►
From YouTube: 2021 CESM Tutorial: Day 5
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
Also,
as
of
today,
as
of
this
morning,
we
still
have
70
000
core
hours
left
great
job,
everyone
on
burning
those
hours.
We
started
with
200
000
and
the
beginning
of
the
week.
That's
what
we
had
the
account
will
be
active
until
august
31st.
That's
your
cheyenne
account,
however.
The
special
reservation
code
that
we
have
been
using
will
expire
today
at
5
pm
mountain
time.
A
So
just
remember
that
if
you
have
an
existing
case
that
you
want
to
re-run
after
5
pm,
you're
going
to
have
to
change
the
job
queue-
and
this
is
the
command
you
can
either
use
the
regular
or
we
can
use
the
economy.
And
we
talked
about
this-
that's
the
very
first
sizzle
lecture
on
the
different
queues
economy.
Queue
will
be
used
in
less
hours.
If
you
use
the
regular,
you
know
that's
kind
of
what,
if
you
think
of
in
terms
of
money,
the
economy,
one
is
the
happy
hour
and
yeah.
A
A
However,
there
are
many
ways
you
can
still
get
engaged,
one
of
which
is
by
joining
the
many
working
groups
that
are
available
within
csm
and
the
way
the
working
groups
are
designed.
We
have
the
co-chairs
which
they
define
and
lead
the
scientific
directions
of
the
working
group.
Then
we
have
the
liaisons,
the
scientific
ones,
which
are
the
first
point
of
contact
for
questions
regarding
science
performed
of
the
working
group
and
questions
related
to
the
model
component
in
general,
and
then
you
have
this.
A
The
software
liaisons,
which
are
the
primary
goal,
is
to
help
with
software
related
issues
for
a
particular
component.
So
we
have
about
12
working
groups
right
now.
You
can
click
on
this
link
and
see
all
of
them
and
get
engaged
in
the
one
that
you're
you
your
most
close
related.
Your
research
is
most
most
closely
related
with.
A
A
A
As
I
mentioned
before,
you
also
have
the
csm
forums.
There
is
a
forum
for
each
component
and
there's
some
forms,
that's
like
cross
components
and
that's
the
best
method
for
getting
help
with
a
problem
that
you
face
from
now
on.
So
if
you,
if
you
run
a
case-
and
you
know
you
run
into
a
problem-
and
you
see
that
it's
like
ocean
model
go
to
the
csm
forum,
go
in
the
ocean
model,
part
and
then
post
a
question,
you
try
to
answer
as
fast
as
possible.
A
There
are
ways
for
you
to
get
an
allocation
on
cheyenne
and
ceso
has
a
web
page.
That
explains
the
ways
for
doing
so.
If
you
are
in
a
u.s
based
university
is
relatively
straightforward,
so
just
click
on
this
link
and
then
and
then
check
out
the
instructions
there
and
again.
If
you
need
help,
please
feel
free
to
reach
out
to
us.
A
For
today's
office
hours
will
be
slightly
different
than
on
the
previous
days.
We
will
split
the
the
groups
by
by
components,
so
we're
going
to
have
an
expert
for
each
component
of
all
these
groups
or
more
than
one
actually,
and
we
have
four
groups.
So
we
have
one
just
for
land
ice,
so
gunther
will
be
moderating
that
one
then
the
second
one.
We
combine
ocean,
cis
and
bgc.
A
The
third
one
is
just
the
land
model
and
then
the
fourth
one
is
atm
and
wacom
chemistry
and
the
way
it's
going
to
work
out
again
we're
going
to
have
a
breakout
room
for
each
of
these
groups
and
there
will
be
all
the
breakout
rooms
in
addition
to
that,
one
where
people
can
go
for
a
one-on-one
interaction
and
we
can
use
you
know
that
the
same
way
we've
been
using
so
far
and
you
can
join
any
of
the
breaker
rooms
at
any
time.
A
And
then
I
want
to
again
thank
everyone
that
was
involved
in
this
tutorial,
starting
from
the
committee
myself,
alice
dubias,
cecilia
renee,
glanter,
vicky
peter
lawrence,
adam
phillips
and
christine
shields
logistics,
a
huge
thanks
to
elizabeth
for
cloth
for
all
everything
that
she
has
done
throughout
this
week
and
before
and
we'll
be
doing
after
education
outreach.
That
was
the
first
session
we
had
a
few
weeks
ago
with
bao
salon,
jerry
saikon
and
elias
mason,
the
vg
for
the
computing
support
web
support
from
ryan
johnson.
A
We
had
over
60
staffs
involved
throughout
the
tutorial
and
you've
met,
most
of
them
from
soft
engineers.
Science
sciences
and
logistics
related
funding
from
nsf
and
then,
finally,
to
all
of
you
for
participating,
and
now
we
really
hope
that
you
had
a
that.
You
enjoyed
this
week
and
you
you
got
what
you
came
here
looking
for
and
we
still
have
one
day
to
go,
so
I
hope
you
enjoy
today
as
well.
We'll
see
you
soon
in
any
other
csm
event
and
again,
don't
hesitate
in
reaching
out
to
myself
or
elizabeth.
A
C
Hi
so
my
name
is
alice.
Duveya
I've
been
at
some
of
the
office
hours,
but
I
think
I
probably
haven't
met
most
of
you
so
far,
I'm
going
to
just
I'm
moderating
this
session,
I'm
going
to
just
do
a
quick
introduction
of
all
the
people
who
will
be
all
the
cesm
experts
who
will
be
providing
feedback
so,
like
I
said,
my
name
is
alice
duvier,
I'm
a
project
scientist.
C
I
work
on
polar
climate,
mostly
sea
ice,
and
I've
been
at
npr
for
about
five
years,
so
I'll
call
on
the
people.
So
next
we'll
stick
with
the
sea
ice
theme,
so
dave
bailey.
Can
you
introduce
yourself.
D
Sure
so
I'm
dave
bailey,
I'm
gonna,
be
leading
the
ocean
ice
breakout
this
afternoon
and
happy
to
chat
with
you
in
the
session
coming
up.
E
Good
morning
or
afternoon
or
evening,
I'm
betty
otto
bliesner
scientist
in
cgd,
and
I
worked
on
modeling
past
climax.
C
Thanks
zhang.
F
G
G
C
E
I'm
here-
but
I
was
just
I'm
just
here
for
watching,
but.
C
All
right,
so,
I
think
we're
going
to
start
the
session
with
some
questions.
Just
to
gauge
your
previous
knowledge,
I
guess
of
the
polar
either
lander
sea
ice
and
paleo
components.
So
elizabeth.
If
you're
able
to
start
those
questions
for
the
poll.
C
E
That's
exactly
true,
it
does
seem
like
awkward,
science
silence
and
and
dead
air
yeah.
