session
by
ren.
When
could
you
please
start
sharing
your
screen
when
duck
stop
sharing,
yes,
and
then
I
will
introduce
the
last
talk
by
ren
smith
from
aecom
he's
talking
about
evaluating
the
representation
of
asian
summer
monsoon
utils
composition
in
csm,
using
airborne
and
in
situ
observations
when
go
ahead.
G
G
Okay,
awesome,
so
thank
you.
So,
like
simone
said,
my
name
is
ren
smith,
I'm
a
postdoc
working
in
the
acom
lab
at
ncar.
So
the
goal
of
this
work
is
to
characterize
how
cesm
is
representing
the
upper
troposphere
lower
stratosphere
during
the
asian
summer
monsoon,
and
this
work
has
been
really
helpful
for
us
and
helping
prepare
our
end.
Car
model
configuration
for
the
a-clip
field
campaign,
which
doug
just
mentioned
and
nicely
set
the
stage
for
me
on
which
is
going
to
take
place
in
august
from
south
korea
this
summer.
G
So
doug
kind
of
mentioned
actually
a
little
bit
of
the
background
here.
So
I
can
kind
of
gloss
over
that
a
little
bit
quicker,
but
the
basically
the
coupling
between
the
anthropogenic
emissions
in
asia
and
the
deep
convection
that's
associated
with
the
summer.
Monsoon
precipitation
has
been
shown
to
have
a
pronounced
effect
on
the
composition
of
the
upper
troposphere
lower
stratosphere
in
this
region.
G
So
this
mechanism
drew
interest
for
the
stratoclem
aircraft
field
campaign
which
took
place
in
the
summer
of
2017
over
south
asia
and
that's
what
we're
going
to
be
using
for
the
present
evaluation,
mainly
because
we're
waiting
about
two
years
now
for
eclipse
observations
because
of
the
pandemic,
and
so
this
is
a
really
useful
data
set
for
us
over
south
asia.
While
we're
waiting
the
flight
tracks
for
the
strata
club
mission
are
shown
here
in
the
black
lines
there.
G
So
we
evaluate
two
different
configurations
of
cesm.
We
have
a
musica
simulation
which
has
the
horizontal
regional
refinement
over
asia,
and
I
think
we've
seen
this
grid
a
couple
of
times
here
now,
but
this
is
sort
of
the
asia
grid.
That's
been
developed
in
with
musica
for
this
purpose
and
then
we
also
have
a
one
degree,
wacom
simulation,
which
has
a
110
vertical
levels,
and
this
reduces
the
vertical
grid
spacing
in
the
utls
layer
to
about
500
meters.
G
So
what
we
see
is
that
at
both
of
these
levels
here
the
regionally
refined,
music
assimilation
on
the
right
is
lofting
more
co
into
the
utls
than
wacom,
and
we
think
that
that's
a
direct
result
of
the
refined
horizontal
grid.
So
it's
it's
interesting
to
see
this
sort
of
direct
translation
of
this
tropospheric
tracer
and
the
differences
there
just
with
the
difference
in
the
horizontal
grid.
G
G
So
I'm
looking
at
this
observations
and
models
in
a
vertical
profile
space,
specifically
a
tropopause
relative
altitude
space
here
on
the
left,
the
grayscale
pixels
are
showing
the
observations.
There's
a
layer
normalized
relative
frequency,
and
then
we
have
the
music
assimilation
in
red
and
the
wacom
simulation
in
orange.
So
the
solid
lines
are
the
means
and
the
dashed
lines
are
the
5th
and
95th
percentiles
of
the
model
information.
So
it's
important
to
note
here
that
the
model
information
on
this
plot
is
actually
subset
over
the
gray
box.
G
That's
on
the
map
on
the
right
and
that's
during
the
entire
measurement
time
period
and
over
a
much
larger,
so
it's
over
a
much
larger
spatial
and
temporal
region.
So
with
that
in
mind,
the
fact
that
the
general
behavior
of
the
model
and
the
observations
falling
reasonably
in
line
with
each
other
is
is
very
reassuring
that
we're
representing
the
general
behavior
of
this
region
so
to
compare
the
specific
models
to
each
other,
rather
than
just
against
observations.
I
want
to
call
out
two
specific
layers
that
are
on
this
plot.
G
The
first
one
is
approximating
the
primary
level
of
convective
outflow.
So
that's
shown
here
the
layer
that
I'm
choosing
here
subsetting
is
the
gray
dashed
lines
here,
or
it's
bounded
by
the
gray
or
black
dashed
line.
Sorry
the
mean
values
of
the
distribution
are
printed
in
the
legend
here
on
the
distributions
on
the
right.
G
So
we
have
the
observations
in
black
again,
music
and
red
and
wacom
in
orange,
and
neither
we
can
see
that
in
this
layer
neither
model
has
quite
as
much
co
as
the
observations,
but
nonetheless,
the
enhanced
co
in
the
music
simulation,
just
like
we
saw
on
the
last
slide,
is
actually
indeed
constituting
an
improvement
compared
to
observations
at
this
primary
level
of
convective
outflow
and
then
the
other
layer
I
want
to
look
at
as
well
is
in
the
lower
most
stratosphere.
This
is
one
to
two
kilometers
above
the
local
tropopause.
G
The
music
assimilation
again
is
falling
closer
to
observations
than
the
wacom
simulation,
and
it's
worth
noting
actually
looking
at
the
shapes
of
these
distributions
that
subjectively
at
least
wacom
seems
to
struggle
with
the
shape
of
that
co
distribution
compared
to
observations
and
really
skews
towards
the
very
low
co
values
in
this
layer.
So
I'm
calling
that
layer
into
question
there
in
that
simulation.
G
That
being
said,
the
next
value
or
the
next
evaluation
is,
is
going
to
be
looking
at
some
tracer
relationships,
and
so
what
I'm
showing
here
is
ozone
versus
co
relationships,
wacom
on
the
left,
music
on
the
right
and
the
models
are
shown
as
the
pixel
data,
and
the
observations
are
shown
as
very
small
black
dots.
So
I
don't
cover
up
the
model
too
much.
G
One
of
the
key
things
to
note
on
these
plots
is
the
difference
between
the
models
in
their
stratosphere
to
troposphere
transition,
which
I'm
showing
here
in
red,
and
this
actually
appears
to
be
better
represented
by
the
wacom
simulation,
and
this
actually
may
be
due
to
its
enhanced
vertical
resolution
in
the
utls,
which
may
be
better
resolving
that
transition.
So
you
know
we
see
some
benefits,
or
it
seems
that
we
see
some
benefits
to
the
horizontal
regional
refinement.
G
We
seem
to
see
some
benefits
to
the
enhanced
vertical
resolution,
and
so
the
impacts
of
these
different
decisions
certainly
need
to
be
carefully
weighed
and
can
both
have
influence
on
the
resulting
simulation.
G
So
this
slide
is
showing
a
collection
of
relationships
in
the
cfc
chemical
coordinate
space
against
species
that
range
in
lifetime
from
a
decade
to
a
century,
so
fairly
long-lived
species
here
and
all
of
them
are
really
excellent,
represented
by
the
models.
When
you
compare
the
the
black
lines
to
the
red
and
the
orange
lines.
G
So
this
gives
us
confidence
that
the
the
chemistry,
the
chemistry
mechanisms,
as
well
as
the
sources,
the
sinks
etc
of
these
species,
are
being
properly
represented
in
the
models,
and
it
makes
us
confident
that
this
model
or
the
model
configurations,
are
trustworthy
for
characterizing
the
impacts
of
these
species
on
the
summer
monsoon
in
a
you
know
whether
that
be
in
a
past
present
or
future
sort
of
setup.
G
And
then
there
are
another
other
relationships
against
cfc11
with
the
shorter
lift
species
which,
sadly,
don't
show
the
same
kind
of
neat.
Consistency-
and
we
think,
that's
probably
because
these
species
have
lifetimes
on
scales-
you
know-
ranging
from
in
this
case
a
few
months
to
a
few
years
scales
that
may
be
influenced
by
convective
activity
and
that
is
obviously
parameterized
in
the
simulation.
So
we're
not
going
to
see
quite
as
nice
of
a
consistency
as
we
saw
with
the
longer
lift
species.
G
But
the
main
point
that
I'm
going
for
here
is
that
this,
this
chemical
coordinate
diagnostic
that
we've
established
here,
allows
us
to
easily
identify
species
which
are
not
up
to
snuff
compared
to
observations
over
the
monsoon
and
that
showcases
the
strength
of
this
approach
towards
model
evaluation.
So
an
example
of
this
is
that
in
the
bottom
row
here
we
have
some
methyl
halide
species,
methyl
chloride
and
methyl
bromide.
That
seem
to
be
really
missing
the
mark
compared
to
the
observation.
G
So,
with
a
little
bit
more
digging
into
those
species,
we
actually
found
that
both
of
those
methyl
halide
species
are
using
zonally
averaged
emissions
in
the
model,
so
it
makes
sense
that
the
models
would
be
too
low
compared
to
observations
since
the
maximum.
The
maximum
emission
values
over
asia
are
actually
getting
smoothed
out
in
this
zonal
average.
So
to
really
prove
this
and
hammer
this
home,
we
actually
plotted
airborne
observations
that
were
taken
over
the
conus
in
blue,
so
those
ones
are
in
blue.
G
The
asia
observations
from
strata
clem
are
still
in
black,
and
you
know
this
tells
us
that
our
models
over
asia
are
conforming
very
nicely
to
the
background
observed
values
of
these
methyl
halides,
and
this
tells
us
that
the
model
transport
here
is
actually
working
excellently,
but
the
issue
with
these
species
is
with
the
emissions,
so
the
takeaway
here
is
much
bigger
than
just
a
need
to
fix
our
emissions
in
these
species,
though
it's
that
we
can
establish
this
chemical
coordinate
diagnostic,
and
that
has
the
allowed
us
to
identify
certain
problem
species
that
and
actually
isolate
the
root
cause
of
why
they're
not
being
represented
compared
to
observations,
and
so
that
shows
that
this
could
be
a
very
powerful
tool
and
then
finally
I'll
end
with
take
home
messages.
G
G
We've
established
this
diagnostic
framework
and
it's
allowed
us
to
infer
and
hypothesize
some
things
about
sort
of
how
the
models
representing
the
monsoon
region.
You
know
enhanced
lofting
of
co
by
musica,
better
troposphere,
stratosphere
transition
in
the
wacom
stimulation,
as
well
as
representing
the
long-lived
relationships
very
well
and
then
finally,
just
to
say
that
the
the
diagnostics
herein
most
broadly
can
be
applied
to
characterize
the
performance
of
other
model
configurations
as
we
as
we
try
out
new
configurations
or
potentially
at
representing
atmospheric
behavior
in
other
parts
of
the
world.
G
So
I
think
I'm
just
about
over
time
here.
So
I
will
end
there
and
say
thank
you
for
your
attention.
B
Thank
you
very
much
friend.
That
was
right
in
time
and,
first
of
all,
I
want
to
thank
all
the
speakers
for
this
session
very
much.
Thank
you
for
staying
so
much
on
time
that
worked
out
really
really
well
and
now
we're
starting
our
discussion.
If
you
want
to
turn
on
your
cameras,
that's
great,
if
not
that's,
okay,
too,
so
we
do
already
have
two
questions
for
ran
and
since
ran.
You
are
right
here.
B
B
H
Hi,
I
was
just
wondering
if
what
emissions
were
you
using
and
how
can
you
compare
ceo
in
a
way
that
you
take
into
account
the
different
resolution
and
that
maybe
you
could
look
at
anthropogenic
ceo
only
since
you
may
have
different
ceo
with
the
different
vertical
and
horizontal
resolution.
F
G
G
H
B
I
Yeah
thanks,
it
was
a
nice
talk,
I'm
just
wondering
the
contrasting
the
music.
You
know
regional
refined
versus
wacom,
higher
vertical
resolution,
whether
it's
obvious
that
the
meteorology
is
leading
to
you
know
greater
transport
into
the
stratosphere
and
whether
you
know
it's
clear
that
you
know
increase
horizontal
resolution.
It
makes
that
more
likely.
J
G
Yeah,
that's
a
that's
a
really
good
point.
I
think
there's
there's
probably
a
little
bit
more
work
to
be
done
to
really
confirm
that
it
is
in
fact
the
higher
horizontal
resolution.
