►
From YouTube: 2020 CESM Tutorial - Land modeling Q&A
Description
Day 3 of the 2020 CESM Tutorial featuring Land Modeling Questions and Answers.
The CESM Tutorial will consist of:
Lectures on simulating the climate system, practical sessions on running CESM, modifying components, and analyzing data.
For more information:
http://www.cesm.ucar.edu/events/tutorials/2020
Community Earth System Model (CESM) is a fully-coupled, community, global climate model that provides state-of-the-art computer simulations of the Earth's past, present, and future climate states.
B
She
wasn't
she
and
her
daughter
there.
Last
night
too,.
A
B
B
B
B
A
A
A
B
A
B
B
B
B
A
B
A
B
B
B
A
B
A
B
I
mean
I
can
delete
them,
but
the
only
problem
I
can
see
is
seeing
how
we're
using
the
links
out,
I
feel
like
it
might
confuse
the
students.
If
they
get
more
links,
it
may
just
be
more
yeah,
I
mean
I'll,
send
them
a
chat
in
the
google
chat
that
we
have
together
with
you
mean
in.
A
A
A
A
Hey
ryan
yeah:
do
we
normally
post
a
participant
list
for
the
tutorial,
or
do
we
maybe
just
only
make
it
available
to
the
participants,
so
they
can
network
beyond
the
event.
B
Yeah
so
in
previous
years
we
haven't
posted
on,
but
I
mean
we
do
kind
of
have
a
participant
list
with
the
page
with
the
bios
and
the
pictures.
A
Yeah
yeah,
I
was
just
wondering
about
them,
I
mean
I
guess
they
can
just
find
each
other
on
google.
You
know
just
find
each
other
if
they
wanted
to
have
a
you
know
a
list
of
each
other's
emails.
B
B
A
B
B
Maybe
we
can
like,
I
don't
know,
let's
see
I
mean
I
don't
know
if
they
want
to
share
their
email,
but
I
mean
maybe
we
can
put
it
out
there
saying
hey
if
anybody
would
like
to
have
their
email
shared
with
everybody
in
the
group.
Let
us
know
or
drop
it
in
the
slack.
B
A
C
B
And
I'll
be
right
back,
I'm
gonna
grab
some
coffee
real.
C
A
C
B
All
right
so
elizabeth
I'm
going
to
go
ahead
and
make
you
the
host
host
and
say
that
co-host.
A
B
Yeah,
it's
been
working
out
good
because
I've
been
starting
it
around
eight
I'm
gonna
get
up
and
started
so
I
can
get
the
youtube
stuff
started.
Okay,
because
it
gives
me
that
control
so
a
good
note
for
future
meetings.
If
you
want
to
stream.
B
B
A
C
C
C
All
right,
let's
get
this
started
good
morning.
Everybody
welcome
to
our
third
day
of
the
tutorial
and
a
few
quick
announcements
before
we
jump
into
the
third
q,
a
panel
session
by
the
landmark
working
group
to
begin
with
some
general
general
topics.
The
csm
lecture
that
took
place
on
monday
morning
and
by
gokan
is
now
posted
online
on
the
csm
website.
So,
if
you'd
like
to
rewatch
it,
it
is
there
the
q,
a
sessions
will
be
posted
shortly.
C
Oh
ryan
is
updating
them
and
uploading
them
as
a
as
to
take
place
for
the
q
a
session.
There
are
more
and
more
questions
that
are
being
asked
on
the
using
the
link,
which
is
great,
and
we
realize
that
maybe
sometimes
we
might
not
be
have
the
time
to
go
through
all
of
them.
So
if
some
of
the
questions
remain
unanswered
either
during
that
session
or
throughout
the
week,
please
let
us
know
and
we'll
try
to
get
answers
for
you
for
tomorrow.
C
Some
of
you
have
requested
private
breakout
rooms
for
today
to
discuss
topics
right
now
we
had
like
when
I
was
sneaking
to
the
slack
channel.
Some
of
you
want
to
talk
about
ocean
and
phytoplankton
and
create
and
ask
these,
that
is
to
create
a
breakout
room.
There's
another
topic
of
discussion
about
exoplanets.
That's
about
to
take
place.
It's
great
to
see
you
self-organized,
so
keep
it
up
and
yeah
if
you
need
a
breakout
room,
of
course,
do
not
hesitate
to
contact
elizabeth
for
this
purpose
about
the
office
hours.
C
Yesterday
we
sent
you,
we
sent
you
an
email
with
some
comments,
and
so
we
did
create
some
links
for
you
to
ask
questions
beforehand,
using
slidero
for
this,
and
just
to
remind
you
that
we're
going
to
only
answer
questions
of
like
for
today's
link
and
on
the
other
links
for
the
next
couple
of
days
will
be
entered
on
the
day
of,
and
it's
the
same
here
if
at
the
end
of
of
the
session,
some
questions
remain
unanswered.
Please
email
us
and
we
will
try
to
get
to
them.
C
Let's
see
and
something
that
we've
already
said
yesterday,
but
just
in
case
you
couldn't
make
it
to
the
office
hour
session
yesterday
or
yesterday's
announcement.
Please
try
to
download
the
latest
version
of
the
practicals
because
I've
updated
them.
I've
updated
some
of
them
here
last
week.
So
maybe,
if
you
downloaded
them
several
weeks
ago,
then
you
might
not
have
the
latest
version
and
try
to
go
through
the
session
before
the
office
hours
to
get
the
most
of
it.
C
So
that's
all
I
have.
If
any
of
you
have
a
we
have
time
for
a
couple
of
questions.
If
some
of
you
have
some
please
go
ahead,
I
can
try
to
answer
them.