I
can
see
that
two
people
have
completed
it
so
far.
C
When
we
get
up
to
about
30
people
completed,
so
that's
three
quarters,
then
I'll
go
ahead
and
start
this,
but
the
point
of
these
questions
is
to
kind
of
start
a
discussion
and
it
doesn't
matter
if
you
know
a
lot
or
a
little
about
them.
I
just
want
to
emphasize
that.
C
All
right,
I
think,
we're
probably
at
that
point.
So
if
you
want
to
end
the
poll
elizabeth
and
show
the
results,
I
think
we
can
just
use
these
as
a
start
off
point
for
this
discussion,
I
want
to
ask
people
we
got.
It
looks
like
we
got
one
question
ahead
of
time
in
that
poll
end
so
I'll.
Ask
that
one
first
after
this
little
discussion.
C
But
if
you
have
additional
questions,
please
put
them
in
the
chat
and
I
will
ask
them
live
so
that
everyone
can
hear
and
then
the
appropriate
experts
will
respond
to
those
all
right.
So
is
everybody
able
to
see
the
response
for
the
the
answers?
Is
that
right,
okay,
dave's
snotting
his
head
great
all
right.
So
the
first
question
was
about
land
ice
and
which
description
best
describes
the
current
state
of
coupled
land
ice
modeling
in
cesm,
and
it
looks
like
about
52
percent
of
people,
said
the
surface
mass
balance.
C
That's
smb
is
computed
in
the
land
model.
There's
an
interactive
greenland
ice
sheet
in
an
interactive
antarctic
ice
sheet
in
development.
That
is
the
correct
answer.
It
looked
like
a
lot
of
people
also
thought
the
dynamic
ice
sheet
model
is
run
offline
with
the
surface
mass
balance
computed
in
the
land
model,
so
bill
or
gunter.
If
you
have
other
comments
about
responses
to
this
question,
I
invite
you
to
to
go
ahead
and
say
whatever
you
want
at
this
point.
G
Yeah
I'll
chime
in
number,
one
would
have
been
the
correct
answer
around
2010
before
we
had
cesm
and
number
two
would
have
been
the
correct
answer
for
cesm1.
G
So
if
you
were
answering
that
question
five
years
ago,
that
would
have
been
correct
and
that
is
still
a
way
that
the
models
often
run,
but
if
you're
interested
in
couple
landice
modeling.
The
current
state
of
the
art
for
csm
is
question
three.
As
for
question
four,
it
would
be
great
to
have
all
land
ice
components
fully
interactive.
Eventually
I
don't
know
that
will
ever
compute
surface
mass
balance
in
the
land
ice
components
themselves
because
it
works
so
well
to
compute
them
in
the
land
model.
C
Great
all
right,
so
then
the
second
question
was
kind
of
a
contrast
between
sea
ice
and
land
ice.
So
what's
the
main
difference
between
sea
ice
and
land
ice-
and
it
looks
like
most
people
said:
land
ice
is
fresh
while
sea
ice
is
saline,
which
is
the
answer
that
we
were
thinking
of
as
well.
So
land
ice
is
like
the
ice
sheets
on
greenland
or
antarctica,
and
it's
formed
from
precipitation
that
compacts,
whereas
sea
ice
is
frozen,
sea
water.
C
So
there's
a
little
bit
of
salt,
it's
not
as
salty
as
the
ocean,
but
there's
salt
in
that
matrix
and
it
makes
them
behave
differently.
Part
two
land
ice
doesn't
move
while
sea
ice
does
move.
In
fact,
both
land
and
sea
ice
move
sea
ice
moves
on
a
much
shorter
time
scale
than
than
land
ice,
because
it's
moving
around
in
the
ocean
and
pushed
around
by
the
winds
and
moving
around
with
the
currents.
C
So
it
moves
pretty
quickly,
but
landis
does
move
it's
just
a
lot
slower
and
so
that's
also
addresses
the
third
part
of
this
question.
Landice
evolves
on
shorter
time
scales.
In
fact,
landice
evolves
on
much
longer
time.
Scales
than
sea
ice,
but
both
evolve
on
longer
time,
skills
than
say
the
atmosphere.
C
All
right.
I'm
gonna
go
to
question
three
now
sea
ice
so
which
is
not
a
parameter
that
can
be
used
to
change
the
sea
ice.
Albedo
dave.
Do
you
wanna
comment
about
this?
One
sure.
D
Sure
so
the
the
current
scheme
that
we
have
in
the
cs
model
is
this
delta
eddington
radiation
scheme,
and
so
there
are
certain
inherent
optical
properties
that
we
are
adjusting,
but
then
those
effectively
end
up
producing
the
amount
of
absorbed
shortwave
in
the
snow
in
the
sea
ice
as
well
as
the
effective
albedo
that
comes
out
of
it.
So
so,
when
you
look
at
this,
I
mean
you
you'll
see
this
this
afternoon
in
the
exercise.
D
Two,
the
dry,
snow,
green
radius
is
actually
one
of
the
most
used
parameters
that
we
use,
because
what
happens
is
when
you
reduce
the
size
of
the
dry
snow
grains
that
increases
the
albedo
and
when
you
increase
the
size
of
them
that
reduces
the
albedo.
So
it's
it's
sort
of
inverse
and
that's
actually
our
number
one
parameter
that
we
will
adjust.
D
Sometimes
it's
necessary
to
go
beyond
that,
and
so
we
will
actually
adjust
the
melt,
onset
temperature,
and
so
that
means
that
that's
the
point
at
which
melt
begins.
D
Typically,
it's
somewhere
between
minus
one
celsius
to
zero
celsius
in
is
the
range
that
we
use,
and
so,
but
you
know
it
varies
by
region
and
time
of
year
and
so
forth,
and
so
we
often
use
that
as
another
parameter
to
adjust
and
then
a
third
one
that
we
sometimes
adjust
is
the
maximum
snow
grain
radius,
and
that
just
means
when
you've
achieved
that
melting
point
at
zero
c.
What
is
that
maximum
snow
grain
radius
that
you're
gonna
get-
and
you
know
it's-
it-
can
be
2500
microns
or
two
and
a
half
millimeters.
D
So
it's
pretty
big
anyway,
but
then
the
final
one,
and
that
was
kind
of
the
tricky
one
which
most
of
you
got,
which
is
or
or
responded
with,
which
is
not
the
right
answer
is
or
sorry
is
the
right
answer
for
this
because
we
asked
which
is
not
so
good
job
snow
conductivity
is
in
fact
something
used
in
the
thermodynamics,
but
it's
not
used
in
the
radiation.
So
it's
no
conductivity
doesn't
actually
impact
the
albedo.
So
so
number
four
was
the
correct
answer
and
nice
job.