Like
you
say,
I
believe
there
are
some.
There
are
some
variables
and
I'd
have
to
check
and
see
if
we
output
them
in
those
simulations.
But
there
are
some
variables
that
we
can
output
to
actually
check.
G
I
think
there's
a
some
kind
of
a
proxy
for
convective
detrainment
or
something
like
that
from
the
jitting
mcfarland
convective
scheme,
and
it
would
be
interesting
to
look
at
that
from
the
different
simulations
to
see
if
it's
really
what's
leading
to
that
deeper
vertical
transport
in
the
musical
simulation
yeah
I
mean
that'd.
I
I
F
G
B
Okay,
thanks
rafa,
you
have
your
hand
up,
go
ahead.
K
Yeah,
thank
you,
so
this
is
for
brandon
or
dac.
I
found
really
interesting
this
not
well
performance
on
reproducing
methyl
chloride
and
methyl
bromide
and
this
relation
with
the
lower
boundary
conditions.
Actually,
this
is
something
we've
been
doing
for
even
shorter
shoreline
species
than
those
are
you
planning
on,
including
like
just
increasing
lower
boundary
conditions
or
or
using
like
an
emission
inventory
for
those
species,
and
I
remember
all
for
methyl
bromide.
F
Yeah,
I
can
comment
on
that,
so
these
two
species
have
mole
fraction,
boundary
conditions,
they're,
not
emission
boundary
conditions
and
and
and
we
get
the
mole
fraction
boundary
conditions
from
you
know,
scenarios
like
from
cm6
we
would
need
to.
We
would
need
to
which
I
think
is
what
you're
implying
a
switch
to
emission
boundary
boundary
conditions.
If
they
exist
the
the
cement
six
boundary
conditions
do
tend
to
have
a
latitudinal
gradient,
but
again,
there's
all
mean
this
ren
is
has
as
specified.
F
F
B
Thank
you.
I
would
have
a
question.
You
don't
see
anybody
else
right
now
for
nori,
and
so
now
you
showed
the
changing
by
biogenic
emissions
and
you
you
change
them.
25
50
75,
but
you
focused
on
the
colorado
area.
What
happens
then?
Do
you
look
also
globally?
What
happens
or
can
you?
Actually?
You
cannot
change
them
just
over
one
region
right,
you
have
to
change
them
globally
and
would
that
help
the
performance
or
have
you
looked
at
that.
D
For
right
now,
I've
only
looked
at
the
colorado
region,
but
I
mean
it
would
be.
To
be
honest,
I
haven't
really
taken
too
much
of
of
a
look
over
the
entire
globe.
Yeah.
B
Yeah,
because
I
think
I
I
think
we
do
have
some
overestimation,
I
think
that
some
parts
of
isoprene-
maybe
luisa,
you
know
also,
but
I
I
do
wonder
you
know
how
do
we
tune.
G
D
I
C
C
It's
definitely
a
challenge
and
I
think
yeah
I
think.
Maybe
we
have
seen
isoprene
overestimated
in
other
regions.
I
think
it
can
depend
quite
a
bit
on
the
vegetation
map,
so
that
was
one
idea
we
had
to
improve
the
resolution
of
the
vegetation
map
and
it's
hard
to
evaluate
isoprene,
because
it's
so
short-lived.
C
B
Any
other
questions
I
would
have
a
real,
quick
question
for
doug,
but
I
also
want
to
move
on.
We
can
discuss
I'm
just
really
interested
in
the
variability
of
the
shedding,
and
I
wonder
if
that
has
some
climatic
connections.
You
know
some
what
is
causing
this,
but
I
guess
that's
not
a
short
answer.
F
Yeah
the
dynamics
is
causing
shedding
you're
to
your
inner
inner
annual
variability.
You
know,
I'm
probably
not
the
right
person
to
answer
that
question,
but
you
know
you
have
different
info
signatures
on
it.
I'm
sure-
and
you
know
laura
pan
and
others
have
even
talked
about
the
influence
of
typhoons
coming
in.
You
know
frequency
of
those
in
cyclones,
so
you
know
something
to
figure
out,
because
the
frequency
is
very
different
year
to
year.
B
L
Okay,
thanks
so
we'll
move
on
to
the
the
next
session.
This
will
be
the
same
format.
10-Minute
talks
followed
by
a
question
and
answer
session
at
the
end,
so
first
talk
is
suda
kamali.
Can
you
go
ahead
and
share
your
screen
and
its
development
of
sc
wacom
with
the
non-hydrostatic
and
pass
a
dynamical
core.
M
So,
as
nick
mentioned,
I'm
student
kamali
and
I'm
an
asp
postdoc
here
at
ncar
and
today,
I'd
like
to
share
with
you
some
of
the
work
that
we've
been
doing
in
the
development
of
sc
vaccine
with
the
non-hydrostatic
and
pass
a
dynamical
core,
and
this
is
a
joint
effort
between
mcube
and
hao
from
hao.
I've
been
working
with
hanley
view
and
francis
witt
and
from
mq
with
bill
scammerock,
joe
clem
and
peter
lauritzen.
M
Here's
an
outline
of
the
talk
I'll
briefly
mention
the
motivation
behind
this
work
and
then
we'll
talk
a
little
bit
about
the
dynamical
core
configurations
and
then
I'll
share
with
you.
Some
of
the
climatology
results
from
the
simulations
that
we've
done
in
order
to
explain
some
of
that
results.
We've
done
some
preliminary
gravity
wave,
forcing
analysis
that
I'll
present
as
well
and
as
you
might
already
know,
geospace
applications
require
atmospheric
models
with
top
stops
in
excess
of
500
kilometers.
M
So,
for
example,
vacuum
x
extends
from
the
earth's
surface
up
to
approximately
around
600
kilometers
and
in
these
regions
the
accuracy
of
the
hydrostatic
approximation
becomes
problematic
and
so
there's
an
increasing
need
for
whole
atmosphere.
Models
with
non-hydrostatic
dynamic
chords,
however,
at
the
moment
the
dynamical
cores
available
to
vaccamics
are
all
hydrostatics,
but
recently
as
part
of
the
sema
efforts.
M
Ncar
has
been
adapting
impasse,
a
the
model
for
prediction
across
scale
atmosphere
to
work
inside
of
cesm,
and
this
will
bring
non-hydrostatic
modeling
capabilities
to
csm,
and
it
opens
up
the
opportunity
for
us
to
be
able
to
use
and
pass
a
vivid
vacuum
x
for
geospace
applications.
However,
the
first
step
towards
that
would
be
to
just
get
back
from
running
with
empaths
and
that's
what
we've
been
working
on
and
what
I
will
presenting
on
today.
M
So
the
dynamical
cores
that
you're,
probably
used
to
using
with
vacuum,
are
the
fv
and
the
se
model
and,
as
the
name
suggests,
that
b
model
is
a
finite
volume.
Solver
and
the
se
model
is
a
spectral
element.
Solver
the
fv
model
uses
a
lat
long
global
grid,
while
the
se
model
uses
a
cube
sphere,
mesh
they're,
both
hydrostatic
models,
and
they
both
use
sigma
pressure
vertical
coordinate.
Now
we
also
have
the
option
to
use
empaths
with
vacuum
and
empass
uses
centroidal
virano
image
and,
as
I
mentioned,
is
a
non-hydrostatic
model.
M
It
is
a
finite
volume
solver.
However,
he
uses
a
secret
staggering,
while
the
fb
model
use
the
degree
staggering
and
for
the
vertical
coordinate
we're
using
a
hybrid
train
following
height
vertical
coordinate
in
impasse.
So,
in
addition
to
being
a
non-hydrostatic
model,
the
differences
in
the
horizontal
staggering
of
the
variables
and
also
the
vertical
coordinate
and
also
some
differences
in
the
filtering
configuration
could
affect
the
results
of
the
simulations
and
to
evaluate
and
validate
that.
M
If
vacuum
was
running
properly
with
mpas,
we
decided
to
set
up
a
case
with
sc
vacuum,
for
each
of
the
different
dynamical
cores
for
a
one-year
simulation
on
a
one
degree
mesh
and
then
compared
the
climatology
results
with
each
other
and
I'd
like
to
mention
for
these
simulations.
The
parameters
gravity
wave
forcing
produced
by
frontogenesis
was
disabled
and
we
looked
at
mean
zonal,
wind
and
temperature
climatology,
and
here
I
will
only
be
presenting
the
mean
zonal
wind
climatology
for
the
benefit
of
time
and
before
we
look
at
the
simulation
results.
M
However,
it's
difficult
to
reproduce
in
parametrized
models
and
I'd
like
to
just
point
that
out
before
we
move
forward
and
looking
at
the
results
from
the
simulation,
so
I
have
here
the
mean
zonal,
wind,
climatology
from
the
three
dynamical
chords
running
sc
vacuum
with
fvse
and
m-pass
a
and
at
the
bottom.
I
have
the
climatology
for
your
reference
that
we
had
on
the
previous
slide,
and
this
is
for
the
month
of
january
and
when
we
look
at
the
winter
hemisphere,
we
don't
really
see
that
reversal
that
we
see
in
the
climatology.
M
The
strength
of
the
reversal
is
much
smaller
than
what
we
see
in
climatology,
but
the
location
of
it
is
roughly
at
a
good
place
and
in
order
to
explain
this
reversal
that
we
were
seeing
in
empaths,
we
decided
to
do
some
gravity
wave,
forcing
analysis,
and
here
is
the
total
parametrized
gravity
wave
forcing
which
we
first
looked
at
and,
as
I
mentioned,
this
is
for
month
of
january.
M
At
the
top,
we
have
the
parameterized
gravity,
reinforcing
for
the
fb,
sc
and
empass
and
at
the
bottom
I've
put
the
corresponding
means
on
all
wins,
for
each
dicor
for
your
reference,
and
we
can
see
that,
where
we're
seeing
this
wind
reversal
in
impasse
and
also
the
slowdown
for
the
sc
and
fv,
we
don't
really
see
much
activity
for
the
parameterized
gravity
weight
forcing,
and
that
was
to
be
expected
because,
as
I
mentioned
earlier,
on,
we've
turned
off
the
frontogenesis
function,
and
so
we
don't
have
the
effects
of
the
parametrized
frontal
gravity
waves,
which
are
the
main
contributors
in
this
region.
M
And
so,
if
it's
not
the
parameterized
gravity
wave
forcing
that
is
giving
this
reversal,
we
thought
well,
maybe
it's
the
result,
gravity
we
forced
to
think
and
so
we've
diagnosed
the
result,
gravity
forcing
using
the
transform
or
layering
and
mean
method
and
at
the
top
you
have
these
result,
gravity
wave
forcing
for
the
fvse
and
empass
dynamical
core
and
at
the
bottom
I
have
the
mean
zone
wind
for
the
corresponding
corresponding
to
each
dynamical
core,
and
this
tem
method
will
allow
us
to
look
at
both
large
scale
and
small
scale
gravity
waves.
M
And
so,
when
we're
looking
here,
we
can
clearly
see
an
eastward
gravity
wave,
forcing
corresponding
to
the
eastward
wind
reversal
that
we
see
in
empaths
and
it's
similar
where
we
see
the
slowdown
in
sdn
impasse.
We
are
seeing
these
gravity
wave
forcing
appear,
and
so
we
can
clearly
see
that
it
is
the
result,
gravity
wave,
forcing
that's
creating
these
reversals
and
if
you
look
more
closely
at
this
region
to
try
and
see,
what's
the
difference
between
the
result,
gravity
based
between
the
three
dynamical
cores.
We
have
here.
M
The
vertical
profile
of
the
result
gravity
wave
forcing
and
we
can
see
that
for
empaths
at
a
lower
altitude.
Where
that
reversal
is
happening,
is
the
forcing
is
much
stronger.
And
although
we
see
a
strong
forcing
for
the
fb
and
sc
that's
happening
at
a
at
a
higher
altitude,
and
so
and
that's
why
we're
seeing
these
reversal
in
empaths-
and
this
is
very
preliminary
studies
and
in
order
to
really
understand
the
difference
between
the
result-
gravity
weight
forcing
between
the
three
dynamical
cores.
M
There
are
certain
differences
between
them
and
in
order
to
study
that
we
have
looked
at
some
gravity
wave,
forcing
we've
done
some
analysis
there
and
we've
realized
that
a
lot
of
the
differences
is
coming
from
the
result
gravity
wave
forcing.