C
C
C
D
D
D
C
C
E
Sorry,
actually,
two
people
just
posted
in
the
chat
asking
if
we're
still
presenting
the
two
slides
on
ourselves.
C
No,
nothing
is
not
not
necessarily
when
I
when
we
saw
like
this.
This
was
just
an
idea
that
we
were
trying
to
to
cook,
but,
like
we
saw
that
you
guys,
like
figure
it
out
on
your
own,
on
the
slack
when
we
saw
the
comments
and
you
I
posted
your
articles
and
talk
about
yourself
already
creating
those
meeting
rooms
about
two
topical
discussions.
F
C
Hi
danica
all
right,
so,
let's
get
started
with
this
third
q.
A
panel
and
keith
olsen
will
be
your
moderator
for
the
session
today
he's
a
project
scientist
in
the
tertiary
science
section
and
the
science
liaison
of
the
landmark
working
group.
His
research
interests
are
urban,
modeling,
land
use
and
land
cover
change
and
the
hydro
hydrological
cycle
in
land,
surface
model
he's
also
the
supports.
C
F
All
right,
thanks
for
the
introduction,
I
will
say
that
I
only
understand
about
10
of
the
model,
but
I
know
people
who
understand
the
other
90.
So
I
think
hopefully
we'll
answer
all
the
questions
here.
So
I'd
like
to
quickly
introduce
each
of
our
panelists
dave.
Lawrence
dave
is
a
senior
scientist
and
the
head
of
the
terrestrial
sciences
section.
F
F
Her
research
interest
includes
crops
physiology,
in
particular,
photosynthesis
and
conductance
ozone
damage
and
vegetation
phenology
peter
lawrence
project
scientists
in
tss
as
well.
His
research
interests
include
land
use
and
land
cover
change,
with
a
focus
on
forest
agriculture
and
human
management
and,
last
but
not
least,
daniel
kennedy,
who
recently
joined
the
land
model,
land
modeling
group
as
a
postdoc
in
january.
Through
the
advanced
study
program.
F
F
F
Then
we'll
quickly
move
on
to
the
next
question.
I
doubt
we'll
get
through
all
the
questions
this
morning.
There's
I
think,
there's
like
22
of
them,
but
friday's
practical
session
includes
a
land,
practical
and
there'll
be
a
few
land
people
there,
including
myself,
and
so
perhaps
we
can
answer
any
questions
we
don't
get
to
then.
F
All
right,
okay,
so
the
first
question
is
in
the
biophysics
geophysics
lecture.
By
dave
he
mentioned
that
tempered
forest
was
strong.
Co2
absorption,
moderate
sun
absorption
and
moderate
evaporative
cooling
have
competing
effects
on
climate.
Could
you
elaborate
on
that
yeah.
G
Should
I
go
ahead
and
answer
that,
since
that's
me
yeah?
What
we
meant
by
this
was
that
you
know
in
an
effort
to
evaluate
the
different
impacts
of
forest
in
different
parts
of
the
world.
G
It
was
seen
in
the
high
latitudes
that
you
know
there's
the
snow,
albedo
feedback
that
very
clearly
dominates,
and
so
that
when
you
have
a
deforestation,
you
tend
to
get
cooling
and
in
the
tropics
you
tend
to
have
a
lot
of
evaporative
cooling
as
well,
so
in
the
mid-latitudes
there's
a
sort
of
competition
between
the
change
in
albedo
and
the
change
in
absorption
due
to
solar
radiation,
in
fact
also
the
change
due
to
roughness.
G
That
means
that
it's
and
the
change
in
the
in
the
co2
absorption,
which
is
which
is
relatively
strong.
That
means
it's
relatively
hard
to
establish.
You
know
without
definitively
whether
or
not
the
force
tend
to
be
warming
or
cooling.
Overall.
F
All
right
thanks
all
right,
so
next
question:
how
are
mountains,
craters,
canyons
and
very
topography
accounted
for
when
these
features
are
smaller
than
the
grid
box
resolution?
How
important
is
topography
in
terms
of
regional
climate
and
global
climate
yeah,
so
topography
certainly
is
important
in
determining
climate,
especially
atmospheric
circulation.
Examples
include
like
the
andes
and
greenland.
For
instance,
we
spent
a
lot
of
time
in
csm
trying
to
capture
effects
of
topography
on
greenland
precipitation,
for
example.
F
I
think
most
of
this
is
kind
of
accounted
for
by
the
atmosphere.
G
I
can
say
that
you
know
that
the
details
of
the
you
know
the
peaks
and
valleys
is
not
captured
by
the
large-scale
model.
G
G
Although
I
don't
think
any
that
are
operational
in
a
you
know,
like
a
cmip
class
model,
we
do
have
an
a
version
of
of
clm
called
the
clm
hill
slope
episode
of
hillside
model,
which
is
able
to
account
for
that
downscaling
and
it's
it's
available
essentially
upon
request.
If
anybody's
interested.
In
that,
I
don't.
I
don't
remember
if
I
talked
about
that
in
my
lecture
last
year
or
not,
but
it's
an
advanced
feature
of
the
model
which
is
which
is
coming
online.
H
Yeah
and
another
thing
that
that
would
capture
is
lateral
movement
of
subsurface
water.
You
know
which
isn't
captured
in
the
main
clm,
but
could
be
represented
better
with
the
hill
slope
hydrology
that
that's
coming
online
now.
F
All
right
thanks,
you
guys
so
next
question
in
the
upcoming
release
of
csm
2.2.
What
specific
changes
will
happen
for
clm
and
among
them,
what
will
be
answer
changing
so
there
will
be
some
there'll
be
answer
changes,
but
my
my
impression
is
that
they're
they're
not
very
big,
there's
no
major
major
changes.