C
All
right
great,
I
see
there
was
one
question
that
came
in
at
that
point:
I'm
gonna
save
that
for
the
the
live
discussion
in
just
a
minute.
I
want
to
go
ahead
and
go
through
these
questions,
so
we
don't
get
totally
derailed
on
one
topic
or
the
other.
So
the
fourth
question
is
paleo.
C
If
you
wanted
to
run
an
ice
age
climate
simulation,
so
that
means
five
to
six
degrees
celsius,
cooler
globally,
starting
from
a
pre-industrial
climate
which
model
component
would
take
the
longest
to
reach
equilibrium,
and
it
looks
like
most
people
said
the
deep
ocean
and
that's
right
so
the
atmosphere
just
for
reference,
spins
up
or
becomes
equilibrated
on
an
average
of
about
like
days
to
weeks,
depending
on
what
you're
looking
at
the
land
carbon
cycle.
I
think
that
does
take
a
while
as
well,
but
not
as
long
as
the
deep
ocean.
C
The
deep
ocean
takes
hundreds
to
thousands
of
years
to
really
equilibrate
and
arctic
sea
ice
takes
like
50
to
100
years,
usually
to
equilibrate.
So
if
you
wanted
to
run
this,
definitely
the
deep
ocean
is
kind
of
the
limiting
factor
in
how
how
much
you're
going
to
need
to
spend
in
computer
resources
to
spin
up
and
then
the
last
question
is
paleo-specific.
C
F
Okay,
maybe
I
will
pick
this
one,
I'm
sorry,
I
cannot
see
the
pool
results
right
now,
but
I
I
know
they.
Oh.
C
It
looks
like
so
seven
percent
said
specifying
the
orbital
year,
21
said
specifying
parameters
for
eccentricity,
obliquity
and
precession
in
the
fixed
parameters.
Mode.
No
one
said
specifying
solar
insulation.
52
said
either
the
first
two
and
21
said
either
the
first
or
the
third.
F
Okay,
it's
cool
that
no
one
selects
the
solar
installation
option.
That's
not
the
correct
answer.
So
there
are
multiple
ways
you
can
set
up
the
orbital
parameters
so
because
orbital
parameters
are
largely
deterministic,
so
for
the
past
climates
in
millions
of
years.
So
you
give
us
the
model
year,
the
time
period
of
your
simulation,
we
can
calculate
the
orbital
parameters
so
there's
one
way
you
can
do.
That
is
to
tell
the
model
the
simulation
years
you
want
to
do.
Then
the
model
can
calculate
the
parameters
on
the
fly
and
there
are
other
ways.
F
F
C
Great,
thank
you
all
right
so
and
sorry.
If
I
cut
anybody
off
who
had
been
almost
done
answering,
but
the
poll
ended
early,
so
I'm
gonna
go
ahead
and
move
to
the
question.
C
The
first
question:
I'm
gonna
go
ahead
and
take
it
from
the
chat
because
it
was
something
we
talked
about
today
and
then
we'll
move
on
to
the
one
that
got
asked
early.
So
vicky
asked
regarding
the
sea
ice.
Is
it
the
same
range
of
snowmelt
temperatures
that
the
land
ice
uses.
D
Yeah,
I'm
I'm
waiting
for
bill
to
chime.
In
you
know,
we
we
derive
the
melt,
onset
temperatures
basically
from
field
campaigns
in
the
arctic,
and
so
those
were
sort
of
typical
values
that
we
would
see
in
the
range
in
the
beaufort
sea,
for
example.
But
we
did
not
at
all
coordinate
with
the
land
ice
people,
and
so
I'm
not
sure
how
you
guys
said
it.
G
I
might
be
missing
some
background.
What
exactly
is
meant
by
the
same
range
of
snowmelt
temperatures.
G
G
G
So
it
would
for
sort
of
for
historical
reasons,
the
snow
schemes
in
the
land
model,
the
snow
schemes
and
sea
ice
model
have
not
been
identical,
also
because
it's
it's
just
snow
just
behaves
differently
when
you're
sitting
on
top
of
an
ice
platform
on
the
ocean
compared
to
a
you,
know
a
land
surface,
but
those
schemes
have
been,
I
think,
converging
a
bit
through
the
years,
but
they're
two
distinct
schemes.
C
Yeah
about
the
what's
different
about
sea
ice
and
land
ice,
and
one
of
those
big
things
is
salinity
so
on
sea
ice,
not
that
we
we
don't
take
all
these
factors
into
account,
but
when
bill
said,
snow
behaves
differently
on
like
sea
ice
versus
land
ice.
One
thing
that
can
change
is
that
some
of
the
the
salt
from
the
ocean
can
get
into
the
snow.
C
That's
on
top
of
the
sea
ice,
so
the
the
snow
becomes
a
bit
salty
and
that
can
affect
things
like
satellite,
retrievals
and
stuff
as
well,
and
that's
beyond
what
we
need
to
talk
about
here.
This
also
dovetails
well
with
the
question
that
was
asked
on
poll
emv,
which
is-
and
I
don't
know
who
asked
this,
but
are
there
any
plan,
developments
for
snow
modeling
and
processes
in
cesm
and
they
didn't
specify
a
particular
component.
So
I
think
this
is
open
to
land
ice,
land
or
sea
ice.
D
So
I'll
I'll
guess
I'll
start.
So
as
bill
alluded
to
you
know,
the
land
model,
snow
treatment
has
sort
of
evolved
separately
from
the
snow
treatment
on
sea
ice
and
I'm
as
far
as
I'm
aware,
I
don't
think
clm
5
has
any
really
new
fancy
snow
developments
going
on
there.
They've
had
wind
blown
snow
compaction
of
the
snow,
and
you
know
wetting
in
the
snow
features
in
there
for
several
years.
I
think,
but
I
you
know,
one
of
you
bill
or
gutter
can
correct
me
on
that.
D
In
terms
of
the
sea
ice
we've
always
had
sort
of
a
constant
density,
snow
pack.
We
don't
account
for
blowing
snow.
You
know
we
just
as
bill
said
again.
We
have
the
complexity
of
the
snow
is
moving
around
with
the
sea
ice,
and
so
it
is
a
very
difficult
thing
to
sort
of
keep
track
of.
D
So
any
new
variable
you
add,
for
the
snow,
has
to
be
carried
around
as
a
state
variable
too
and
ejected
with
the
sea
ice
and
so
forth.
That
being
said,
just
yesterday,
elizabeth
hunky
just
issued
a
pull
request
into
the
zeiss
consortium
for
brand
new
advanced
snow
physics
treatment,
and
so
this
will
introduce
a
new
snow
tracer
into
the
size
model.