However,
in
order
to
really
understand
these,
differences
between
the
models
will
have
to
do
further
study.
M
Perhaps
on
spectral
analysis,
study
and
I'd
like
to
mention
here
that
this
was
the
first
step,
as
I
said
earlier,
towards
the
main
goal
for
us,
which
is
to
adapt
the
non-hydrostatic
deep
atmosphere
and
pass
a
to
work
with
vacuum
x.
So
thank
you
for
your
time
and
we'll
be
happy
to
answer
your
questions
in
the
q.
A
section
of
this
session.
L
Okay,
thank
you.
Sudha
on,
as
we
did
mentioned,
either
put
your
question
in
the
chat
or
save
them
for
the
question
and
answer
at
the
end
of
the
session.
So
next
talk
tao.
Could
you
go
ahead
and
start
sharing?
Your
screen
will
be
given
by
wang
and
it
is
the
parameterization
scheme
of
thermospheric
gravity
waves
in
wacom
x,
and
we
can
see
your
screen
it's
not
in
the
presentation
mode.
Yet,
okay,
okay,.
N
Okay,
hi
everyone,
I'm
tall
a
phd
student
at
united
university
and
visitor
in
incredible,
I'm
glad
to
to
give
to
give
a
presentation
in
cs
css
workshop
and
today.
My
topic
is
the
summerslam
graduate
parameterizations
in
welcome
max,
and
we
know
the
the
definition
and
the
propagation
of
gratitude
is
important
throughout
the
whole
atmosphere
and
in
the
lower
atmosphere,
the
low
and
mid-atmosphere.
N
The
dissipation
is
mainly
due
to
the
wave
breaking
and
in
the
upper
atmosphere.
When
the
we
can
see
with
the
increasing
kinematic
viscosities.
The
molecular
damping
plays
an
important
role
on
the
dissipation
of
relativist
in
the
thermosphere
for
the
current
vector
max.
This
is
a
f3
worker
max
we
can
see
there
is
a
large
display,
forcings
near
the
mesopots,
due
to
the
real
breaking
effects
and
it's
implemented
by
the
lincoln
legion
center
recent
theories
and
well
there's
a
the
molecular
damping.
N
Facts
are
not
well
considered
in
the
sphere
in
the
current
welcome
acts,
so
this
work
will
aim
to
develop
a
thermospheric
gradually
planetary
parallelizations
in
welcome
x.
N
Our
we
use
the
the
governing
equation
of
boston's
carbon
equations
and
the
viscous
term,
including
kinematic
viscosity
and
similar
diffusivities,
is
on
the
right
hand,
and
the
mu
is
the
economic
kinematic
viscosity,
and
it
is
the
thermal
diffusivities
and
the
notes,
the
curalis
force
and
the
iron
drug
is
not
neglect
unconnected.
N
In
these
studies
and
the
viscous,
we
can
derive
the
viscous
dispersion
relations
from
the
above
gaming
creations
and
using
the
these
five
definitions
and
hypotheses
and
the
notes,
the
the
imaginary
imagination
of
a
vertical
fibonacci
numbers.
I
might
press
the
inverse
decay
rate
and
when
we,
when
we
use
the
first
two
definitions,
we
can
get
the
dispersion
relations
in
the
same
as
the
equation.
19
invaders
and
the
fridge
in
2005.
N
and
our
dispersion
relations
is,
it
can
be
separated
into
the
real
and
the
imaginary
parts
of
and
into
imagine
the
part
and
the
real
part,
and
we
use
the
iteration
method,
iterative
method
to
get
the
solution
of
mr
and
imi.
N
So
here
are
the
solutions
for
the
mi
skill
height
with
a
100
kilometer,
horizontal
wavelength
and
as
a
function
of
intrinsic
fixed
velocity
and
the
height
we
can
see
the
ionized
skill
height
is
rapidly
decay
rapidly.
Decay
with
the
increase
in
kinematic
viscosities
and
the
red
dotted
line
denotes
the
dissipation
height.
N
N
And
here
are
the
solutions
for
mr
to
easily
understand?
We
calculated
the
mr
into
the
vertical
wavelength
and
the
left
figure
shows
the
vertical
wavelength
considering
molecular
damping
and
we
compare
it
with
the
the
wavelength
or
vertical
wavelength
without
molecular
damping
and
we
can
see
the
the
vertical
wavelength
is.
The
is
the
the
the
the
shorter
the
shorter
vertical
wavelength
is
damped
and
the
longer
virtual
wavelength
is
received
with
the
increasing
kinematic
viscosities.
N
And
here
we
are,
we
said
we,
we
developed
a
calling
recognition,
a
column
vertical
model
to
present
the
upward
protein
propagating
breakfast
at
a
60
degree,
nose.
The
zoning
temperature
and
zonal
wind
at
60
degree
knots
in
january
are
set
as
the
background
fields
and
the
gaussian
graduated
source
from
the
ground-based
phase.
Velocity
from
-100
to
100
meter
per
second
are
launched
at
the
drop
sphere,
and
the
moment
flux
due
to
the
molecular
damping
is
defined
here.
N
So
the
here's,
the
vertical
profile
of
the
momentum,
flux
tau
only
due
to
the
molecular
damping
and
at
for
the
specific
speed
and
the
left
is
the
westward
speed
and
right
on
the
right
hand,
is
the
eastward
phase
speed.
We
can
see
the
wave
stop
propagating
when
they
reach
the
critical
layer
and,
in
general,
the
the
faster
with
steel
with
speed
and
has
a
higher
attitude
of
the
inflection
points.
N
And
here
we
are.
We
have
the
the
the
vertical
profile
of
the
zonal
zone,
forcings
for
the
summary
of
the
wave
spectrum
spectrums
and
the
the
the
solid
lines
the
molecular
dampings
on
the
the
dissipation.
Only
due
to
the
damping
and
the
dashed
line
is
the
only
due
to
the
ray
breakings
and
the
black.
N
The
black
line
is
the
phase
velocity
from
the
minus
100
to
minus
to
100,
and
the
right
red
line
is
the
phase
either
phase
phase
velocity
velocity
from
-60
to
60
meter
per
second
and
the
red
line.
Is
the
the
blue
line
minus
red
lines?
N
Now
we
can
see
there's
a
westward
forces
near
the
mass
box,
meaning
due
to
the
lower
of
speed,
grad
waves,
and
we
can
see
the
molecular
the
the
forcings
due
to
molecular
damping
is
larger
than
the
larger
than
the
real
breaking
so,
which
implies
that
the
molecular
damping
may
contribute
to
the
zono
window
reversals
near
the
master
plus
and
in
the
in
in
the
thermosphere
we
can
see.
The
eastward
forcing
is
shown
occurs
due
to
the
molecular
damping
and
the
faster
phase.
N
Speed
has
a
higher
peak
height
in
the
thermosphere.
N
So
so,
when
we
consider
both
the
molecular
damping
and
the
rebraking
effects
here,
the
vertical
profile
of
zono
lunar
forcings
for
the
for
for
both
consider
the
molecular
damping
and
breaking
effects
and
the
reset
of
the
real
force
at
the
troposphere,
and
we
take
the
minimum
value
of
the
tall
by
real
breaking
and
the
damping
in
the
lower
and
the
middle
atmosphere
to
remain
the
the
forcings
unchanged
there.
N
So
we
don't
really
want
to
change
the
the
the
forcings
in
the
lower
and
the
mid
at
sphere
and
and
when
the
and
for
the
thermosphere,
we
only
consider
the
molecular
molecular
damping
in
the
thermosphere,
because
the
length
and
saturation
there
is
no
longer
active
in
the
thermosphere
and
when
we,
when
we
compare
it
originally,
when
we
compare
it
with
the
sc,
we
can
max
results.
We
can
see
the
the
peak
height
the
peak
height
is
around
to
250
kilometers
in
sc
internet.
N
In
estonia,
I
see
ykmx,
which
is
gradually
resolved
in
in
the
seo
max
the
the
the
attitude
is
higher
than
the
parameterization
results
so,
which
implies
that
the
the
thermospheric
reactive
forces
is
not
only
originated
from
the
troposphere,
but
also
they
can
originate
from
the
mid
and
the
higher
altitude
higher
attitude,
which
is
closer
to
the
second
wave
series
series.
N
So
we
are
now
we
are
working
on
a
set
wave
source
with
a
wider
phase
velocity
in
the
mid
and
upper
atmosphere
to
get
a
higher
peak
attitude
and
closer
to
the
secondary
wave
theory
here
and
some
summaries.
N
The
the
pro
the
properties
of
a
relative
in
the
thermosphere,
including
a
vertical
illness
in
inverse
decrease
and
the
dissipation
height,
is
derived
from
a
vertical
viscous.
Relative
dispersion
relation
and
the
large
reinforcing
in
the
thermosphere
occurs
in
velcro
max
implemented
by
a
thermospheric,
relatively
parameterization
and
the
forcing
in
the
lower
mid
and
the
lower
mid
and
atmospheres
remain
unchanged.
N
We
are
working
on
a
new
scheme
which
is
closer
to
the
secondary
theory
to
more
comfortable
with
sme
recommends
and
the
full
worker
max
model
results,
including
the
thermospheric
gratitude
forcing
and
additive
adiabatic
effective,
edit
vertical
diffusion
will
be
performed
in
the
of
the
future
to
reduce
the
the
amount
of
model
bias.
We
know
the
further
for
the
current
fy
vector
max
the
the
the
column
in
integrated
auto
entry
is
more
higher
than
the
the
observations.
L
All
right,
thank
you
tao.
So
if
you
could
stop
sharing
your
screen
and
raphael,
do
you
want
to
go
ahead
and
share
yours?
So
next
talk
will
be
given
by
rafael
pedro
fernandez,
a
special
temporal
distribution
of
inorganic
halogen
sources
and
sinks
in
the
troposphere
evaluation
of
vsl
chemistry
between
csm1
and
csm2.
L
K
Okay,
thank
you
very
much.
So
it's
a
pleasure
for
me
to
present
these
results
in
the
assessment
workshop
of
the
inorganic
halogen
chemistry,
implementation
in
campchem
and
the
comparison
between
our
results
in
session
one
and
the
porting
into
season
two.
So
this
work
has
been
developed
in
cooperation
with
many
groups
and
people
from
the
department
of
atmospheric
chemistry
and
climate
in
madrid,
spain,
some
students
working
with
me
here
in
argentina
and
mendoza,
and
also
our
colleagues
from
from
anchor.
K
So
I
will
split
my
stock
in
three
different
subsections,
so,
first,
like
a
technical
description
of
which
are
the
main
sources
and
things
in
of
halogens
in
the
troposphere.
K
Basically,
because
we
don't
only
want
to
compare
session
one
and
session
two
in
atmospheric
burdens,
but
also
on
the
sources
and
sinks,
a
comparison
between
natural,
virtually
hello
carbons
with
respect
to
long
live
or
some
depression,
the
pressure
substances
will
be
given
and
also
the
impact
of
sea
salt
aerosol
dehydration.
K
So
then,
I
will
present
some
rapid
results
of
the
most
important
features
on
this
evaluation
of
session,
one
versus
sm2
and
the
role
of
the
boundary
layer
in
limiting
the
transport
of
of
halogens
from
the
boundary
layer
to
the
free-to-sphere
and
at
the
end,
some
nice
results.
K
We
have
just
published
on
the
impact
of
chlorine
on
missing
lifetime
and
burden,
so
basically
to
talk
about
the
contributions
source
gases
are
our
halogen
surge
gases
are
transported
from
the
surface
to
the
stratosphere,
and
I
will
split
the
contribution
from
long-leaf
halogens,
those
that
have
a
lifetime
more
than
six
months,
and
I
will
have
a
special
attention
on
methylchloride
and
methyl
bromide.
That
ren
also
provides
some
results
related
to
the
eclipse
campaign
for
sure
the
contribution
of
charlotte
halogen
species
which
are
photo-decomposed
in
the
troposphere.
K
So
the
long-lived
species
are
assumed
to
be
completely
unreactive
until
the
stratosphere.
But
then
the
halogen,
the
short-lived
halogens
react
in
the
troposphere
and
they
provide
chlorine
and
bromine
atoms
which
participate
in
in
the
and
alters
the
oxidative
capacity
of
the
troposphere
and,
finally,
the
contribution
from
sea
salt.