I
don't
know
if
dave,
if
you
know
specific
changes
that
are
included
there.
G
F
Okay
yeah
the
next
question.
Actually
I
answered
that
in
a
I
had
a
discussion
with
a
couple
people
yesterday
about
the
introduction
paper.
C15
says
urban
late
landings
cannot
change
the
simulation.
Then
how
can
we
simulate
the
effects
of
urban
sprawl?
F
Both
the
extension
of
urban
area
and
change
of
urban
properties
did
the
cmip
historical
and
future
simulations
consider
urban
sprawl
effects.
So
the
short
answer
to
that
is
that
yeah
we
don't.
We
don't
account
for
changes
in
urban
area
urban
properties
in
a
dynamic
sense.
In
other
words,
we
don't
do
that
in
a
transient
simulation
like
we
do
for
like
natural
vegetation
and
crop
changes,
and
so
the
scene
of
historical
and
future
simulations
didn't
didn't.
F
Consider
changes
in
urban,
but
you
can
simulate
that
by
you
know,
conducting
like
time,
slice
experiments,
we
change
the
surface
data
set
and
increase
or
decrease
urban
area
or
change
urban
properties
and
compare
it
to
simulation.
You
know
would
say
present
day
urban
and
we
have
a
tool
now
to
call
urban
properties
tool
that
helps
with
that,
especially
with
changing
building
properties
and
such.
F
G
G
Real
quick
that
a
graduate
student
in
advance
at
the
university
of
brussels
has
just
implemented
variable
lake
area
in
time
that
is
currently
under
review
for
a
pull
request,
so
that
will
be
available
soon.
F
All
right
so
next
questions
what
are
some
of
the
most
uncertain
parameters
or
processes
in
clm
and
in
particular,
salem?
U,
what
are
some
improvements
that
are
planned,
slash
desired
in
future
releases
of
clm?
I
mean
every
everything's
uncertain
in
the
model.
All
the
parameters
have
some
uncertainty
range,
in
particular
the
the
urban
model.
Certainly,
the
building
properties,
particularly
the
thermal
properties,
are
quite
uncertain
what
the
roofs
and
walls
roads
are
made
out
of
and
and
how
that
translates
into
thermal
properties,
such
as
thermal
conductivity
and
heat
capacity,
etc.
F
And
so
one
one
project
overall,
we're
embarking
on
for
clm
is
a
parameter.
Uncertainty
ensemble,
where
we're
trying
to
pull
all
the
parameters
out
of
clm
and
assign,
and
some
assign
some
kind
of
uncertainty,
bounce
for
them
and
do
some
sensitivity,
studies
and
maybe
dave
you
want
to
expand
on
that.
A
little
bit.
H
Yeah
so
far,
we've
identified
about
200
parameters
that
we're
interested
in
pulling
out
and
perturbing.
H
So
we're
going
to
do
that,
one
at
a
time
to
check
which
ones
affect
the
model
and
then
we're
going
to
do
more
than
one
at
a
time.
Everything
at
one
time,
experiment
to
make
sure
that
we
capture
any
interactions
and
we're
particularly
interested
in
effects
on
the
long-term
carbon
and
hydrological
cycles.
So
how
much
uncertainty
in
our
long-term
carbon
projections
is
there
from
our
parameters?
H
So
it's
exciting
we're
putting
a
lot
of
computing
time
into
this
experiment,
and
you
know,
have
some
big
questions
we
want
to
answer,
but
hopefully
also
provide
a
community
resource
for
whichever
you
know,
subcomponent
of
the
model
you're
interested
in
you
can
look
at
the
output
from
when
we
adjust
these
different
parameters.
H
I
should
say:
there's
also
another
paper-
I
I
don't
know
dave
if
it's
still
in
review
or
if
it's
out
by
katie
dagin
that
did
a
first
pass
and
identified
eight
or
ten
parameters
that
have
very
large
effects
on
the
water
cycle.
So
you
should
look
out
for
that.
I
what
what
journal
said
in
dave.
G
I
forget
the
journal,
it
was
a
statistics
journal
and
it's
still
in
review.
Okay,.
F
All
right
thanks
daniel
yep,
so
the
practical
homework
we're
asked
to
turn
off
urban
parametrization.
What
does
it
mean
by
turning
off
urban
parametrization?
Specifically,
how
will
clm
and
clmu
work
with
urban
parametrizations
turned
off
so
in
that
specific
practical?
What
was
you
were
asked
to
do
was
to
simply
turn
off
the
heating
and
air
conditioning
in
the
urban
model.
There's
a
nameless
parameter
that
they
want
you
to
experiment
with.
So
it's
not
turning
off
the
entire
urban
parametrization.
F
F
So
you
know
it
depends
on
if
you're,
if
you're,
trying
to
add
in
a
new
land
unit
or
a
new
vegetation
type,
it's
not
certainly
not
straightforward
to
do
that.
People
in
the
past
have
done
this,
maybe
by
like
taking
over
an
existing,
say,
plant
functional
type,
that's
not
in
their
modeling
domain
and
then
adjusting
the
properties
that
are
contained
in
the
clm
parameter
file.
G
I
I
can
say
that
it's
hard
to
answer
this
question
without
knowing
more
about
it,
because
every
every
different,
every
different
land
surf,
adding
a
new
type.
It
can
mean
different
things
you
could.
You
can
go
from
one
extreme
they're,
trying
to
add
in
a
new
land
unit
which
is
quite
challenging
to
to
trying
to
introduce
a
new
plant
functional
type
which
is
keith
just
described.
So
unfortunately,
there
is
no
there's.