Now
this
is
not
exactly
the
version
that
we
have
in
csm
currently,
but
it
will
be
one
that
we'll
be
using
as
of
csm3
in
a
few
years.
G
G
It's
been
shown
and
for
ins
land
ice
working
group
co-chair
john
lennox
has
worked
some
on
this
in
regional
models.
It's
been
down,
it
makes
a
difference
when
snow
is
to
the
surface
mass
balance.
If
snow
is
blown
around
from
one
region
to
another
region
and
that's
something
we
don't
yet
capture
in
cesm
but
we'd
like
to.
C
All
right,
so
one
thing
I
forgot
to
say
earlier
too,
was
that
I
would
invite
all
the
participants
to
go
ahead
and
turn
on
your
cameras.
So
this
can
be
more
of
a
discussion.
There
are
no
other
questions
in
the
chat
right
now.
So
if
you
have
a
question,
I
would
invite
you
to
either
put
it
in
the
chat
or
you
can
go
ahead
and
raise
your
hand,
and
I
can
call
on
you.
It
would
be
nice
for
us
answering
to
be
able
to
see
people's
faces
so
that
we
don't
feel
like
we're.
E
I
mean
I
I
do
have
some
some
things,
I
wonder
about
which
size
I
mean
yeah,
you
broke
the
eyes
now,
so
I
can.
I
can
speak
out
sure
what
what
what
is
like
what
is
thermodynamically
happening
at
the
base
of
the
sea
ice
and
and
the
exchange
with
with
the
operation
layer
like
what
what
is
physically
terminally
happening
over
there.
D
It's
a
good
question:
I'm
I
mean
there's
a
lot
going
on,
but
I
I'm
trying
to
get
at
some
of
the
specifics
of
what
you're
looking
for,
but
we
we
exchange
the
fluxes
of
fresh
water,
heat
and
salt
at
the
base,
and
so
this
happens
during
growth
and
melt,
and
so
you
can
have
positive
or
negative
fluxes,
depending
on
whether
you're
in
a
growth
period
or
a
melt
period.
And
you
know
traditionally
we're
actively
working
on
this
alice.
Has
a
paper
hopefully
accepted
soon
talking
about
the
way
that
we've
traditionally
done?
D
The
salt
fluxes
in
the
model
we've
always
sort
of
naively
assumed
that
the
sea
ice
was
4psu
and
we
used
that
for
our
flux,
calculation,
it
simplified
things,
but
the
new
thermodynamics
scheme
that
we
have
so
from
adrian
turner.
D
We
have
a
freezing
point
at
the
base
of
the
ice
that
depends
on
the
salt
and
so
there's
a
lot
more
complex
interactions
going
on
and
so
we're
actively
working
on
improving
that
flux
exchange
at
the
base
of
the
ice
with
the
ocean.
E
Okay
and
and
how
does
it
work
with
like?
If
I
mean
this
is
pure
theoretically,
but
if
you
have
like
a
more
thinking
about
energy
or
heat
exchange
like
if
you
have
like
a
warm
current
or
something
floating
below
your
eyes
or
like.
D
It
doesn't
right
so
we
I
mean
we
also
couple
momentum
at
the
base
too,
but
yeah,
basically
at
the
bottom
of
the
ice.
There's
an
energy
balance
equation,
that's
solved,
whereas
the
growth
melt
is
equal
to
the
amount
of
heat
coming
from
the
ocean,
but
also
balanced
by
the
amount
that's
conducted
away
from
the
base
and
what
happens
is
if
you're
conducting
more
away
from
the
base,
then
you're
actually
getting
from
the
ocean
you're
going
to
have
growth,
whereas
if
you
have
more
heat
coming
from
the
ocean
than
is
conducted
away,
then
you
get
melt.
D
So
that's
that's
sort
of
the
simplistic
view,
at
least
at
the
bottom
of
the
ice.
So.
C
Great,
so
we
have
a
follow-up
question
here
from
rudra.
Do
you
want
to
ask
your
question:
live
rudra.
H
Yeah
hi
everyone.
So
recently
I
read
a
paper
in
which,
in
c2,
observations
were
used
for
determining
that
how
arctic
cyclones
and
the
following
processes
may
affect
dci's
loss
in
the
arctic
right.
So
they
mentioned
that
the
observations
and
the
results
were
pretty
accurate
as
they
used
very
small
spatial
scale,
whereas
in
model
huge
means,
comparatively
larger
special
scales
are
represented
right.
So
how
accurately
the
oceanic
activities
beneath
the
ci,
such
as
heat,
transfer
between
subsurface
solution
and
ci,
or
they
mixingly
attacked
or
if
they
ecman
pumping.
So
how
accurately.
C
So
I
think
the
the
answer
is:
we
know
that,
like
the
mixed
layer,
depth,
for
example,
is
we
probably
don't
have
high
enough
ocean
resolution
to
really
adequately
resolve
it
in
some
places
and
they're,
especially
in
polar
regions
or
the
arctic,
where
you
get
different
sorts
of
stratification
because
of
the
freshwater
lens
at
the
top
and
stuff
like
that,
we
don't
necessarily
have
all
those
fine
details
in
the
ocean
model,
so
I
would
say
that,
because
of
the
large
scale,
and
also
because
of
you
know
vertical
resolution
as
well
as
horizontal
resolution,
there
are
going
to
be
some
inadequacies
compared
to
like
observations
of
those
sorts
of
features
in
the
real
ocean.
C
We
also
know
because
of
the
the
coarse
horizontal
resolution.
Primarily
we
don't
resolve
coastal
currents
super
well
so
like
around
antarctica
or
places
around
the
arctic.
If
you're
interested
in
how
the
the
coastal
currents
are
going
to
affect
the
ci
sneer
near
places
where
people
live,
for
example,
those
are
not
going
to
be
super
well
resolved,
and
one
thing
that
is
now
in
the
size
model
that
we
don't
have
in
cesm2.
C
So
you
can
get
sea
ice
that
in
an
area
where,
where
the
the
shoreline
is
pretty
shallow,
the
ice
can
be
thick
enough,
that
it
basically
wedges
into
the
to
the
bottom
of
the
ocean,
and
then
it
gets
stuck,
and
so
that's
land
fast
ice
and
that's
pretty
important
actually
for
a
lot
of
settlements
in
in
the
arctic,
and
it
can
be
important
as
well
in
the
antarctic,
for
biology
and
stuff
like
that,
like
penguins
use,
landfast
ice
a
lot,
so
we
do
have
that
in
incise
and
it
will
be
in
cesm3,
but
it's
not
in
cesm
2
right
now.
C
So
I
think
one
thing
you
just
have
to
accept
with
any
sort
of
modeling
is
that
you're
never
going
to
quite
resolve
all
the
processes
or
all
the
features
that
are
going
on.