The
this
sea
salt,
the
halogenation
source,
which
is
further
explained
in
the
next
slide.
K
K
So
this
was
implemented
initially
for
only
for
halogen
reservoir
species,
which
could
capture
and
uptake
the
bromide
content
or
chloride,
content
from
sea
salt
and
in
the
literal
paper
I
represent
today,
we
updated
this
chemist,
this
mechanism
for
chlorine,
in
particular,
because
acidification
of
the
sea,
salt,
encases,
the
the
capture
of
chloride
and
and
then
the
release
of
more
chlorine
to
the
atmosphere.
K
So
this
process
is
strongly
dependent
on
surface
area
density
of
the
model
and
and
then
depending
on
the
distribution
this
this
source,
the
magnitude
of
this
source,
changed
significantly
between
different
model
configurations.
K
So
our
initial
results
here
I
highlight
this
is
the
chlorine
production.
So
this
is
the
release
of
chlorine
atoms
at
different
vertical
levels.
In
the
model
we
can
see
that
the
dominant
source
in
green
are
the
sources
arising
from
sea
salt,
the
halogenation,
which
is
maximized
in
the
boundary
layer.
But
then
it
also
has
a
great
magnitude
in
the
free
throw
sphere.
Then
in
second
place.
So
the
contribution
for
for
from
short-lived
species
is
in
blue
and
we
can
distinguish
the
contribution
from
anthropogenic
species
and
natural
species
released
from
the
ocean.
K
So
basically,
their
anthropogenic
contribution
is
the
only
significant
one.
Mostly
the
chloromethane
is
transported
and
is
releasing
chlorine
atoms
in
the
free
troposphere,
but
very
interesting.
We
can
see
that
the
contribution
for
methyl
chloride,
which
is
here
in
orange
in
the
troposphere,
is
even
larger
than
the
contribution
from
the
anthropogenic
arbitrary
and,
let's
keep
in
mind
that
methyl
chloride
has
mostly
natural
sources
and
it's
not
controlled
by
the
montreal
protocol
and
is
included
in
the
model
as
a
lower
boundary
conditions
and
without
any
special
resolution.
K
So
that
would
be
really
interesting
to
evaluate
in
the
future,
as
highlighted
before.
So.
The
good
things
is
that,
with
what
we
have
been
able
is
to
quantify,
which
is
the
most
important
protein
source
in
the
troposphere,
and
basically
it's
mostly
sea
salt,
halogen
sea,
the
halogenation
both
for
chlorine
and
for
bromine,
comparing
sesame
one
and
session
two.
So
this
is
the
the
same
plot
on
the
left
for
session
one,
and
these
are
the
new
results
for
season
two.
K
What
we
see
is
that
there
is
a
like
a
stronger
barrier
here
in
the
in
the
top
of
the
boundary
layer,
so
sea
salt
is
not
transported
so
efficiently
to
the
free
troposphere
and
then
the
the
magnitude
of
the
chlorine
released
from
from
this
recycling
is
largely
reduced
by
approximately
14,
which
I
we
believe
is
more
realistic
representation
of
the
atmosphere
so
in
comparison
for
bromine,
which
is
what
we
are
more
familiar
with.
K
We
can
see
that
the
this
barrier
for
bromine
recycling
on
sea
salt
is
also
present
in
session
two
and-
and
this
is
a
good
thing-
a
good
improvement
in
the
bottle,
but
then
also
that
the
contribution
from
natural
hollow
carbons,
virtually
hello
carbons
in
comparison
with
the
lower
with
the
long-lived
methyl
bromide,
is
also
much
important
and
and
this
the
inclusion
of
virtually
species
for
bromine
is
really
required
for
the
proper
representation
of
bromine
in
the
troposphere.
K
So
this
is
just
a
comparison
of
which
are
the
places
I
don't
know.
I
will
not
go
through
the
all
of
these
numbers,
but
where
are
the
the
first
second
contributions
of
halogens
to
the
troposphere?
So
for
both
cases
for
chlorine
and
for
bromine
natural
sources
are
the
dominant
ones,
although
for
chlorine
is
mostly
basil,
chloride
and
for
bromine
is
mostly
bromophore.
K
So
this
is
the
final
comparison
that
we
have
for
session
two.
We
have
now
a
freeze
version
of
a
halogen
chemistry
in
session.
Two
we
can
see
the
plots
on
the
left
are
results
of
inorganic
chlorine
in
session
one
and
on
the
right
in
session
two.
We
can
see
this
small
reduction
here
in
the
in
the
free
troposphere
due
to
the
sea
salt
recycling
that
was
too
efficient
in
season,
one
that
has
been
updated
but
consistently.
K
We
can
see
that
the
source,
gas
injection
and
the
product
does
injection
of
of
organic
source
gases
and
inorganic
product
gases
is
very
similar
between
the
two
models
which
are
here
on
thick
and
dashed
lines.
So.
Lastly,
I
just
wanted
to
introduce
some
of
these
really
interesting
results
of
the
impact
of
clotting
on
methane,
which
connects
halogens
with
climate.
K
So,
basically,
all
these
improvements
that
I
have
just
mentioned
have
been
implemented
in
this
model.
All
these
results
are
have
been
performed
with
system,
one
version
of
campaign
with
halogens.
We,
we
improved
the
representation
of
of
hcl
and
many
surfaces
around
the
world
surface
observations
around
the
world
which
are
represented
here
on
the
right
and
also
we
have
a
really
nice
representation
of
the
basic
vertical
profile
of
the
chloromethane,
which
was
the
main
chord
in
the
source
from
charlie
halogens.
K
So,
what's
the
role
of
chlorine
in
methane,
so
basically
general
wisdom
says
that
if
you
increase
an
oxidant
which
is
the
chlorine
atom
by
including
all
this
halogen
chemistry,
then
we
will
increase
the
loss
of
methane
by
the
direct
reaction
with
chlorine
with
methane.
But
we
have
to
keep
in
mind
that
the
main
the
main
destruction
channel
channel
for
methane
is
o
h
and
o
h
is
strongly
affected
by
halogens
indirectly,
because
that's
one
of
the
main
implications
that
have
halogen
chemistry
in
the
troposphere
they
affect
ozone
and
affecting
altering
ozone.
K
Then
they
can
change
the
o-h
partitioning
and
abundance
in
the
troposphere.
In
the
clean
environments,
the
inclusion
of
halogens
decreased
the
total
overage
abundance,
but
then
in
polluted
environments,
if
they
can
increase
this
so
which
of
these
contributions
wins
when
we
include
halogen
chemistry
on
methane,
and
we
can.
We
have
computers
on
on
the
on
the
loss
rate
like
quantitatively.
K
So,
basically,
here
on
the
right
on
the
left,
sorry,
we
can
see
that
enhancing
the
the
chlorine
loss
rate
increases
by
100
or
200
percent
when
we
include
short-lived
species
in
the
model.
But
then
this
also
results
in
a
reduction
of
eight
percent
of
the
total
oah
losses,
and
this
compensates
and
the
reduction
in
in
a
wh
the
destruction
of
missing
by
the
reaction.
K
You
know
age
drives
the
overall
behavior
of
methane
in
the
troposphere,
so
we
have
a
lifetime
of
enhancement
of
approximately
0.6
jrs,
so
that
is
almost
sorry
six
percent
of
the
total
lifetime,
and
we
can
also
see
that
the
burden
is
increased
regardless
of
the
simulation.
So
this
is
occurring
for
rcp,
6.0
and
rcp
8.5
and
that
the
net
reduction,
if
relative
forcing
sorry
the
increase
in
relative
forcing
is
approximately
0.05
watts
per
square
meter.
So
that's
all
for
today
and
a
big
pass
of
time.
So
thank
you
very
much.
L
All
right,
thank
you.
If
you
could,
please
stop
sharing
your
screen
and
we
can
move
on
to
the
next
talk
which
will
be
given
by
sebastian.
Easton.
Do
you
want
to
go
ahead
and
start
sharing
your
screen?
L
Sure
and
the
presentation
is
geoscam
as
a
chemistry
module
in
csm,
so
we
can
see
your
slides,
so
please
go
ahead.
J
Excellent,
thank
you.
First,
I
apologize
if
this
talks
a
little
low
energy
or
strange.
I've
been
recovering
from
stomach
bugs
the
past
three
days,
so
a
little
out
of
it,
but
thank
you
so
much
for
this
opportunity
to
talk.
I'm
going
to
be
talking
today
as
introduced
about
geoscam,
not
as
a
standalone
model,
but
instead
now
as
a
potential
chemistry
module
available
in
cesm
2.1.1,
and
this
is
really
work.
J
That's
been
completed,
not
so
much
by
me
so
as
by
my
student,
thibaut
fritz,
who
unfortunately
couldn't
be
here
today
with
incredible
help
from
collaborators
both
at
encar,
such
as
ladies
and
gentlemen,
steve
goldhaber
and
at
harvard
such
as
hype,
england,
elizabeth
lundgren
and
danielle
jacob
before
we
get
started.
I
just
wanted
to
acknowledge
that
this
work
is
was
supported
by
the
nsf
and
that
we
made
use
of
the
generous
resources
at
cheyenne
to
complete
it.
J
So
the
basic
question
is:
why
do
this?
We
have
a
perfectly
good
chemistry
module
in
csm,
of
course,
already
in
camcam,
never
mind
welcome,
and
the
basic
idea
is
that
we
want
to
advance
a
few
different
ideals.
So
we
are
looking
to
address
the
so-called
chemistry
brand
challenge
brought
up
by
the
national
research
council.
This
need
for
chemistry
to
be
an
online
process
in
ccms
and
which
was
not
not
a
capability
of
the
geoscan
standalone
model.
Prior
to
this
point,
really
we
want
to
increase
accessibility.
J
That
means
making
gscam
available
more
available
to
the
cesm
community,
while
also
bringing
cesm's
capabilities
through
to
the
gs
cam
community.
We
think
that
by
knitting
those
communities
together
through
a
common
access
point,
we
will
allow
a
greater
transfer
of
knowledge
and
a
greater
understanding
of
the
relative
capabilities
and
deficiencies
of
the
models
in
question.
J
So
I
wanted
to
come
back
to
that
question
of
consistency.
Geoscan's
ozone
is
thought
to
be
biased
low,
whereas
relative
to
this
we
think
the
cam
kemps
is
not.
This
was
something
that
was
evident
in
parkinson's
2021
paper,
looking
at
comparisons
to
the
chorus
aq
measurements,
but
at
the
same
time,
in
some
other
areas,
gs10
might
get
a
better
fit
than
canton,
and
historically
certainly,
the
community
that
I'm
most
familiar
with
gscan
community
would
think
of
this
as
being
a
chemistry
problem.
But
of
course
there
are
multiple
possible
reasons
for
this.
J
Could
it
be
because
of
different
emissions,
inventories,
different
chemistry,
as
we
thought,
or
perhaps
even
different
meteorology,
and
the
truth
is
that
when
we
don't
have
a
common
framework
in
which
to
test
multiple
chemical
modules,
it's
difficult
to
say
which
of
these
factors
is
the
case
or,
if
there's
a
common
deficiency
between
them,
which
causes
different
outcomes.
J
So
to
that
end,
we
have
gscan
now
integrated
as
a
chemistry
option
in
cesm
as
it
stands.
This
is
a
downloadable
set
of
code
that
I'm
happy
to
share
with
anyone
if
they're
interested,
where
we
have
geoscam
as
a
parallel
option
alongside
camcap.
This
is
the
basic
geoschem
code.
It
doesn't
actually
involve
any
modifications
from
the
same
gs10
code
that
powers,
the
gscan
classic
ctm
or
the
gchp
ctm,
or
indeed
the
jscom
code,
which
is
within
nasa's
geos
model.
The
only
differences
are
that
there's
a
new
interface
which
allows
it
to
interface
with
cesm.
J
This
also
allows
it
to
present
the
information
that
is
needed
to
the
modal
aerosol
model
for
man4
in
order
to
generate
modal
aerosols,
which
cesm
can
interpret.
This
is
fully
coupled.
We
can
run
online,
although
the
simulations
I'll
show
you
in
a
second
are
not,
and
this
uses
anthropogenic
emissions
through
the
hemco
module.