D
I
was
just
going
to
add
the
yeah
the
data
structures
of
a
lot
of
the
way
the
vegetation
works,
require
software
engineering
to
actually
add
new
plant,
functional
types,
those
that
exist
all
or
land
units.
So
I'm
just
going
to
back
up
dave
that
you
know
if
you
do
follow
keith's
idea
of
lights,
replacing
a
plant
functional
type
in
a
domain
that
that
can
be
quite
simple,
but
if
you
would
actually
want
to
add
something
to
the
data
structure
and
the
whole
way
that
clm
is
structured,
you
would
need
some
software
engineering
associated
with
it.
F
Yeah,
I
do
know
that
you
know
we
have
a
ton
of
crop
types,
for
instance
in
the
in
the
model,
most
of
which
aren't
aren't
used.
F
F
All
right
so
can
vegetation
parameters
be
explicitly
spatially
explicitly
defined
similarly
to
a
soil,
texture,
ancillary
file,
yeah,
so
soul,
texture
is,
is
spatially
explicit
on
as
defined
on
the
surface
data
set
vegetation
parameters
are
generally
defined
by
plant
functional
type,
with
the
exception
of
of
things
like
if
you're
running
in
satellite,
phenology
mode
leaf
area
index
and
and
canopy
top
and
bottom
heights
and
stem
area
index,
you
can
you
can
change
those,
and
I
could
change
those
in
sp
mode.
But
right
now
I
don't
think
most
vegetation
parameters
are
spatially
explicitly
defined.
F
H
The
other
thing
you
can
do
is
define
your
own
set
of
pfts
right
and
then
make
your
own
pft
map.
So
I
think
some
people
have
done
this
for
regional
simulations
where
they
want
more
specific
plant
parameterizations
and
those
tend
to
be.
You
have
to
incorporate
your
own
map,
so
I
think
paulie
bowati
right
did
this
in
california,
where
they
wanted
to
represent
a
bunch
of
specific
trees.
H
So
you
would
implement
your
new
pfts
and
give
a
map,
so
you
wouldn't
necessarily
be
using
that
kind
of
the
soil,
texture
type
of
implementation,
but
just
giving
us
the
new
pft
map
and
your
own
set
of
pfts,
and
that
would
be
kind
of
a
workaround
for
making
your
parameters
spatially
explicit.
D
Also
going
to
say
for
the
crops,
fertilizer
is
both
spatially,
explicit
and
also
temporally
explicit,
so
it
changes
through
time
and
also
a
particular
crop
has
a
different
amount
of
fertilizer,
depending
where
it
is
sort
of
globally.
So
there
are
examples
where
we
have
done
spatially
explicit
crop
level
parameters
well,
in
this
case,
there's
data
going
in,
but
you
could
also
do
the
same
sort
of
method
for
parameters.
If
you
wanted
to.
F
All
right
thanks,
I
am
trying
to
run
csm
on
a
finer
grid
and
have
created
a
surface
data
set
with
mixed
surf
data.
F
When
I
ran
comp
set,
I
2000
it
failed
with
error,
saying
need
glc
to
land
mapping
file.
How
can
I
create
such
a
file?
I
did
not
make
a
domain
file
as
it
needed
yeah
so
that
that's
a
grid,
that's
supported.
There
should
be
mapping
files.
F
It
may
just
be
that
the
model
can't
find
that
mapping
file.
You
do
need
a
domain
file
again
that
that
should
exist
for
something
specific.
F
I
suggest
you
post
on
the
on
the
discuss
csm
forum,
with
your
with
the
details
of
what
you're
trying
to
do,
and
I
can
help
you
on
that
forum.
There's
a
bunch
of
you
might
also
search
on
there's
a
bunch
of
instances
of
people
trying
to
do
similar
things.
F
G
Can
I
follow
up
a
little
bit
keith
and
just
say
that
we
have
a
lot
of
supported
resolutions,
but
we
do
recognize
that
users
need
to
make
their
own
and
this
make
sure
if
dat
tool
does
work
in
in
most
cases,
although
it
is
a
little
bit
cumbersome
to
use-
and
we
do
have
an
ongoing
project
to
try
to
streamline
this
whole
process
to
make
it
easier
for
users
to
create
their
own
domains,
especially
because
we
are
coupling
the
model
to
wharf
the
regional
model
and
and
every
wharf
user
tends
to
have
their
own
domain.
F
F
Yes,
you
can
do
that.
There's
a
building
temperatures
stream
file
in
the
land
in
the
land
name
list
that
controls
the
maximum
building
interior
air
temperature,
and
it's
spatially
explicit.
It's
split
up
to
33
regions
across
the
globe,
so
you
can
manipulate
that
file
to
you
know,
turn
down
or
turn
up
your
air
conditioning,
and
you
can
do
that
dynamically
because
it's
it's
a
yearly
file
with
a
yearly
temporal
resolution.
So
that's.
F
F
Do
we
have
options
for
turning
off
the
air
conditioning
or
heating
system
and
clmu?
After
all,
air
conditioning
is
still
unavailable
over
many
places
in
the
world?
Yes,
that
that
nameless
parameter.
You
can
turn
off
that
we
I
referred
to
earlier.
You
can
turn
off
the
air
conditioning
heating
system
and
you
can
also
control
the
maximum
minimum
interior
building
temperatures,
either
through
the
streams
file
or
what's
on
the
surface
data
set.
F
I
think
some
aspect
of
this
will
will
probably
come
in
when,
when
fates
comes
in,
I
believe
maybe
one
of
the
panelists
can
expand
on
that.
G
I
can,
unless
someone
else
wants
to
yeah,
so
the
classical
dgvm
which
existed
in
in
prior
versions
of
clm
is
still
in
the
model.