You
can
always
go
to
a
you
know,
a
finer
vertical
or
horizontal
resolution
and
find
something
new
that
you're
not
going
to
resolve,
and
so
I
think
we
do
a
reasonable
job,
but
I
do
think
there
are
probably
some
inadequacies
there.
C
E
Yeah
sure,
well,
actually,
honestly,
I
I'm
not
a
ocean
scientist.
I
was
just
wondering
because
I
found
in
the
lecture
you
are:
you
are
developing
some
biogeochemical
component
and
also
water
isotopes
in
season
in
sea
ice.
I
was
just
wondering
what
kind
of
biogeochemical
process
will
be
included
in
future
release
versions.
C
I
don't
dave,
probably
knows
more,
so
I
was
going
to
say
one
of
the
things
that's
new
in
cesm2
is
that
we
have
this
prognostic
sea
ice
salinity,
which
means
the
salinity
throughout
the
column
of
the
sea
ice
changes
over
time.
In
the
past
we
just
had
some
fixed
profile,
and
we
said
this
is
what
it
is
everywhere,
but
that's
not
really
how
sea
ice
works.
C
Another
reason
that
that
was
important
is
that
having
a
prognostic
salinity
is
relevant
for
biology
and
biology
that
like
lives
in
the
ice
strata,
so
I
would
say
this
is
a
development,
I'm
not
sure
who
is
working
on
this
directly
for
cesm
right
now.
I
do
know
that
seis
itself
has
the
capacity
to
look
at,
like
you
know,
phytoplankton
and
including,
like
diatoms
and
stuff,
like
that
within
the
ice
column
and
looking
at
like
different
nutrients.
C
D
Yeah,
I
haven't
worked
on
it
personally
either,
but
I
know
that
nicole,
jeffrey
at
los
alamos
national
lab
is
she's
actually
working
on
the
full
column
biogeochemistry,
which
will
be
available
in
the
next
release
of
the
size
consortium
model
which
again,
as
I
mentioned,
is
not
going
to
be
available
in
csm
until
csm3.
D
We're
we're
developing
it
currently,
and
so
this
contains
numerous
bgc
tracers
within
this,
and
it
is
the
full
column.
Currently
in
csm2
we
have
what's
called
the
skeletal
or
just
bottom
layer
bgc
in
there,
and
so
there
is
the
possibility
of
doing
something
like
primitive
coupling
with
the
sea
ice
in
the
ocean.
D
But
we
have
not
really
explored
that
and
but
you
know,
keith
lindsey
from
the
biogeochemistry
working
group
have
we've
been
talking
a
lot
about
trying
to
do
some
experiments
where
we're
coupled
with
the
marble
ocean
biogeochemistry
component
and
try
to
make
sure
we
get
those
exchanges
better
represented.
You
know,
and
we
would
try
to
do
it
mostly
from
the
marble-centric
perspective.
So
what
are
the
main
species
and
components
of
that
model
that
we
could
also
simulate
in
the
sea
ice?
And
so
that's?
C
Great
thanks,
I
see
that
tim
says
he
has
a
follow-up
question,
so
go
ahead
and
ask
your
question
tim
and
then
there's
someone
else
after
that.
So
tim
go
ahead
and
then
you'll
go
after
him.
E
Yeah
sure
yeah,
I
was
chewing
a
little
bit
within
my
head
with
the
fact
that
you,
I
think
you
explained
right
away
that
that
sea
ice
when
it
moves
into
the
land
it
kind
of
becomes
grounded
so
to
say,
like
land
trust
ice,
you
call
it
if,
if
it's
large
enough
like
theoretically
it
could
change
the
it
could
change
like
the
the
sanded
blow
or
whatever,
like
the
c
ground
blow
it
like,
like
you
ripple
it
a
little
bit
onwards
does.
Is
it?
Is
it
representative.
D
D
I
I
think
that
there's
talk
again.
This
would
be
development,
but
there
would
be
talk
about
some
sediments
from
the
grounded
ice
getting
trapped
within
the
ice
and
being
carried
around,
because
that
sort
of
sediment
iron
can
be
very
important
for
the
biogeochemistry
cycle.
But
it's
again
we
just
we
cannot
represent
that
at
this
point
explicitly,
but
it
is
something
we
we
hope,
maybe
in
the
next
four
to
five
years,
so.
C
I
also
want
to
bring
up
for
for
bill.
Hopefully
he
can
elaborate
on
this
some,
but
right
now,
at
the
face
of
a
glacier,
for
example,
I
I
don't
think
we
have
ice
shelves
and
at
the
face
of
a
glacier,
we
don't
really
represent
calving.
We
have
like
an
iceberg.
C
For
example,
we
just
have
like
foxes
that
just
go
in
entirely
to
the
ocean,
and
so
it's
not
it's
not
exactly
like
an
iceberg,
but
bill
might
be
able
to
comment
on
either
potential
where
what
the
state
of
that
is
now
and
then
what
potential
developments
there
are,
especially
with
the
new
upcoming
ocean
model,
where
we,
we
could
have
potentially
changing
ice
shelves
or
glaciers
at
the
the
interface
of
the
ocean.
G
Sure,
well,
a
big
thing.
That's
coming,
which,
which
you've
heard
about
is,
is
mom
6,
replacing
pop
and
pop.
You
may
have
heard
when
you
get
to
the
ice
shelf
pop
doesn't
have
any
cavities
under
the
ice
shelf.
It
just
has
a
wall
where
the
where
the
observed
shelf
edge
is
whereas
mom
6
has
the
capability
to
circulate
the
ocean,
underneath
the
ice
shelf
and
and
back
out
the
top,
and
in
some
cases,
for
example,
underneath
the
big
ross
and
dulcinea
shelves.
G
You
actually
have
marine
ice
being
deposited
at
the
surface
of
the
ice
shelf
because
we
haven't
had
an
ocean
model
that
can
do
this.
We
haven't
worried
about
this,
but
it's
something
that
we'll
be
talking
about
how
to
handle
the
future.
Another
thing
we'll
be
talking
about
is
when
you
have
ice
shelves
advancing
and
retreating.
G
The
land
ice
requires
a
surface
mass
balance
from
a
computer
in
the
land
model,
and
so
that
means
that
the
boundary
between
the
land
and
the
ocean
has
to
change
in
some
sense,
or
rather
the
overlap
region
between
the
land
and
the
ocean
has
to
change,
because
the
floating
ice
is
like
part
of
the
land
model
and
the
sea
beneath.
It
is
part
of
the
ocean
model,
and
so
that's
an
upcoming
software
engineering
challenge
to
allow
all
that
to
change
dynamically.
C
Great
thanks
so
ben
you
said
in
the
chat
that
you
have
a
question.
So
go
ahead
and
unmute
and
ask
your
question.