J
Hemco
is
essentially
a
grid
independent
emissions
handler
which,
through
this
grant,
is
now
available
for
both
geoscanning
csm
and
camcam
and
cdsm.
J
I
don't,
I
suspect,
I
don't
really
need
to
introduce
music
into
this
community,
but
the
general
idea
being
that
we
create
a
modular
approach
to
chemistry
for
integration
into
any
atmospheric
model.
Through
this
project,
we
are
eventually
hoping
to
have
not
just
geoscan
monolithically
available
in
csm.
J
So
as
a
basic
evaluation,
we
wanted
to
ask
how
these
models
compare
when
simulating,
say,
2016
and
comparing
to
observations
so
in
the
slides
that
are
coming
up.
You'll
see
results
from
geoscan
chemistry
embedded
in
cesm.
We've
got
that
cgc
cam,
chem
chemistry
embedded
in
cesm
c,
slash
c
cc
and
then
geoscan
sound
alone
in
its
own
ctm
s,
just
s
g
c.
In
this
case
they
are
all
using
mostly
harmonized
emissions,
something
I
can't
see
later
and
mirror
to
meteorology
in
the
case
of
standalone
geoscans.
This
is
running
as
a
ctm.
J
In
the
case
of
both
the
cesm
simulations.
This
is
nudged
meteorology.
So,
if
we
look
at
say,
zonal
mean
ozone.
Here
this
is
the
simulated
zone
amino
zone
for
2016
from
cesm
gsk.
It
looks
plausible,
but
if
we
compare
it
to
what
we
get
in
say,
cesm
cam
cam.
We
know
that
we
have
this
low
bias
that
we
might
have
expected
in
this
part.
J
Blue
means
that
csf
geoschem
is
estimating
less
ozone
than
cesm
cam
can,
and
this
is
most
likely
related
to
the
bromine
chemistry
in
geos
kevin,
in
particular
the
additional
bromine
emissions
through
sea
salt,
which
is
something
that's
now
being
addressed
in
more
recent
versions
of
geoscan
and
csmgc.
J
However,
what
was
most
interesting
to
us
was
then
the
comparison
to
standalone
gs10,
where
we
find
that,
as
we
expected,
that
underestimate
in
the
northern
hemisphere,
was
largely
gone,
but
we
still
get
a
strong
underestimate
in
the
southern
hemisphere.
Now
this
is
most
likely
due
to
greater
sea
salt
emissions
when
we
calculate
them
through
cesm
than
or
rather
through
cam
than
when
we
were
calculating
the
standalone
geoscale.
J
So
in
this
case
we
know
that
the
nudged,
geoscam
and
cesm
and
enlarged
cancun
and
cesm
are
using
identical
surface
water,
flux,
temperature,
winds
and
surface
pressure
for
merit2,
but
then
using
cam's
physics
to
diagnose
other
meteorological
variables
and
that's
resulting
in
differences
in
terms
of
things
such
as
the
water
field,
between
geoscan,
standalone
and
geostem.
When
it's
running
in
cesm,
and
we
and
you
can
see
that
in
the
central
plot.
Here
we
have
more
water
vapor
in
the
upper
troposphere
when
running
geosphere
in
cesm
than
we
do
in
standalone.
J
Where
it's
being
read
in
from
merit2.
And
then
less
in
the
lower
troposphere,
whereas
when
we
compare
the
cesm
geoscan
to
cesm
camcam,
we
don't
really
have
such
differences,
and
this
also
manifests
in
differences
in
cloud
units.
Say:
cesm
geostem
has
almost
twice
as
much
cloud
fraction,
but
roughly
two
thirds
as
much
cloud
water
and
cloud
ice,
and
this
is
going
to
result
in
differences
in
the
simulation
that
are
more
reflective
of
differences
in
interpretation.
J
J
I'll
finish
it
up
here,
I
try
to
be
short
with
this
talk
just
by
saying
that
we
are
starting
to
understand
as
a
result
of
this
work,
some
of
the
differences
that
we
see
between
cam,
cam
and
gstown,
and
hopefully
we
can
use
this
to
further
advance
those
models.
J
The
paper
that
describes
this
this
advance
is
now
in
gmdd
and
you
can
get
to
it
through
that
qr
code
in
the
top
right
and
talks
in
more
detail
about
how
we
dealt
with
aerosols,
with
halogens,
with
washout
and
so
on,
and
we're
very
pleased
to
say
that,
as
a
result
of
this
work
fully
coupled
runs
and
as
as
are
indeed
nudged
runs,
are
possible
with
geoscan
through
cdsm
and
we're
looking
forward
to
continuing
to
work
on
this,
to
implement
it
as
a
standard
option
in
cesm
rather
than
t1,
and
to
continue
the
modularization
of
gs10
within
cesm
as
part
of
the
musical
project.
L
All
right,
thank
you,
sebastian
and
the
rest
of
the
speakers
in
this
session.
There
were
several
questions
in
the
chat
I
for
suda.
I
think
they
have
all
been
addressed.
If
I'm
wrong,
please
speak
up
and
I
think
the
rest
of
the
questions
for
raphael
I
believe,
have
also
been
addressed.
Are
there
any
other
questions
that
anyone
has
for
any
of
the
the
speakers.
F
J
For
the
agreed,
I
think
the
seesaw
compass
is
one
that
we've
become
increasingly
interested
in,
especially
because
in
the
latest
version
of
geoscan
version,
14
we're
actually
seeing
that
we
have
to
turn
off
say
deep,
bromination
of
sea
salt,
because
it's
becoming
such
an
important
component
of
how
we
treat
ozone
in
moles.
So
getting
a
better
understanding
between
the
different
models
of
how
seasonal
is
treated
is,
I
think,
actually
a
very
high
importance
to
us.
K
E
I
I
had
a
question
for
from
suday
and
bill.
I
Your
results
seem
to
suggest
that
gravity,
waves
or
some
other
sort
of
momentum,
flux
carrying
things-
are
quite
a
bit
bigger
in
the
stratosphere
of
empaths
than
they
are
in
the
other
other
two
dicors,
and
I'm
just
wondering
if
you
guys
have
looked
at
energy
spectra
of
the
divergent
modes
in
these
three
dicors
and
whether
they're
different
all
the
way
you
know
all
the
way
through
or
do
they
start
off
about
the
same
and
decrease
less
in
in
impasse,
I
mean
I'm
just
wondering
if
it's
a
difference
in
the
you
know,
damping
of
the
of
the
dicor
or
whether
it's
a
difference
in
the
amplitude
of
the
source
at
fronts
or
other.
M
I
Sure
yeah
we're
definitely
running
less
filtering
in
terms
of
the
horizontal
background
filters
and
and
the
the
upper
level
damping
that,
for
example,
the
sc
core
is
is
using,
and
I
think
that
may
be
the
the
primary
difference
and
plus
we're
on
a
c
grid.
So
we
don't
have
to
damp
as
much
because
we
don't
have
a
null
space
in
our
pressure
gradient
divergence
operator.
B
Yeah,
I
was
just
on
the
on
the
sea
salt
again,
but
I
think
we
should
follow
up.
I
just
think
that
the
connection
between
sea
salt
and
then
obviously
the
halogens,
is
very
tight,
and
that
is
you
know
when
the
sea
salt
is
changing
with
model
versions.
It's
a
problem
and
I
will
talk
about
it
in
our
timeline.
B
We
do
implement
the
new
mem5
version
or
we
want
to
at
least
that
we
started
testing,
that
that
would
change
again
emissions
and
burden
of
sea
salt
and
dust
because
it
reverts
the
moats,
the
moat
width
from
the
coarse
mode
of
sea,
salt
and
dust
back
to
what
we
had
in
csm1,
and
so
it's
likely
that
that
will
change
something
and
other
things
are
that
we
are
testing
some
reordering,
basically
of
when
emissions
are
taking
place
in
the
model
and
the
diffusion
that
got
changed
with
the
implementation
of
club,
and
so
those
things
may
change
again.
B
K
Just
a
comment
on
this
also
simone
is:
we
haven't
really
finished
our
comparison
for
different
vertical
resolutions.
K
That's
that's
the
first
thing
without
changing
into
a
new
version
or
a
new
diffusion
scheme,
but
at
least
using
the
same
one
trying
to
understand
the
response
of
the
model
due
to
different
vertical
resolutions
would
be
important
also
for
us,
and
we
are
actually
working
on
that.
L
If
not,
I
I
do
have
a
question
for
for
tao.
You
you
mentioned
adding.
You
know
some
sort
of
additional
wave
source
in
the
in
the
mesosphere,
I
think
to
get
secondary
waves.
Do
you
have
an
idea
of
how
how
that
would
work
in
in
sort
of
practice.
N
Yeah
there
are
the
actually,
there
are
two
ways:
the
ones
that
kindly
discussed
with
me
the
once
the
wheel
is
breaking
and
we
add
the
we
added
the
molecular
damping
terms.
No
all
right.
The
way
we
reap
is
the
perfect
is
upward,
propagating
and
once
it's
breaking,
and
we
only
consider
the
molecular
damping
and
we
set
a
new
wave
spectrums
and
we
consider
the
the
molecular
damping
turn
effect
and
that's
that's
more
closer
to
the
closer
to
the
the
secondary
theory
and
all
all
we
can.
N
We
can
say
that
we
just
said
a
damping
molecular
damping
effects
above
a
center
new
new
new
new
spectrums
above
the
120
kilometers
yeah.
We
decide
to
set
a
new
spectrum
of
border
spectrum
above
the
100
120
kilometers,
to
to
consider
the
molecular
damping
effects
and
and
below
below
below
in
the
lower
energy
meter
atmosphere.
We
can
see
the
molecular
damping,
so
that's
the
two
ways
and
I
think
the
more
the
one.
The
first
way
is
the
more
reasonable,
more
physical
yeah
but
yeah
working
on
this
all
right.
Thanks.
L
Since
I
don't
see
any
other
questions
or
comments,
I
think
we
can
conclude
the
this
session
here.
We
will
have
15
minute
break,
and
so
we
will
restart
again
at
10
40
mountain
time.
K
B
Yeah,
so
good
is
nick
still
sharing
we're
sharing
the
slides.
K
Of
science
and
model
evaluation,
so,
as
with
the
first
two
sessions
we
have
before
the
break,
we
will
go
through
this
10
minute
talks
and
then
please
include
all
your
questions
in
the
chat
or
keep
in
mind
for
the
for
the
join
questions
and
answer
discussions
again.
So
the
first
one
will
be
presented
by
the
sound
job.
K
Improvement
of
secondary
organic
aerosol
representations,
income
can
to
match
radiation
fields
in
full
chemistry
version.
O
Can
hear
you
thanks
yeah?
My
name
is
du
song
and
carl
econ.
So
thank
you
for
giving
me
the
opportunity
to
share
some
results
today,
I'm
going
to
talk
about
secondary,
gaining
aerosol
parameterization
in
cam,
and
so
there's
a
new
parameterization
for
consistent
radiation
fields
between
chem
and
chemcam.
O
Organic
aerosols,
which
are
shown
in
green
color,
contribute
to
significant
fraction
of
some
microneedle
cells
globally
and
here's
another
pie
chart
now
showing
further
classifications
of
organic
aerosols
into
primary
and
secondary
organic
aerosols,
secondary
organic
aerosol,
which
is
shown
in
purple.
Color
here
is
much
more
abundant
than
primary
organic
aerosol
in
gray,
color
organic
aerosol
is
also
important
in
terms
of
radiation
balance
and
so
climate
modeling,
because
organic
air
cells
cater
sunlight
efficiently,
as
its
radius
is
small.
O
I'm
not
going
to
give
every
detail
in
the
interest
of
time
here,
but
basically
cam
can
use
10
tracers
with
5
for
guests
and
the
other
5
for
aerosol
phase.
In
addition
to
number
of
tracers
here
are
some
discrepancies
like
online
pressures
of
flying
biogenic
emissions
and
precursor
gases
and
loss
processes.
O
O
O
O
But
briefly,
we
added
a
tracer
called
sua
here
to
efficiently
represent
the
chemistry
of
precursor
gases
in
chemcam,
and
we
try
to
match
model
parameters
between
chemcam
and
cam,
but
the
new
scheme
doesn't
add
number
of
sva
tracers
simulated
in
the
model.
As
a
result,
the
computational
cost
only
as
two
percent
and
this
two
percent
increases
actually
within
internal
model.