G
We
call
it
functionally
supported
in
that
we
believe
it
will
run,
but
we
have
not
evaluated
it.
Nor
do
we
intend
to
evaluate
it.
So
it's
at
a
user's
discretion
to
to
try
to
run
that,
and
our
expectation
is,
is
that
the
vegetation
distribution
would
probably
look
quite
poor
if
you
tried
to
run
it
right
now,
because
we
haven't
looked
at
it
in
many
years,
so
we
don't
intend
to
support
that
anymore
in
the
future.
G
So
what
we
are
doing
is
transitioning
to
the
fates
model,
functionally
assembled
terrestrial
ecosystem
simulator,
and
that
has
many
different
levels
of
dynamic
vegetation,
biogeography
of
varying
levels
of
complexity.
In
this
demographic
cohort
based
model-
and
so
you
know
that
requires
an
entire
lecture,
and
I
can't
remember
whether
that
was
covered
at
all
in
the
lecture.
But
we
do
intend
over
the
next
several
years
to
transition
to
this
spates
model
as
being
either
our
our
default
model
or
our
sort
of
equal
partner
model
with
our
with
our
current
existing
biogeochemistry
model.
G
So
if
you're
interested
in
in
fates,
you
know
please
get
in
contact
with
with
you
know,
one
of
us
in
the
group,
and
we
can
point
you
in
the
right
direction.
A
lot
of
users
are
starting
to
to
clm.
Users
are
starting
to
use
states,
but
it
remains.
F
F
I
mean
we
do
everything
in
a
kind
of
a
meter
per
meter,
squared
basis
to
calculate
fluxes
and
such
but.
H
Yeah
we
we
kind
of
leave
that
to
the
flex
tower
group
to
determine
their
fetch
and
scale
their.
You
know
eddy
covariance
measurements
accordingly,
and
then
we
would
take
their
per
meter
squared
observational
record
and
compare
it
to
our
per
meter
squared.
So
we,
the
point
mode,
doesn't
have
any
area
basis,
we're
just
doing
a
per
meter,
squared
and
and
using
the
core,
like
the
coordinating
observations
from
the
flux
tower.
E
I'll
just
add
to
that
that
you
can
change
what
plants
are
growing
there
and
I
think,
or
what
plant
functional
types
are
growing
in
the
veg
and
so
and
some
other
properties
of
the
ecosystem.
And
so
I
think
that
that
can
really
help
to
match.
The
fetch
of
the
flux
towers
is
by
adjusting
some
of
the
land
surface
properties
so
that
they're
similar
to
what
the
flux
tower
is
experiencing
and
then,
as
others
said,
that
it's
most
of
our
fluxes
are
calculated
on
a
per
meter
squared
basis.
F
Thanks
when
we
talk
about
the
effect
of
albedo
and
temperature,
what
kind
slash
scale
of
temperature
is
that
for,
for
example,
increase
in
albedo
will
cool
line
surface
or
two
meter
or
canopy
or
top
of
atmosphere
temperature?
This
might
be
linked
to
dave's,
slides
on
different
force
types
response
to
climate,
peter
you're
kind
of
our
expert
in
land
cover
change.
Do
you
wanna.
D
I'm
gonna
say
that
yes,
it'll
it'll
affect
all
of
those
those
temp
like
the
two
meter
temperature.
It's
about
the
surface
energy
change.
That's
also,
then,
going
to
have
an
influence
on
the
on
the
surface
fluxes
and
then
the
top
of
atmosphere
will
also
be
impacted
by
that
there'll
be
depending
whether
you're
looking
at
a
fully
coupled
system.
D
If
it's
fully
coupled
then
you'll
see
a
response
back
through
the
atmosphere
as
well,
so
top
atmosphere
is
going
to
be
impacted
by
the
fluxes
to
the
atmosphere
and
then
changes
in
cloud
cover
changes
in
circulation
patterns.
So
it's
more
complex
than
just
saying
that
the
albedo
did
this
it's
and
that's
why
we
do
have
to
do
ensembles
of
experiments
to
pull
out
a
a
signal
from
a
fully
coupled
experiment,
and
then
what
happens
is
that's.
D
The
response
of
the
atmosphere
is
very
dependent
on
the
vegetation
type
and
also
the
climate
zone
you're
in
so,
if
you're
in
a
tropical
area-
and
you
deforest
you're
not
going
to
see
the
albedo
being
the
major
signal.
What
you'll
see
is
the
changes
in
the
surface
hydrology
and
the
energy
fluxes,
whereas
if
you're
at
high
latitudes,
you
will
see
a
big
impact
because
of
snow
snow
plant
interactions.
D
D
One
of
the
things
is
we
see
across
different
climate
models,
different
responses
to
the
same
land
cover
change,
so
you
could
cut
down
forests
in
clm
and
you
might
see
a
very
different
response
than
if
you
would
in
the
hadley
center
model
or
if
you
were
in
gfdl's
model.
So
what
we
say
is
that
okay,
I'm
trying
to
have
an
explicit
relationship
between
this
tree
cuts
down,
and
then
you
have
this
climate
response.
It's
like
this
climate
response
evolves
in
this
model.
D
So
maybe
that's
a
bit
too
much
information,
but
I
think
that
sort
of
captures
the
nature
of
how
complex
it
is
to
try
and
pin
down
exactly
what
happens
when
a
certain
forest
type
gets
cut
down
in
a
certain
climate
place
dave.
Do
you
want
to
expand
on
that
or
you?
You
think
that
really
captures
that
yeah?
That
captures.