I
Sure
yeah,
so
it's
maybe
a
bit
specific,
but
I
know
that
historically,
like
a
lot
like
most
unit
models,
have
a
lot
of
trouble,
simulating
sort
of
realistic
polyneias
like
sort
of
open
ocean
plenty,
isn't
all
about
like
the
mother,
moderates
bernie
and
the
big
ones
that
sometimes
happen.
I
In
addition-
and
I
know
there
are
some
papers
recently
because
I
was
just
looking
at
it-
that
use
csm
at
a
high
resolution
and
and
it
works
a
lot
better-
I'm
wondering
I
guess
I
guess
my
question
is-
is
sort
of
like
how
do
you
expect
that
to
change
as
with
resolutions
and
also
with
changing
an
ocean
model?
Does
that
kind
of
throw
a
lot
of
stuff
out
the
window?
I
don't
know
if
anyone
like
the
sort
of
specific.
I
D
To
yeah,
I
I
I
have
a
strong
opinion
on
this,
so
yeah
open
ocean
plenty
is
are
an
interesting
challenge
because
they
are
really
ultimately
the
interaction
of
local
winds,
but
also
mixing
of
warm
waters
from
below
in
the
ocean,
and
so
you
just
have
this
confluence
of
events
that
happens
the
real
problem,
the
real
challenge
with
these
things
is,
you
know
in
the
observed
sea
ice
record,
we
have
seen
two
three
would
lc
plenias,
you
know,
and
so
they're,
not
very
common,
and
so
the
the
thing
is
it
is.
D
You
know,
I
I
think
it's
fair
to
say
that
the
course
resolution
models
don't
really
simulate
them
and
that's
true,
and
it's
probably
because
the
resolution
is
not
fine
enough
in
the
ocean
to
really
resolve
that
sort
of
topographic
upwelling.
That
is
leading
to
those
sort
of
events
that
may
change
with
mom
six,
because
we're
gonna
go
to
sort
of
a
standard
resolution
of
two-thirds
of
a
degree
instead
of
one
degree,
so
maybe
that
would
be
sufficient.
D
Maybe
not,
however,
the
the
reason
I
have
a
strong
opinion
on
this
is,
I
feel,
like
the
high
res.
The
tenth
of
a
degree
goes
too
far
the
other
direction.
You
know
what
you
see
in
those
high
resolution
runs.
D
Is
you
see
madras
polyneias
like
every
10
years,
you
know
they're
they're
almost
like
clockwork
and
and
and
and
this
is
in
an
1850
sort
of
cold
climate.
So
I
find
that
kind
of
unrealistic.
In
my
opinion,
although
you
know
there's
people
who've
debated
on
this,
we
don't
really
know
how
often
they
might
have
occurred
in
1850
or
whatever,
but
there
is
a
paper
by
frank,
brian
and
peter
gent.
D
That
shows
that
the
high
resolution
tends
to
have
too
much
poleward
heat
transport
in
the
ocean,
so
you're,
probably
getting
excessive
heat
into
these
regions,
and
that's
what
makes
it
much
easier
in
the
hi-rez
case
to
form
these
open
ocean
polynias
there's
also
a
subtle
distinction
about
the
definition.
Are
you
just
seeing?
Basically
the
wood
lc
circulation
coming
around
and
then
closing
off
and
all
of
a
sudden
you
have
this
big
hole
in
the
middle.
You
know
so
so
you
know.
D
There's
some
interesting
debates
on
this,
and
I
know
that
a
lot
of
people
working
in
particular
on
this
ihes
project,
which
I
think
you're
referring
to
they
get
very
excited
about.
The
bellini
is
in
this,
and
so
but
anyway,
that's
that's
sort
of
my
two
cents
worth.
C
I'll
I'll
chime
in
here
not
about
the
waddell
plenty.
I
know
you
asked
that
specifically,
but
one
thing
we
have
noticed
in
cesm2
compared
to
ces
m1.
C
Is
that
there's
a
lot
more
open,
open
water
ice
production,
basically,
which
we
call
frazzle
ice
production
in
the
model
in
cesm2
compared
to
cesm1,
and
a
lot
of
this
is
in
coastal
regions
and
so
there's
a
lot
of
ongoing
work
right
now
by
myself
and
others
at
ncar
and
not
at
ncar
to
understand
what
exactly
is
going
on
like?
Are
we
getting
more
polinias,
or
is
it
just
like
a
shift
in
processes
that
are
going
on
in
the
model
and
we're
not
actually
seeing
a
change
in
polyneias?
C
So
I
would
say
that's
kind
of
an
area
of
open
research.
We
do
know
that
in
the
standard
not
high
resolution
version
of
the
model,
as
I
mentioned
before,
like
we
don't
adequately
simulate
coastal
currents
and
we
don't
adequately
adequately
simulate
winds
that
the
downslope
winds
that
drive
most
coastal
pollenias
and
so
they're
going
to
be
deficiencies.
C
All
right,
I
think,
that's
all
of
the
questions
that
are
in
the
chat.
I
don't
see
anyone's
hands
up,
but
I
would
invite
people
to
go
ahead
and
put
your
hand
up
or
oh,
we
got
one
in
the
chat
here.
Charles
do
you
want
to
go
ahead
and
mute
yourself
and
ask
yourself.
E
Yes,
I'm
looking
at
the
surface
behind
you-
and
I
was
wondering-
is
that
a
typical
sea
ice
surface
or
you
know,
I
mean
the
ice
sheets.
I
believe
that
take
and
blocky
and
cover
the
entire
sea
surface.
Or
do
you
treat
that
surface
behind
you
as
say
a
regular
ocean
surface
from
the
radiation
point
of
view?
And
how
do
you
know
where
those
kind
of
surfaces
exist
right.
C
Okay,
I'm
going
to
pin
myself,
which
feels
really
weird-
and
I
don't
know
if
I
know
how
to
un.
Oh,
I
know
how
to
unpin.
So
what
you're
seeing
behind
me
is
a
picture
of
sea
ice.
This
is
what
you'd
call
pancake
ice
it's
from
the
antarctic
and
what
we
do
in
the
model
is
for
each
grid
cell.
We
would
look
at
a
picture
like
this.
For
example,
imagine
this
is
a
grid
cell
behind
me
and
there
is
a
ice
thickness
distribution.
C
C
So
we
do
model
the
distribution
of
thicknesses
of
ice
within
a
grid
cell
and
we
also
model
the
percent
of
a
grid
cell,
that's
covered,
so
you
could
say
for
this
grid
cell.
I
might
say
that
two-thirds
of
it
are
covered
with
ice
and
so
the
concentration
that
the
bulk
concentration
would
be.