Variability.
O
This
slide
shows
the
model
result
or
global
atmospheric
s3
burden
on
the
left
and
vertical
profiles
of
swa
and
pom
primer
are
getting
better
using
one
year.
Not
jit
simulation
camcam
is
shown
in
black
and
can
default
is
shown
in
blue,
and
you
can
see
some
discrepancies,
such
as
seasonal
variations
and
vertical
profiles
in
the
upper
atmosphere
and
slight
differences
in
primary,
getting
better
concentrations.
O
O
Actually,
the
advantages
of
using
the
as
new
sua
scheme
is
not
apparent
here
at
the
surface,
but
for
high
altitudes,
for
example,
same
plot,
but
for
500,
hectopascal
region
default
cam
fails
to
capture
the
high
concentrations
over
south
america
and
africa,
and
this
is
this
flood
is
for
100
hectopascal
again.
O
O
Here
paragraphs
show
annual
mean
global
burden
of
sua
black
carbon
and
pom
cam
cam
shown
in
black
again
came
in
blue
and
usually
scheming,
cam
shown
in
green.
The
new
scheme
shows
better
agreements
compared
to
camp
camp,
especially
improvement
of
black
carbon
is
important.
I
think,
because
black
carbon
is
important
for
absorbing
sunlight
in
the
model.
O
Radiation
balance
is
also
important,
especially
for
climate
research
applications.
Here
I
show
you
net
shoot,
wave
and
long
wave
at
the
table
model
from
cam
cam
and
differences
between
cam
and
cam
cam
for
the
default
and
new
sua
scheme
global
mean
value
of
difference
is
shown
in
parentheses,
because
organic
also
sunlight.
So.
O
This
survey
first
thing
is
changed
and
differences
between
cam
and
chemcam
are
decreased
in
the
news
soa
scheme,
especially
over
the
pearl
region.
Here
this
is
a
slide,
is
for
1850s
and
for
2000
simulations
again.
The
new
scheme
shows
better
agreements,
but
for
long
a
version,
the
global
mean
value
gets
worse,
but
this
is
because
the
default
cam
shows
some
strong,
negative
bias
and
also
strong
positive
bias
over
some
reasons,
so
biases
are
cancelled
in
the
default
cam.
O
This
is
a
summary
slide,
showing
bar
graphs
for
global
average
net
short
wave,
long,
wave
and
short
wave
cloud,
firsting
and
long
wave
cloud
version
again
following
the
improvement
of
aerosols.
Radiation
fields
are
also
improved,
especially
for
shoot
wave
and,
furthermore,
in
terms
of
differences
such
as
residual
energy
balance,
which
is,
should
weigh
minus
long
wave
version
came.
O
We,
the
new
sv
scheme,
shows
a
less
bias
compared
to
chemcam
and
for
radiate
firsting,
so
which
is
a
short
wave
on
the
left
and
long
wave
on
the
right
so
which
is
the
difference
between
2000
and
1850s
again.
Can
we,
the
new
asus
cam,
shows
better
agreement
with
cam
cam
here
for
both
shortwave
and
long
way?
O
So
in
summary,
organic
aerosols
represent
a
significant
fraction
of
aerosols
in
the
atmosphere
and
also
important
for
earth's
radiation
balance
and
can
use
the
simplified
sva
scheme
for
computational
efficiency,
which
results
in
discrepancies
in
aerosols
and
also
radiation
fields
between
chem
and
chemcam,
and
I
think
this
is
a
big
problem
when
it
comes
to
climate
studies,
as
most
climate
modeling
studies
use
cam
instead
of
chemcam
because
of
computational
cost.
K
Thank
you
all
the
sun
and
for
the
nice
presentation
we
can
get
the
questions
later.
Next
talk
if
you
want
to
share
your
screen,
is
from
ben
gobert
and
it's
titled
two
decades
of
carbon
monoxide
from
space
and
surface
in
c2
global
networks
and
evaluation
of
camcam
simulations.
H
H
So
yeah
muppet
has
been
launched
in
1999
and
has
been
providing
consistent,
reliable
results
since
2001
and
it's
consistent
with
also
other
infrared
measurements
from
other
satellites
and
have
been
launched
later.
As
shown
here
and
co,
column
average
show
pretty
marked
seasonal
cycle,
but
one
of
in
the
northern
hemisphere
and
differences
in
cycle
and
thousand
hemisphere
due
to
biomass,
burning
and
long-term
trends
that
show
overall
decrease
over
the
last
20
years,
but
also
before.
H
If
you
look
at
surface
data,
but
that
decline
has
been
slower
in
the
second
half
of
the
2000s.
H
H
So
I
can
plot
and
apply
the
averaging
kernels
to
camcam
for
the
monthly
million
as
well
and
for
camcam,
I'm
using
anthropogenic
camps,
emissions
and
q
fed
for
biomass
burning
and
mirror
to
to
adjust
for
the
methodology
and
I'm
applying
I'm
doing
another
simulation
with
repeated
emissions
to
account
for
to
estimate
the
natural
viability
in
nco,
and
so
as
as
usual.
If
you
compare
the
moped
and
the
model,
you
see
a
large
underestimation
of
co
in
the
northern
hemisphere.
H
H
This
is
due
to
the
course
grid
likely
and
the
lack
of
vertical
resolution,
and
because
that
makes
oh
to
be
too
high,
but
we
weren't
trying
to
go
beyond
this
type
of
typical
comparison
and
in
particular,
over
the
last
20
years,
there
have
been
a
lot
of
change
in
entrepreneurial
emissions
with
decrease
over
the
industrialized
region
and
increase
in
overtropics.
H
So
when
you
ask
the
question
it's
like,
can
we
look
at
the
change
in
trend,
seasonal
and
internal
variation
and
that
what
can
we
learn
from
those
changes
to
evaluate
the
model?
Further,
and
in
particular,
can
we
see
these
in
in
these
changing
emissions
in
the
muppet
record?
So
here
it's
I'm,
showing
the
moppy
total
column
over
the
intel,
gangtube
plane
so
of
another
ninja,
and
so
I'm
applying
the
stl
to
decompose
the
time
series
into
a
trend
component,
a
seasonal
component
and
a
remainder.
H
So
the
trend
you
can
see
is
pretty
much
the
background
around
100
ppb
and
it's
slow.
It's
slow
variation
over
time.
So
the
trend
represents
a
change
in
background
over
time.
H
The
seasonal
cycle,
around
10
of
the
ceo
and
large,
inter
annual
variability
from
what's
remains
so,
for
instance,
the
2015
amino.
H
So
you
can
see
the
change
in
background
on
the
seasonal
component
and
the
iav
for
every
latitudinal
band,
and
so
you
can
see
again
the
decrease
of
the
background
over
time
and
very
marked
seasonal
cycle
due
to
chemistry
and
photochemistry
and
large
internal
availability
of
tropics
due
to
fires
and
again
from
2015-
and
you
know,
has
a
very
large
signal
in
of
internal
viability.
H
H
So
if
we
look
at
the
seasonal
cycle,
I
calculated
the
seasonal
amplitude
for
every
year
and
look
at
that
change
in
seasonal
appetite
over
time.
And
so
here
are
the
results
from
muppet.
You
can
see
that
the
actual
seasonal,
larger
seasonal
amplitude
occurs
in
the
tropics
of
the
southern
hemisphere
and
they
have
been
decreasing
over
time
by
quite
a
lot
and
it's
decreasing
and
also
in
other
regions.
Why
look
at
the
northern
hemisphere
tropics
is
increasing
slowly
or
fairly?
It's
fairly
stable.
H
What,
although
the
in
the
beginning
of
the
record
of
this
underestimating
the
seasonal
amplitude
for
the
the
tropics
southern
hemisphere,
region
and
the
model,
show
a
decrease
in
seasonal
amplitude
for
in
the
tropics,
while
it
doesn't
show
that
in
the
northern
hemisphere,
but
you
can
see
that
this
is
this
change
of
seasonal
amplitude
due
to
emissions,
because
the
the
repeated
emissions
doesn't
show
that
we
can
go
further
and
look
and
focus
on
that
change
in
the
tropical
southern
hemisphere
region
and
looking
at
the
monthly
data
and
seasonal
amplitude
time
series
using
co
tags.
H
You
can
see
that
the
peak
of
co
of
the
maximum
co
is
due
to
biomass
burning
and
that
the
emissions
are
decreasing
over
time,
so
they
are
responsible
for
the
decrease
in
the
the
peak
of
the
signal
cycle
and
therefore
the
seasonal
amplitude.
If
you
repeat
emissions
there
would
have
been
a
slight
increase.
H
We
can
also
look
at
other
variables,
such
as
like
megan,
ceo
from
medicine.
That
is
also
tagged,
production,
chemical
productions,
chemical
loss
and
there
is
a
strong
dependence.
A
strong
correlation
between
ceo
from
biomass,
burning
to
the
ceo
column,
to
look
at
regional
scale,
I'm
taking
the
example
of
northern
india
again,
the
the
train
component
of
muppet
is
very,
doesn't
change
much
of
the
time,
but
there
is
kind
of
a
reversal
in
the
sign
that
is
due
to
emissions
according
to
the
model.
H
If
you
want
to
look
a
bit
more
into
details,
we
I
removed
the
trend
component
and
then
you
can
see
that
the
signal
of
increasing
on
topogenic
emissions
in
the
region,
it's
much
more
clear,
although
it's
everything
is
on
the
same
sketch
here.
So
it's
harder
to
see,
and
we
can
also
look
at
the
co
tags
to
to
quantify
those
those
changes,
and
it's
also
suggesting
that
this
is
due
to
the
change
in
autocratic
emissions
over
time.
K
P
Thank
you,
hello,
everyone.
I
will
talk
about
our
csm
ii
mem
simulation
of
the
hangar
tonga
eruption.
First,
I
want
to
thank
all
my
co-authors,
especially
encar
team
are
so
supportive
and
many
of
the
people
have
incurred
involved
in
this
work.
P
P
P
P
P
P
Saturated
water,
vapor
curve
is
very
sharply
changed
because
of
the
temperature
is
changed
rapidly
at
this
area,
so
at
20
millibar,
which
means
at
20
millibars.
If
we
inject
amount
a
big
amount
of
water
in
in
there
and
about
2000
ppm
water
can
stay
in
the
gas
phase.
However,
if
we
inject
the
water
into
the
30
millibar,
it
can
only
hold
50
50
500
ppm
of
water,
so
this
figure
we
explore
the
so2
burden
as
a
function
of
time
which
indicate
the
so2
lifetime
in
after
the
tonga
eruption.
P
P
This
is
because
the
h2o
provides
significant
amount
of
o-h.
You
can
compare
the
left
panel
in
the
right
panel.
You
can
see
the
left
panel,
the
so2
deplete
the
oh
and
it
take
time
to
for
more,
oh,
to
react
with
so2,
while
in
our
on
the
right
side,
you
can
see
in
in
the
plume
that
we
always
have
abundant,
oh
to
react
with
so2.
P
This
figure
is
the
the
aerosol
optical
ex
stratospheric
aod,
compared
between
the
amps
and
the
two
model
simulations
with
without
the
water
injection
in
the
middle
and
with
water
injection
in
the
middle.
It's
pretty
obvious
that
with
the
water
injection,
we
get
a
higher
optical
signals
and
we
have
the
sulfate
form
right
away,
because
the
shortened
so2
lifetime.
P
So
we
investigate,
if
this,
the
the
the
large,
so
we
find
the
larger
effective.
A
larger
aod
is
because
the
particle
is
larger
in
the
when
we
injected
water
and
the
effective
radius
of
the
particle
increased
by
50
percent.
P
So
we
wonder
if
the
swelling
or
the
coagulation
of
the
particle
are
the
reason
for
this
large,
effective
radius,
and
we
find
that
the
particle
mass
only
increased
about
50
percent
of
this
because
of
the
swelling
of
water.
So,
to
conclude,
this
part
is
that,
because
of
the
short
so2
lifetime,
subsequent
coagulation
creates
larger
sulfate
particles
that
doubles
double
the
stratospheric
aerosol
optical
depth.
P
So
we
don't
see
much
signals
inside
the
polar
vortex
from
the
h2o,
because
inside
the
polar
vortex
it
forms
polar
stratospheric
clouds.
It
creates
lots
of
variability
in
our
ensembles.