G
D
G
I
guess
in
just
in
the
you
know,
to
sort
of
introduce
new
capabilities
that
are
going
to
be
coming
online.
We've
a
graduate
student
at
university
of
washington
has
implemented
a
simple
land
model
called
slam,
marisa,
the
goo,
and
that
is
not
yet
available
within
within
csm
for
the
broad
user.
Although
you
could
contact
her
if
you
wanted
to
use
it,
but
we
are
this
far
going
to
be
implementing
that
and
the
nice
thing
about
this
slim
model
is
it
allows
you
to
control
very
directly
things
like
albedo.
G
F
All
right
thanks
that
was
good.
Good
answers.
Can
land
biogeochemistry
be
switched
off?
If
so,
how
would
this
be
done?
Yes,
you
can
switch
it
off.
F
You
know
basically
there's
two
modes
of
operation
with
clm
there's
we
call
sp
mode
or
satellite
phonology
mode
where
the
leaf
area
index
and
canopy
heights
are
prescribed
from
satellite
data
and
there's
no
no
biogeochemistry
going
on
or
the
bgc
mode
where
btc
is
is
on.
So
anybody
want
to
expand
on
that
at
all.
F
F
Oh
yeah,
I
mean
there's,
there's
comp
sets
available
for
sp
mode
and
for
bgc
mode.
You
know
kind
of
complementary
comp
sets.
You
can
select
comp
set
to
switch
off
by
g
chemistry.
F
I
don't
think
you
really
need
to
change
anything
any
anything
in
any
primarizations
in
particular
in
clm.
They
should
run
at
those
resolutions.
One
thing
you
have
to
be
careful
about
particularly
like
five
kilometer
resolution
is:
if
you're
making,
for
instance,
making
service
data
sets
using
the
make
surface
data
tool.
A
lot
of
the
input
data
sets
are
coarser
resolution
than
that
so
you'll
get.
F
F
The
other
thing
is
the
atmospheric.
Forcing
you
know
that's
provided
by
default
is,
I
guess,
at
a
half
a
degree
resolution.
So
if
you
want
also
fine
resolution
atmospheric
force,
you
need
to
have
to
provide
that
you
know
if
you
want
to
match
your
atmospheric,
forcing
to
your
land
grid.
D
D
To
it's,
we
do
actually
have
an
1850
and
a
2005
five
kilometer
resolution
raw
data
set.
That
is
part
of
the
the
data.
So
if
you
don't
want
transient
data,
we
do
have
two
snapshots
at
a
higher
resolution,
so
the
the
quad
degree
resolution
is
for
the
full
time
series
so.
F
Yeah,
so
the
there's
other
data
sets
such
as
soil
textures,
all
color,
that
are
at
coarser
resolution.
So
you
just
have
to
be
yeah
careful
about
that.
G
In
terms
of
forcing
we
also
have
on
hand
the
anal
deaths
forcing
over
the
north
america,
which
is
about
an
eighth
degree
resolution,
if
you're
doing
a
north
american
project,
yeah.
F
All
right
next
question:
in
sealand,
five
different
types
of
vegetation
are
saying:
sharing
one
soil
column.
Is
it
possible
to
assign
separate
soil
column
to
each
pft
by
changing
model
setup
or
modifying
the
csm
scripts?
Any
suggestions
on
doing
this
modification
yeah.
I
just
discussed
this
briefly
with
someone.
Yesterday
we
used
to
have
a.
F
We
called,
I
think,
a
no
compete
option
in
clm
where
each
pft
did
have
their
own
soil
column,
so
they
didn't
compete
for
water,
like
they
do
in
the
default,
where
they're
they're
all
sharing
the
soaks
same
soil
column,
but
I
wasn't
sure
if
that
was
still
operational
in
the
present
model,
and
I
feel
like
somebody
has
maybe
tried
to
do
this
before.
I
don't
know
if
any
of
the
panelists
have
any
more
information
on
this,
so
natalie.
D
G
Well,
it's
not
available
to
m5,
but
there
is
a
pull
request
with
this
feature
in
in
it
in
in
in
the
biomass
heat
storage
code.
So
it
should
be
available.
G
D
Yeah
and
there's
there's
a
jgr
paper
on
this
as
well.
So
if
you
look
at
the
schultz
where
she
actually
explicitly
looked
at
the
fluxes
with
them,
without
competition
were
within
them
without
shared
saw
columns.
So
this
has
been
documented
and
it's
available
in
the
literature.
If
we
want
to.
F
Look,
that's
great
yeah,
the
student
that
posed
this
question
emailed
me
about
it.
So
I
I'll
forward
her
question
on
to
you
guys
and
maybe
can
provide
some
additional
information
to
get
a
chance
to
her.
To
get
her
started.
Is
that
okay,
sure,
okay.
F
Be
his
spin-up
okay,
there's
like
a
couple
questions
here.
We
kind
of
discussed
this
just
a
bit
yesterday.
F
F
I
kind
of
my
answer
was
that
in
general,
if
you,
if
you're
interested
in
the
kind
of
the
transient
nature
of
of
of
your
changes
and
something
with
with
fire,
where
you're
changing
going
to
change
fire,
you
might
want
to
start
from
an
equilibrium
state,
and
so
you
might
want
to
do
a
spin
up,
but
do
any
panelists
have
a
another
opinion,
different
opinion.
D
I
would
just
say
that
fire
isn't
something
that
we
explicitly
prescribe.
We
prescribe
population
and
gdp
in
terms
of
like
how
fire
suppression
and
fire
ignition
work,
and
so
that
would
be
something
you
could
use
to
try
and
constrain
the
fire
behavior,
but
the
fire
model
itself
is
sort
of
dynamic
and
it
responds
to
to
deforestation
and
all
sorts
of
other
elements
like
climate
and
lighting
strikes.