You
know
like
.66,
which
is
the
concentration
of
sea
ice
in
this
grid
cell
and
then
within
that
we'd
say.
Okay,
what
percentage
of
that
concentration
is
this
thicker
ice
versus,
maybe
like
this
thin
small
ice?
C
And
this,
as
I
said
this
picture
specifically,
is,
is
pretty
small
scale,
so
this
is
much
smaller
than
what
we
we
look
at
in
our
model.
But
this
sort
of
thing
exists
as
you
go
up
in
scale
as
well,
so
so
yeah
and
then
you
know
we
for
each
one
of
those
different
thicknesses.
We
calculate
the
heat
fluxes
and
we
calculate
the
snow.
C
D
And
so
again
the
pot
model
has
its
own
surface
flux,
treatment,
its
own
albedo
treatment.
It's
not
a
very
sophisticated
albedo
treatment.
It
basically
just
sets
the
albedo
of
all
that
dark
area
to
0.06,
but
you
know,
but
then
you
you
for
each
grid
cell
there's
a
fraction
of
open
water
and
there's
a
fraction
of
sea
ice
cover
and
the
open
water
part
is
simulated
by
pock
and
the
sea
ice
covered
part
is
simulated
by
sice.
C
All
right
again,
I
don't
see
any
questions
yet
in
the
chat
or
raised
hands.
We
haven't
had
any
on
paleo.
So
if
there's
anyone
who
has
any
questions
about
paleo,
this
is
your
chance.
C
B
Yeah
go
ahead.
Yeah
I
actually
do
have
a
question
about
paleo
for
either
jung
or
betty
so
csm
when
you're
developing
it
you're,
basically
validating
it
against
historical
simulation
or
historical
records.
The
instrumental
record
right,
so
I
was
just
wondering
what
you
actually
use,
what
kind
of
data
sets
or
whatever
to
make
sure
that
the
csm
that
was
validated
for
running
kind
of
modern
climate
accurately
can
accurately
simulate
these
paleo
climates?
You
know
lgm
or
the
pliocene,
or
you
know,
however,
far
back.
E
Well,
let's
let
john
answer
that,
because
he
has
been
working
a
lot
on
this
for
cesm2
looking
at
the
last
glacial
maximum,
which
is
21
000
years
ago,
and
also
warm
climates
like
esc,
50
millimeters
go
with
high
co2.
So
would
you
like
to
expand
on
that.
F
Yeah
sure
so
I
think
firstly,
if
you
compare
the
different
model
components,
maybe
a
atmosphere
is
making
a
biggest
change,
see
in
this
context,
for
a
development
from
csm1
to
csm2.
F
So
most
of
the
change
is
from
the
atmosphere
component.
So
when
you
evaluate
the
model
against
observation,
so
usually
people
do
animate
only
simulation
first,
so
you
use
a
prescribed
observation,
observation
of
sea,
surface
temperature
and
cis
to
force
the
atmosphere
model.
So
you
compare
the
simulation,
say
the
precipitation
cloud,
radiative,
forcing
etc
against
present
day
to
say
how?
How
do
you
simulate
those
variables?
F
So
you
put
a
transcend
4c
into
the
simulation,
including
aerosol
emission.
Then
you
can
perform
this
transient
historical
simulation.
That's
really,
I
think,
very
important
here,
because
this
simulation
enables
you
to
compare
with
the
satellite
observations
more
closely.
So
that's
exactly
the
same
time
period
with
the
same
errors
of
forcing
greenhouse
gases,
etc.
So
you
can
take
a
closer
look
at
the
historical
simulation
and
for
different
model
components.
There
are
a
lot
of
people
looking
at
their
their
own
model,
components
that
they
are
interested.
F
So
there
are
a
lot
of
eyes,
look
at
the
simulation
and
to
to
examine
the
variables
they
are
interested
in.
So
if
there's
a
problem,
for
example,
if
you
have
too
much
cis
cover
over
the
laboratory,
then
we
know
there's
a
problem,
so
we
need
to
diagnose,
what's
going
on
what
went
around
there
and
we
can
make
changes,
make
improvements
from
there.
E
Great,
thank
you,
john.
Do
you
want
to
also
talk
about
what
kind
of
paleo
records
we
use
and
some
of
the
new
techniques
for
working
on
those.
D
F
So,
okay,
I
will
talk
about
the
water
isotopes
first,
so
the
what
isotopes?
Currently
it's
only
in
the
fully
coupled
1.2
or
1.3.
So
in
fully
coupled
csm2,
we
don't
have
this
capability
and
this
is
really
unfortunate,
but
we
are
working
on
that.
Hopefully
we
can
get
this
capability
online
with
csm2
later
so
this
is
particularly
important
for
paleo.
F
F
F
So
those
can
be
used
to
evaluate
model
and
of
course,
the
spatial
coverage
and
the
availability
are
in
general
limited,
but
you
can
still
get
some
idea,
for
example,
for
the
ice
age
climate.
If
we
can
get
a
thousand
proxy
sst
records,
we
can
have
a
more
or
less
idea
about
how
cold
the
climate
is,
and
you
can
also
take
one
step.
F
C
Great-
and
I
want
to
point
out
too,
that
ryan
put
in
the
chat
the
link
to
the
paper
that
ron
just
mentioned,
I
think
we're
closing
in
on
the
time.
There
is
one
last
question,
john:
you
do
you
want
to
ask
that
live.
I
Yeah.
Thank
you.
My
questions
regarding
to
the
lecture
like
there's
a
plot,
comparing
the
eye
set
observation
of
the
eye
thickness,
comparing
with
the
model
stimulation
from
the
wall,
cam
and
cam,
and
I
see
there's
a
little
bit
large
discrepancy
between
them.
I'm
just
wondering
like
when
you
see
a
discrete
this
difference
between
observations
and
model
which
variables
that
you
think
is
important
to
causing
this
difference
in
the
model,
and
how
do
you
find
it
out
from
when
you
analyze
when
you
analyze
the
models
and
yeah?
C
Okay,
so
dave
put
a
link
to
a
paper
that
I
led
on
this
sort
of
question
in
the
chat,
but
for
this
instance
with
looking
at
the
ice,
and
actually
this
is
not
too
dissimilar
from
what
I
do.
What
I
would
do
if
I
were
looking
at
the
atmosphere
or
something
like
that
is
when
you
see
something,
that's
not
what
you
expect.
C
I
start
looking
at
some
of
the
processes
so
like
I
looked
at
energy
budgets
and
I
looked
at
mass
budgets
between
wacom
and
cam
and
tried
to
figure
out
how
those
were
different
and
then
in
this,
in
this
particular
case
for
cesm2,
when
we,
it
wasn't
necessarily
a
discrepancy
from
observations,
but
a
big
difference
between
the
models
that
we
weren't
expecting
and
then
also.