P
P
K
Thank
you
very
much
for
the
presentation
we
can
now
get
into
the
questions
and
asks
and
answers
of
the
different
talks.
If
you
could,
please
stop
sharing
your
screen.
Thank
you.
K
K
Do
you
have
an
estimate
of
how
large
this
sink
bio
h
is,
and
if
this,
the
oh
has
a
large
seasonality
and
if
that's
impact
on
the
on
the
model
or
in
the
on
the
observations
promote
bit.
H
H
But
if
you
use
the
transcom
orange
derived
from
messy
chloroform
that
suggests
no
inter
hemispheric
gradients,
so
you
would
have
less
storage
in
the
northern
hemisphere.
H
If
you
use
that,
then
you
would
have
much
more
co
in
the
winter,
and
so
typically,
when
you
have
knocks
emissions
in
one
degree
grids
you
end
up
with
too
much
storage
in
the
winter.
That
is
responsible
for
that
bias.
Think
otherwise
orange
is
pretty
well
represented
in
the
model.
It's
maybe
slightly
overestimated.
H
I
F
F
Do
you
want
to
comment
on
that
is?
Do
you
have
any
idea
why
the
model
doesn't
want
to
spread,
spread
latitudinally
more
than
observation.
P
Yeah,
so
this
is
one
thing
that
we
we
keep
discussion
discussing
so
first,
the
we
we
try
to
just
inject
in
a
larger
area,
so
we
don't
have
the
we
have
wider
in
wider
spreading,
but
the
spreading
still
cannot
spread
as
large
as
observation,
and
one
particularly
problem
is
spreading
across
the
triple
tropical
tropics
is
because,
like
the
modeled
mary
delano
winged
is,
is
really
small,
and
I
so
I
I
think
in
reality,
there's
wing
shares
that
more
mesoscale
maybe
get
pushed
the
plume
faster
meridianally.
P
Also
it
could
be
the
plume
itself
is.
It
has
a
different
temperature
than
the
surroundings,
and
that
has
some
plume
dynamics
that
press
make
the
plume
spreading
faster.
So
we
we
probably
would
like
some
help
from
other
people
if
other
ones
have
ideas.
Yeah
thanks.
Thank
you.
K
Okay,
so
there
is
a
question
from
susan
on
the
chat:
how
are
the
lower
boundary
conditions
for
greenhouse
gases
determining
wacom
and
are
they
compared
with
observations
similar
to
the
co
carbon
monoxide
study?.
H
K
A
All
right
can
y'all
see
that
yep,
okay,
great
I'll,
just
keep
it
in
this
non-presentation
mode,
so
yeah.
Our
our
hope,
with
this
discussion
session,
is
to
rehash
some
of
our
development
timelines
and
the
sort
of
strategies
and
and
goals
we
have
for
the
remaining
life
of
csm2,
the
csm
x
release
and
then
csm3.
A
So
we'll
present
the
just
represent
the
whole
atmosphere,
timeline
first
and
then
I'll
hand
it
over
simone
and
rafa
for
the
chemistry,
climate
timeline,
and
so
I
think,
actually
you've
seen
quite
a
few
talks
today
on
some
of
the
things
on
the
left
hand,
side
of
this
timeline
related
to
them.
So
we've
had
some
discussion
of
updated
chemistry
and
physics
and
some
improved
and
new
dicor
support
so
again
we're
in
mid-2022.
A
So
we
have
here
initialized
prediction:
comp
sets
which
was
sort
of
a
I'll
say
right
now,
pi
in
the
sky,
but
I
think
we're
pushing
to
get
this
done
because
it
has
a
lot
of,
I
think,
ability
to
support
community
projects.
It's
a
good
follow-on
for
our
existing
earth
system
prediction,
real-time
forecasts
and
then
moving
forward
again
things
like
tm
and
transport
diagnostics,
being
an
online
available
output.
So
folks
don't
have
to
save
you
know
high
time
resolution
output
to
look
at
dynamics
and
transport
see.
A
Okay
and
then
updating
the
the
sd
model
or
the
sd
approach.
So
we've
had
a
few
studies
published
recently
in
about
the
last
year
talking
about
some
of
the
current
deficiencies
with
that
and
how
we
can
improve
it.
There
are
a
lot
of
physics
updates
coming
on,
including
things
like
karma
working
towards
scale,
aware
gravity,
wave
schemes
here
down
the
line
which
is
important
not
just
for
regionally
refined
simulations
like
folks
have
been
showing,
but
even
some
of
our
meshes
like
the
cube
sphere.
Mesh
have
quite
a
variation
in
the
grid
size
over.
A
Just
the
you
know
uniform
resolution
mesh,
and
so
we
want
to
make
sure
we're
treating
everything
properly
and
finally,
we've
shown
the
you
know
future
vertical
grid
for
the
wacom
integration
with
the
l93
workhorse
model
and
we're
we're
thinking
of
two
options
for
how
to
handle
that,
whether
we
stack
wacom
on
top
or
whether
we
transition
in
the
stratosphere
to
a
just
generally
higher
resolution,
wacom.
B
B
Yeah
yeah:
well
it
doesn't
matter
okay
I'll
just
say
I
did
it
or
change
something.
So
the
this
one
is
more
separated
between
developments,
so
this
is
basically
the
first
slide
is
more
like
developments
towards
this
winter
and
fall
and
not
summer
2023
and
then
the
next
one
is
more
longer
term
developments
still,
hopefully
going
towards
chem.
Seven
we're
not
talking
really
about
csm2
csm3
at
this
point,
so
that's
what
I
removed
so
please
disregard
that
at
the
title
yeah.
So
for
our
cam.
B
We
we
already
had
a
talk
today
on
the
new
soa
scheme.
So
what
we'll
try
to
do
is
get
into
nukem
some
some
new
features
that
will
in
particular
be
important
for
chem
that
are
relevant,
of
course,
that
are
aerosol
relevant
changes
and
chemistry,
relevant
changes
for
simple
chemistry
and
things
like
dms
emissions.
Already
in
the
development
code,
a
new
dust
team
is
currently
developed,
but
what
I
also
pointed
out
earlier
is
the
mam5
scheme,
which
is
both
important,
not
only
for
cam
cam.
B
I
mean
it's
actually
designed
for
wacom
to
have
a
course
mode,
sulfate
aerosol,
but
it's
also
important
because
we're
changing
the
sigma
ranges
back
to
what
it
was
to
csm1,
as
I
mentioned
before,
which
is
something
we
think
that
needs
to
be
added
quickly.
So
whenever
tuning
is
happening
in
chem,
7
and
webcam
and
other
models,
we
have
that
already
integrated,
as
well
as
the
discussion
of
marine
organic
aerosols
that
that
possibly
change
the
radiative
forcing
but
then
for
chemistry
on
point
2.
B
We
do
have
new
developments
coming
in
people
I
mean
there
is
updates
on
the
chemical
mechanism
and
the
photolysis
lookout
table,
but
eventually-
and
hopefully
by
the
end
of
the
year,
they
do
work
on
the
new
photolysis
gene,
which
we
call
called
tovx
and
that
will
be
eventually
coupled
to
cesm,
hopefully
sooner
than
later,
that
it
can
be
integrated
and
we'll
be
allowing
much
more
great
science
that
already
people
do
in
other
versions
of
the
model.
B
We
heard
about
vsl
chemistry
today
and
we
added
already
upper
boundary
conditions
for
the
l58
model,
and
we
also
heard
about
the
emission
to
wilhemko
today.
So
what
I
want
to
discuss-
and
it
would
be
great
if
we
have
a
little
discussion
on
the
concepts
that
the
community
would
desire
for
the
new
developments
and
in
particular,
when
you
talk
about
the
workhorse
model,
you
know
what
what
do
we
need?
Also
in
terms
of
horizontal
resolution.
B
We
go
up
in
vertical
resolution,
but
in
some
places
you
know
the
two-degree
model
is
still
valid
for
some
research
and
it
said
something
what
we
want
to
move
forward
and
test
the
more
concepts,
the
more
of
course
work
and
support
is
needed.
But
that's
something,
I
think
is
important
in
the
second
slide,
please,
nick
to
the
next
slide.
B
Yeah,
and
so
that's
more,
hopefully
by
next
summer,
so
we
talked
about
the
in
flexible
errors
interface
in
my
overview
talk,
and
while
we
are
developing
the
aerosol
schemes
like
karma
and
others,
we
think
we
will
merge
that
over
into
the
development
version
with
this
new
aerosol
interface
and
that
will
allow
other
capabilities
also
and
hopefully,
either
into
easier
integration
like
the
brown
carbon
that
is
in
a
sandbox
for
a
while,
as
well
as
other
things.
B
B
You
know
we
had
that
great
overview
talk
on
monday
on
also
the
needs
for
methane
emissions
and
also
other
chemistry,
and
finally,
what
we
really
need
is
community
support.
We
have
seen
the
great
work
together
with
the
harvard
group
here
with
the
hamco
emission
data
set
on
kim
7
and
also
gs
kim
and
then
other
things
that
are
worked
on
in
the
community
and
some
are
missing
here.
B
For
example,
yeah
things
like
nox
emissions
or
also
land
and
megan
developments
that
that
should
be
added
here
and
finally
we're
working
on
melodies.
So
that's
all
ongoing
and
more
planned
towards
next
year
summer.
That's
it
and
now
I
think
what
we
want
is,
I
don't
know
if
we
want
to
continue
sharing
or
if
we
just
stop
sharing
and
have
a
discussion
nick
go
ahead.
A
I
guess
what
we
can
say
is:
may
I
mean
maybe
it'd
be
good
to
just
bring
it
back
to
big
picture
whole
atmosphere,
side,
we're
very
concerned
about
vertical
levels
of
the
new
workhorse
configuration,
but
for
a
lot
of
you
using
not
the
workhorse
configuration
of
the
model,
you're,
probably
interested
in
the
improved
physics
coming
online
and
some
of
these
scheme
changes.
But
from
our
perspective
we
really
want
to
know.
A
B
Yeah,
you
can
write
your
thoughts
in
the
chat
again
or
you
can
just
open
it.
Just
raise
your
hand,
any
comments
are
really
helpful,
as,
as
nick
said,
on
webcam
vertical
grid
on
what
do
you
want
to
see
happening?
Also
priorities?
Well,
we
presented
as
priorities.
Do
you
see
anything
that
really
you
think
we
should
push
forward?
That
really
needs
to
go
in
the
model?
B
Yeah?
That's
I
mean
we'll
we'll
try
to
do
our
best,
but
it
really
is
important
to
get
the
in
the
information
from
the
community
on
here
as
well.
B
B
F
So
you
know,
there's
one
one:
maybe
development
process
or
a
project
we
haven't
really
addressed-
is
that
we've
kind
of
had
a
situation
set
up
in
the
past,
where
we
have
heterogeneous
chemistry
in
the
stratosphere.
We
have
heterogeneous
chemistry
in
the
troposphere
and
there's
really
no
reason
why
we
can't
make
that
more
of
a
universal.
I
Just
that
part
of
that
is
a
different
treatment
of
aerosols
and
the
sulfates,
and
the
way
mam
was
designed.
There
was
three
tropospheric
aerosols,
so
sulfates
are
treated
in
volatile,
solid
ammonium,
sulfate
particles
and
then,
when
we
added
the
stratospheric
sulfates,
we
magically
turn
them
into
liquids
above
the
tropopause
and
the
correct
way
to
do
that
really
is
to
keep
track
of
ammonium
in
the
aerosol,
which
I
think
ma'am.
Seven
does
right.
L
B
Does
actually
keep
track
of
and
and
this
song
has
worked
on
that
version
quite
a
bit
and
shahong's
group
has
initially
worked
on
this,
but
so
I
think
that's
really
a
good
good
point.
You
know
part
of
the
interface
of
the
aeros
interface
will
make
the
integration
of
mosaic
helpful.
B
Then
we
do
have
nitrates,
and
I
think
that
could
be
the
next
step
really
to
say:
let's,
let's
also
try
to
get
rid
of
this
disconnect
or
this
difference
between
the
stratosphere
and
the
troposphere,
because
that's
yeah,
if
it's
really
nitrogen
ammonia
dependent,
I
think
we'll
first
have
to
have
mosaic
included
in
the
model.
I
Thank
you
so
much
for
the
session.
I
have
I'm
kind
of
a
little
bit
newer
to
the
modeling,
so
I
just
have
a
background
question.