D
So
it's
it's
one
of
those
times
where
it's.
It
would
be
more
complicated
than
just
saying.
Well,
let's
prescribe
1850
fire,
it's
not
something
you
actually
prescribe,
and
so
how
would
you
spin
it
up?
So
when
we
did
the
no,
we
we
actually
turned
off
the
fire
for
the
land
use
model
into
comparison
project.
We
did
some
land
on
the
simulations
and
we
just
went
from
the
same
spun
up
state
as
to
all
the
other
other
experiments
and
then
just
turned
off
the
fire
suppression,
the
fire
ignition
from
changing
human
population.
D
But
it
is
complex
because
there
is
both
the
natural
wildfire
side
of
it
and
then
there's
also
the
deforestation
side
of
it.
So
I'm
not
sure
exactly
how
you
would
turn
that
off
and
then
how
you
would
spin
it
up
to
a
different
state
if
you
would
just
and,
as
you
said
keith
like,
would
you
be
doing
it
from
like
a
background
like
natural
vegetation
state?
G
G
Especially
in
the
coupled
model
to
get
a
good
equilibrium
state
in
1850
right,
and
so
we
we
provide
files
that
give
a
good
equilibrium
climate
from
from
a
known
model
configuration
once
you
modify
that
yeah,
then
you
probably
do
need
some
new
spin-ups,
but
the
process
it
really
depends
on.
Unfortunately,
it
really
depends
on
the
question
you're
asking,
and
so
how
much
do
you
need
the
system
to
be
in
equilibrium
and
also
how
much
does
your
perturbation
affect
the
coupled
simulation?
G
But
we
do
see
users
frequently
get
tripped
up
by
this
and
there's
just
no
way
around
this.
This
challenging
problem
of
equilibrium
at
the
beginning
of
a
run
and
in
general
I
think
every
you
know
if
you're
trying
to
do
something,
that's
going
to
be
a
big
costly
experiment,
especially
if
it's
going
to
be
a
fully
coupled
experiment
in
a
lot
of
ways.
G
I
recommend
getting
in
contact
with
somebody
at
ncar
to
discuss
how
to
how
to
get
the
model
right
started,
because
because
it's
very
hard
to
answer
this
question
without
knowing
the
full
context.
F
Yeah
and
then
in
terms
of
spin
up,
I
I
suggested
that
you
know
I
mean
one
way
we
spin
up
is
we'll
use
the
atmospheric
forcing
from
the
fully
coupled
1850
simulation
and
force.
You
know
clm
with
that
and
spin
it
up
to
equilibrium.
So
that's
one
way
to
get
a
to
get
a
spin
up.
This
kind
of
consistent
with
a
couple
model
climate
to
start
off
a
historical
simulation.
D
Can
I
just
quickly
just
add
to
that
last
point:
there's
different
levels
of
spin
up
like
there's
spin
up
to
equilibrium
as
keith
saying,
which
is
like
you,
do
this
accelerate
decomposition
and
and
that's
sort
of
like
very
time
consuming.
Sometimes
you
can
just
go
from
a
spun
up
state
and
run
it
for
like,
say,
150
years
until
you
converge
to
a
carbon
balance
that
you,
you
think
is
within
tolerances,
so
it
doesn't
always
have
to
be
this
full
spin
up,
but
it's
a
case-by-case
basis.
D
F
All
right
getting
near
the
end
right
now,
I'm
sorry,
I
haven't
been
keeping
track
of
the
chat
or
the
raised
hands
things.
I've
kind
of
lost
that
ability
to
monitor
that,
but
I'll
check
that,
at
the
end
of
this,
after
we
finish
these
questions
and
see
what
what
else
we
have
in
cm5
do
this
different
soil
columns
interact
with
each
other,
for
example
horizontal
fluxes?
How
is
that
modeled.
F
They
don't
they
don't
do
that
right
now.
The
hill
slope
capability
that
I
think
dave
mentioned,
I
think,
does
try,
tries
to
model
that
I
don't
know
if
anyone
wants
to
expand
on
that.
G
Yeah
I'll
just
say
that
this
representative
hill
slope
capability,
where
we
mod
within
each
good,
so
we
model
one
to
as
many
hill
soaps
as
you
think
is
needed
to
capture
the
you
know.
The
sort
of
features
of
that
of
that
grid
cell
hill
slips
of
that
grid
cell,
including
slope
and
aspect,
does
you
know,
do
lateral
transfers
of
water
down
the
hill
slope
and
we
do
see
some
nice
features
of
that
where
you
get
riparian
zones
and
upland
dry
regions
and
things
like
that.
So
it's
a
new
feature
to
the
model.
G
H
And
sean
swenson
and
and
dave
published
a
paper
on
this,
so
you
can
go
find
that
to
check
out
some
of
the
details.
F
Next,
how
do
you
parameterize
fire?
I
suppose
the
model
computes
a
fire
index,
but
how
do
you
include
the
anthropogenic
components
since
most
fires
are
caused
by
humans?
F
D
And
I'll
just
repeat:
there's
also
deforestation,
fires
that
happen
as
you
cut
down
forests,
there's
a
fire
spread
that
comes
out
of
that
component
as
well,
and
then
there
are
agricultural
fires,
so
crop
residue
and
things
like
that.
G
And
fire
itself
is
dependent
on
fuel
load
in
the
surface
climate,
how
dry?
It
is
so
it's
fairly
fairly
complex,
again,
probably
reading
the
technical
description
of
the
papers
associated
with
it
to
get
a
full
picture.