You
know
that
then
can
relate
to
a
discrepancy
with
observations.
C
You
know
I
found
that,
like
the
surface
melting
was
really
different
between
the
two
and
so
then
I
think
about
okay,
well,
what's
different
in
the
model
in
the
surface
like
what
would
cause
differences
in
the
surface
melt,
they're
using
the
same
snow
scheme.
So
that
can't
be
it.
C
You
know
that
sort
of
thing
and
in
this
particular
case
for
the
paper
that
dave
linked,
we
found
that
the
atmosphere
was
really
different,
and
so
I
think
it
becomes
a
little
bit
of
like
a,
I
don't
wanna
say
a
goose
chase,
but
you
know,
like
you,
find
one
little
piece
of
something
that
might
be
a
nugget
that
helps
you
get
to
the
answer
and
you
just
kind
of
keep
following
that
trail,
and
sometimes
you
do
that
and
you
get
to
a
dead
end
and
you're
like.
I
really
have
no
idea
like
this.
C
That
wasn't
the
reason
and
then
you
kind
of
have
to
step
back
and
think.
How
else
can
I
do
this
and
I
think
that's
a
lot
of
just
the
scientific
process
that
probably
most
of
us
go
through
if
any
of
the
end
car
scientists
know
it.
You
know
a
priori
what
your
results
are.
Gonna
be
then
go
ahead
and
chime
in,
but
usually
for
me,
it's
very
much
a
discovery
process
and
you
can
the
longer
you
go
on
and
more
you
work
with
models.
I
think
you
you
get
a
better
intuition
for
okay.
C
This
might
be
a
culprit
for
this,
and
so
you
can
kind
of
start
on
the
maybe
the
right
path,
what
appears
to
be
more
easily
but
there's
still
things
that
even
for
those
of
us
who've
worked
with
models
a
long
time.
I
mean
I've
worked
with
models
for
now
like
almost
15
years
and
dave's
probably
worked
with
models
for
a
lot
longer
than
that,
and
I'm
talking
about
sea
ice
models
specifically
here,
but
then
you
get
something
like
with
cesm2
development.
This
is
not
a
secret
there.
C
There
was
a
big
problem
we
had
with
the
labrador
sea
freezing
over
and
we
did
this
sort
of
thing
where
we
tried
to
investigate
fluxes
and
we
tried
to
investigate.
You
know
all
the
things
that
we
thought
might
be
causing
it
and
we
we
kind
of
just
we
still
don't
exactly
know
what
the
the
culprit
is.
But
you
know
yeah
it's
just
a
process,
and
it
sounds
like
that.
C
It
looks
like
betty
linked
a
paper
where
zhang
did
the
same
sort
of
thing
for
the
for
the
lgm
in
the
chat
as
well.
If
you're
interested
in
linking
that
we
are
at
9
45,
I
think
the
session
was
supposed
to
end
at
9
45.
So
I
want
to
be
respectful
of
people's
time.
One
last
question
did
come
in
from
rudra,
so
if
you
need
to
log
off,
we
understand
that
that
makes
sense
rudra.
If
you
want
to
ask
this
last
question,
live,
please
go
ahead.
H
Yeah,
I
was
just
wondering
that
in
in
the
csm2
model,
what
role
does
the
ice,
albedo
and
sea
ice
roughness
would
play
in
the
outgoing
and
incoming
long
wave
radiation?
D
I'm
happy
to
answer
this
one.
I
I
mean,
I
think,
if
I
understand
your
question
correctly,
the
shortwave
and
the
long
wave
computations
are
completely
separate,
so
the
long
wave
computation
has
associated
with
it
an
emissivity
which
is
basically
how
much
of
the
lung
wave
goes
up.
D
And
then
you
have
incoming
long
wave
from
the
atmosphere,
and
so
you
end
up
with
a
net
flux
of
long
wave,
so
we
use
a
constant
emissivity
in
the
site
model,
and
so
so
really
I
mean
there
are
probably
coupled
feedbacks
that
would
eventually
lead
to
changes
in
the
long
wave,
but
just
directly
in
terms
of
the
surface
sea
ice,
it's
solely
a
function
of
surface
temperature
and
a
constant
emissivity.
D
So
I
I
don't
see
how
the
albedo
would
directly
impact
that
I
mean
there
would
be
an
indirect
effect,
possibly
in
terms
of
the
sea
ice
roughness.
Similarly,
we
have
a
constant
sea
ice
roughness
everywhere.
So
the
only
thing
that
affects
the
drag
over
the
surface
is
a
stability
in
the
atmosphere
above
and
so
that
will
change
the
exchange
coefficients
over
sea
ice.
D
But
the
roughness
currently
is
is
constant
and
I
think
earlier
there
was
a
discussion
a
bit
about
the
fluxes
at
the
bottom
of
the
ice
and
the
fluxes
on
top
of
the
ice.
D
We
are
hoping
to
move
to
a
more
sophisticated
form,
drag
formulation
which
will
actually
explicitly
take
into
account
roughness,
underneath
the
ice
and
above
the
ice
as
well
so-
and
I
was
going
to
add
something
a
little
bit
to
alice's
point-
that
when
we
do
these
long
pre-industrial
simulations,
what
we
try
to
do
is
we
try
to
find
a
balanced
climate
for
1850
and
we
try
to
adjust
the
sea
ice
parameters
to
sort
of
fit.
D
What
we
believe
is
the
best
sea
ice
conditions
for
1850,
and
we
so
we
did
this
for
the
wacom
simulations
and
we
came
up
with
a
set
of
albedos
and
then
the
idea
was
we
wanted
to
use
the
same
parameters
in
the
cam
style
simulations,
but
it
turns
out
with
the
cam
style
simulations.
D
The
sea
ice
was
much
much
thinner
and
and
some
of
the
reasons
for
the
sir
analysis
paper,
as
I
pointed
out,
analysis
talked
about.
But
then
there
was
sort
of
this
feeling
that
we
weren't
going
to
separately
adjust
those
parameters
for
the
cam
simulations,
and
so
we
sort
of
had
to
live
with
the
thinner
ice
in
the
cam
style
simulation.
So
but
it
is,
it
is
a
process
when
we're
developing
a
model
where
we're
trying
to
achieve
a
balanced
climate
for
1850,
and
then
we
can
start
our
20th
century
runs
from
that.
So.
C
All
right:
well,
I
think
we
are
at
time
we
got
some
really
good
questions,
so
I
just
want
to
thank
everybody
for
attending
this.
Thank
you
to
the
ncar
experts
for
providing
your
expertise
and
we
hope
to
see
all
of
you
guys
as
part
of
our
working
groups
and
into
the
future,
when
you're
doing
whatever
cool
research
and
science
you're
doing
as
well.