I
guess,
which
is
I
think
today
you
were
mentioning
that
they
that
the
the
chemistry
models
that
there
are
different
kinds
of
chemistry
models
that
are
run
depending
on
the
situation,
because
some
of
the
chemistry
take
models
take
so
much
computing
power.
I
So
I
just
was
wondering
for
wacom,
x
and
wacom
are
the
chemistry
models
that
are
run
in
the
troposphere
or
do
they
include
lower
boundaries
that
have
this
is
a
follow-up
question,
but
that
have
that
have
geographic
dependence
and
or
regional
dependence
in
terms
of
the
lower
boundary
sources
for
the
greenhouse
gases.
F
In
general,
the
boundary
conditions
at
the
surface
are
taken
from
assessments
like
cmip6
and
they
they
typically
have
a
latitudinal
dependence
and
in
a
seasonal
dependent
dependence,
but
they're,
not
there's
no
longitudinal
component
to
it.
There's
all
means
right.
I
For
the
long-term
green,
the
long-lived
greenhouse
gases
short-lived
things
like
so2,
pollution
as
as
regional
latitudinal,
longitudinal
dependence
yep,
and
what
about
methane
methane
is
that
in
is
that
a
it's?
Is
that
considered
a
shorter
term
or
a
longer
term?
I
It's
probably
it
probably
just
has
a
zonal
mean,
but
I'd
have
to
check
that.
F
K
L
B
Any
other
questions
or
thoughts.
I
do
wonder
since
I
don't
know
we
have
a
bunch
of
people
in
this
working
group,
how
many
people
would
be
actually
thinking
that
a
support
of
a
two
degree
version,
a
cheaper
version
of
the
workhorse
model-
would
be
something
you'd
be
interested
in.
So
I
think
that
currently
the
costs
will
increase
and
we're
still
figuring
out
the
cost
for
the
58
and
the
93
level
model.
But
some
maybe
studies
like
duck
is
using
often
a
two
degree
model
for
for
quicker
studies.
A
I
think
yeah,
I
think
especially
going
to
a
higher
vertical
resolution
in
these.
We
need
to
keep
supporting
two
degree.
Okay,
I
would
say
for
wacom,
2
degree
and
2
degree
m
a
is
something
we
absolutely
can't
leave
behind
and
probably
need
to
develop
closely
parallel,
because
otherwise
just
too
expensive.
I
know
a
lot
of
folks
here
have
to
get
allocations
and
things,
but
even
for
people
with
access
to
lab
accounts
and
stuff,
it's
still
expensive
to
run
one
degree.
F
I
will
also
add
that,
for
we
have
a
two
degree
version
of
wacom
6
finite
volume,
where
it's
110
levels
and
it
it
gives
a
really
nice
qbo,
and
so
you
can
get
a
qbo
2
degrees.
So
we
should
probably
think
about
working
towards
that
with
a
spectral
element,
two
degree
93
level
or
something
like
that
whatever
or
whatever
nick
decides.
The
vertical
gradient
is
going
to
be
very
vertical
width
of
the
model.
P
Yeah
from
the
university
and
also
several
of
people
from
university,
I
know
about
it's
just
very
too
expensive
for
us
to
run
the
one
degree
model
and
using
our
own
resource
or
using
the
in-car
resource
that
we
apply
for
so
we
most
of
us
usually
just
go
with
two
degree
first
and
then
maybe
one
wrong
degree.
When
our
test
is
more
complete
and
just
running
one
degree
is
taking
a
long
time
and
also
too
expensive
for
us
to
afford.
B
Yeah
and
what
I
wonder
when
we
see
this
really
nice
musical
version,
zero
runs
when
we
are
interested
in
the
local
air
quality,
regional
air
quality,
the
that
doesn't
matter
anymore,
then,
because
you
really
spend
most
of
the
time,
probably
on
that
original
refined
grid
and
just
save
a
little
bit.
So
I
think
that's
why
most
of
the
people
that
are
looking
at
music
are
probably
not
don't
need
that,
but
yeah,
it
seems
to
be
for
more,
like
also
for
the
karma
configurations,
more.
The
stratospheric
applications
important
in
the
long
term.
O
B
I
don't
think
it
improves
much
at
this
point
I
did
to
test
simulations
with
the
93
level
and
with
chemistry,
and
it
was
still
pretty
expensive.
I
think
peter
is
on
here
lauritsen.
I
saw
him
somewhere.
B
I
So
I
hope
that
once
nick
has
the
new
grid
ready
that
we
can
start
fine-tuning
these
things,
but
right
now
we're
we're
using
like
two
or
three
x,
shoulder
time
steps
for
the
vertical
remapping
that
you
would
like
to
use.
I
hope
we
have
room
for
improvement
right
there.
I
B
I
Well,
when
you
have
a
model
top
at
80,
kilometers
we're
not
having
any
stability
issues
and
we're
running
with
optimal
time
steps.
That's
what's
happening
in
cam
right
now,
but
we're
getting
similar
throughput
as
fv
with
moderate
or
moderate
core
counts,
and
we
can
scale
it
out
to
get
more
throughput
than
fv.
But
when
the
top
is
above,
80
kilometers,
the
vertically
breaking
gravity.
Waves
really
grow
large
and
it's
been
a
challenge
to
stabilize
and
that's
what
I
want
to
explore
with
with
nick
once
he's
ready.
I
B
B
Yeah,
I
think
online
meetings
are
tiring
and
the
discussions
are
just
not
the
same
as
almost
sitting
all
in
one
place,
hoping
to
be
back
at
one
point
in
person
and
have
better
discussions.
B
I'm
yeah,
I
think
we'll
can
wait
a
little
longer
for
people.
We
can
also
resume
early.
I
don't
know.
A
I
mean
we
don't
even
just
have
to
take
questions
if
people
also
have
major
issues
with
the
model.
That's
also
a
good
time
to
say
something
yeah,
I
will
just
say
to
re
reiterate
like
a
lot
of
this
development
is
adding
computational
expense.
So
I
think
these
lighter
weight
configurations
didn't
get
the
same
resources
and
attention
this
last
cycle,
probably
because
it
wasn't
so
expensive
just
yet,
but
for
sure
going
forward.
We
know
this
is
an
issue
yeah
peter.
I
A
B
I
I
A
Yeah
I'll
I'll
just
say,
I
think,
probably
the
folks
doing
the
asian
center
monsoon
work
are
a
pretty
good
cut
of
vertical
levels
that
that
topic
seems
to
have
bridged
the
two
communities,
but
in
regard
to
anything
else,
I'm
not
sure
it
seems
much
less
integrated.
At
least
that's
my
impression,
but
I
know
trump
come
folks.
Look
at
tagged
ozone,
a
lot
like
the
3s,
so
there's
at
least
some
use
there
of
trying
to
understand.
A
I
will
say:
there's
probably
so
the
session
tomorrow
and
earth
system
prediction
is
probably
like
the
dry
dynamics
version
of
this
community
kind
of
uniting.
So
there's
a
lot
of
folks
in
that
tomorrow
that
are
going
to
be
focused
on
strategic
coupling
and
even
some
some
stuff
higher
up
just
because
the
dynamics
are
a
fast,
fast
mechanism
there
but
yeah.
A
I
guess
one
issue
going
forward
and
this
kind
of
relief
yeah,
I
was
gonna,
say
I
think
this
relates
so
this
economical
configuration
need
is
stuff
like
the
m,
a
scheme,
the
middle
atmosphere,
chemistry
scheme,
so
for
a
lot
of
longer
term,
geoengineering
stuff,
there's
pressure
to
use
ma
because
it's
three
times
cheaper
to
run
that
model
than
it
is
to
run
with
the
comprehensive
chemistry,
but
obviously
that
there's
some
issues
because
we're
neglecting
some
important
tropospheric
processes
and
like
dan,
has
shown
that
strat
aerosols
tend
to
be
mucked
up
by
not
getting
the
right,
so2
oxidation,
and
so
I
think,
there's
probably
I
don't
know
if
there's
any.
A
I
am
not
a
chemistry
person,
but
I
I
would
say,
like
the
ma
scheme,
I
think
going
forward
is
really
important
for
a
lot
of
this
stuff
too.
That
folks
talk
about,
including
even
the
paleo
community,
but
it
would
be
good,
maybe
at
some
point
for
the
that
scheme
to
maybe
be
adjusted
in
a
way
that
the
aerosols
in
the
stratosphere
can
be
a
little
more
accurate.
A
Complexity-
simone-
I
don't
know
if
that's
your
arena
or
not,
but
it's
something
that
dan
and
I
have
been
talking
about,
because
it
would
be
nice
to
find
a
fix.
B
Yeah,
I
think
there
is
one
thing
the
song
started
with
one
fix.
I
think
there
is
a
difference
between
tropospheric
aerosols
just
because
you
are
prescribing
the
chemistry
and
what
we
did.
What
he
just
showed
will
benefit
the
middle
atmosphere
weka,
it's
not
only
chem.
It
actually
will
also
make
wacom
middle
atmosphere
more
similar
to
wacom.
B
So
that's
one
thing.
The
second
thing
is
that
sulfates
are
different
in
the
troposphere
you,
you
said
yeah
in
the
middle,
in
the
due
to
ls
or
so
there's
different
oxidation,
and
that's
also
for
the
troposphere.
You
do
have
prescribed
fields
of
oxidants
and
those
simply
result
real
result
in
different
sulfate,
because
you
don't
have
the
variability.
So
there
are
possibilities-
and
I
think
there
are
people
from
also
the
hemco
group
and
louisa
on.
C
Yeah,
so
there
are
maybe
two
different
things:
middle
atmosphere.
Chemistry
is
interactive
chemistry,
so
it's
not
specified
oxidants
in
the
troposphere,
but
it
only
has
methane
as
a
hydrocarbon
in
the
troposphere,
so
you're
missing
all
the
voc
chemistry.
So
that
may
be.
You
know
that
could
be
an
important
thing
to
to
come
up
with
some
simple
chemistry
to
expand
on
just
methane
oxidation
so
that
you
have
better,
oh
and
thus
better
sulfate
in
the
troposphere.
C
You
know
that
there
could
be
a
lot
of
different
viewpoints
of
what
what
needs
to
be
what's
most
important
to
get
right,
but
if
we
could
get
ozone
and
oh
closer
to
cam
chem
when
we
use
middle
atmosphere,
chemistry
that
might
be
a
good
goal
and
I'm
not
sure
we've
really
compared
that
at
all.
So
that's.
C
Right
we
could
improve
the
soa
with
the
same
scheme,
but
it
would
be
using
the
calculated
oh
and
ozone
oxidates
oxidants
yeah.
We
should
look
at
that.
That
may
be
one
area
of
them.
You
know
some
one
simple
improvement,
but
to
improve
the
hydrocarbon,
co,
nox
chemistry.
We
probably
need
to
add
some
more.
C
More
vocs,
another
thing
I
was
thinking
about
was
so
I
think
that's
that's
that
should
be
in
you
know,
that's
certainly
something
to
think
about
how
to
improve
the
middle
atmosphere,
tropospheric
chemistry,
another
thing
we've
been
talking
about
a
long
time
is
improving
the
or
we
want
to
keep
adding
tropospheric
chemistry,
and
so
would
there
be
some
way
to
go
back
to
having
a
specified
stratosphere
to
reduce
the
costs
that,
if
we're
primarily
interested
in
lower
tropospheric
chemistry
and
want
to
add
a
lot
of
vocs
and
other
other
things.
C
We
could
specify
stratospheric
components
somehow,
and
I
know
ben
is
interested
in
that
you
know
for
data
also
for
data
assimilation,
studies
where
we
don't
need
to
calculate
all
the
stratospheric
chemistry.
So
that's
something
else
we
could
look
at
and,
if
anyone's
interested
out
there
in
helping
with
any
of
these
questions,
let
me
know.
B
Yeah,
send
us
any
emails
or
thoughts
or
interests
in
ongoing
stuff.
This
will
be
posted.
All
the
talks
will
be
posted
and
hopefully
we
have
in-person
meetings
again.
That
makes
it
easier
and
yeah,
I
think,
more
fun
as
well.
Eventually,
so
I
think
we
can,
if
there
are
really
no
any
more
questions,
I
would
say
we
can
stop
recording
now
and
resume
or
end
the
session
and.