F
E
Yeah
there
are
eight
prop
types
that
are
modeled
in
clm5
and
those
include
temperate
corn,
tropical
corn,
temperate
soy,
tropical
soy,
cotton,
sugar
cane
and
rice
there's,
one
other
one
that
I'm
forgetting
wheat.
Yes,
thank
you,
and
so
those
are.
Those
are
the
eight
crop
types
that
we
include
and
then
there's
irrigated
and
rain
fed
fractions
for
each.
E
There
are
62
crop,
functional
types
that
have
space
on
the
land,
surface
data
set
and
the
land
use
change,
data
set
and
there's
so
there's
an
irrigated
and
rain
fed
fraction
for
each
so
there's
actually
31
different
crop
types
that
have
space,
but
then
only
eight
of
those
have
active
parameterizations
that
we
use.
D
And
I'll
just
add
that,
like
the
crops
that
we
don't
have
parameterizations
for
that
are
mapped
get
remapped,
they
get
run
as
a
different
type
of
crop
that
we
do
actually
have
running
in
the
model.
So
if
you
have
something
that
is
mapped
out
as
barley,
it
gets
run
as
wheat
because
we
don't
have
a
barley
parameterization,
but
we
still
have
barley,
mapped
in
the
world.
F
All
right,
thank
you
next
question.
What
is
the
spin
up
time
for
the
land
so
in
satellite
phenology
mode
really
doesn't
take
that
long
model
years,
maybe
20
to
30
model
years
to
to
kind
of
get
soil,
temperature
and
soil,
moisture
and
equilibrium?
G
I'll
just
say
that
there's
we
have
a
wiki
page
on
this
on
the
github
site,
where
we
try
to
sort
of
track
our
current
understanding
of
the
spin-up
time
scales
and
there's
many
there's
really
many
spin
up
time
scales
and
it
really
really
again
depends
on
the
question
you're
asking
so,
for
example,
if
you're
interested
in
sort
of
climate
feedbacks
due
to
vegetation-
and
you
just
really
care
about
the
vegetation
state
coming
to
equilibrium-
that's
much
faster
than
the
solar
carbon,
and
so
you
can
get
away
with
a
much
shorter
time
scale.
A
G
Depending
on
the
depending
how
far
away
you
are
from
the
original
configuration,
so
again,
it
comes
down
to
this
question
of
what
you're
asking,
but
the
wiki
is
one
page
one
way
to
to
guess
and
and
just
in
general
what
I
tell
users.
G
A
lot
is
just
there's
very
little
to
do
other
than
just
do
it
and
then
run
these
scripts
and
see
when
it's
in
equilibrium,
because
it's
hard
to
anticipate
ahead
of
time
so
just
be
in
the
habit
of
of
doing
that
to
check
for
equilibrium
before
you
kick
off
your
full.
You
know
expensive
experiments.
D
Yeah
go
ahead,
I'm
just
going
to
add
that
it's
important
that
you're
in
equilibrium
with
the
climate
as
well
like.
So
if
you
spun
it
up
with
gswp3
climate
and
then
you
go
to
a
fully
coupled
model
run
it's
a
different
climate,
so
there'll
be
a
change
in
the
carbon
balance,
and
things
like
that.
So
it's
important
to
be
in
balance
with
the
the
model
that
you're
using.
F
You
know
I
was
just
going
to
add
that
there,
in
the
ceiling,
user's
guide,
there's
instructions
on
how
to
do
the
spin
up
and
there's
some
examples
of
how
the
spin-up
will
might
look
for
you
and
the
expected
times
also
so,
okay.
So
the
last
question
we
kind
of
already
addressed
we'll
get
back
to
you
on
that
to
help
you
with
that,
I'm
going
to
stop
sharing
my
screen.
E
Hi,
thank
you
yeah.
I
just
had
a
quick
follow-up
on
the
albedo
question.
I
was
wondering
if
you
could
briefly
discuss
what
albedo
scheme
is
used
for
lakes
and
how
it
transitions,
as
the
lake
moves
from
liquid
to
ice.
H
G
Just
okay:
what
is
it
for
lakes?
There
is
an
albedo
for
for
open
water,
so
it's
just
a
specified
albedo
of
the
surface
which
does
allow
for
some
penetration
of
radiation
into
the
water
and
then,
as
the
as
it
goes
into
in
the
frozen
situation,
they
would
calculate
the
fraction
of
the
ice.
That
is
a
fraction
of
the
lake
that
is
ice
covered
and
then
it
would
use
a
bare
ice
albedo,
which
is
also
specified,
and
then,
if
snow
falls
on
top
of
that
lake,
then
we
track
that.
G
And
then
you
know,
depending
on
how
much
snow
is
on
top
of
the
ice,
you
would
use
the
snow
albedo,
which
is
calculated,
I
believe
by
our
model
called
snicker.
So
hopefully
that
answers
the
question.
So
it's
it's.
It
is
using
all
the
the
physics
of
clm,
essentially
except
just
counting
for
lake
ice
and
then
snow
on
top
of
the
eye
on
top
of
that
ice,
and
it's
all
frozen.
Fractionally
weighted.
G
E
C
No,
I
mean
it's
fine,
so
let's,
let's
stop
it
here
and
give
everybody
a
five-minute
break
before
we
enter
the
media
sciences
session
and
please
come
back
at
10.
If
you
can
it'll
be
great
and
then
you'll
need
the
best
that
10
will
send
you
off
with
your
scientists
in
all
the
different
breakout
rooms,
and
thank
you
so
much
for
this
session.
F
Yeah
thanks
thanks
to
the
panelists,
I
really
appreciate
you
being
here
helping
answer
the
questions
and
thanks
to
everyone,
the
students
participants
listening.
I
appreciate
it.