►
From YouTube: 5th PAWS Webinar
Description
The fifth webinar from the Paleoclimate Advances Webinar Series (PAWS) which took place on September 1st 2022.
Dr. Georgy Falster discussed "New reconstruction elucidates intrinsic & forced behavior of the Pacific Walker Circulation over the past 800 years" and Dr. Ran Feng discussed "Late Pliocene as an analogue for studying sensitivity of Pacific Walker Circulation to CO2 forcing"
For more information and to signup for the PAWS Google Group visit:
https://www.cesm.ucar.edu/events/webinars/paws/
A
A
Okay,
great
so
by
joining
the
webinar,
you
agree
to
follow
our
code
of
conduct
during
the
webinar.
Please
consider
new
ideas,
encourage
innovation,
offer
constructive
feedback,
acknowledge
teamwork,
show
appreciation
and
share
the
air.
The
webinar
has
two
talks.
Each
tag
has
20
minutes
with
additional
5
minutes
for
qna.
A
A
A
A
It
is
my
great
pleasure
to
introduce
today's
two
speakers:
doctors,
georgie
foster
and
ryan
phone
georgie
foster
is
an
arc
center
for
excellence
for
climate
extreme
postdoctoral,
fellow
at
the
australian
national
university
georgie
combines
information
from
observation
model
simulation
and
the
paleoclimate
proxy
archives
to
better
understand
water
cycle.
Variability
on
spatial
skills,
from
regional
to
global
and
time
scales
from
inter-annual
to
centennial.
B
All
right,
we're
good!
You
can
see
my
screen.
Yes,
all
right!
Well,
thank
you.
So
much
for
having
me
at
this
walker
themed,
pause,
webinar,
I'm
really
looking
forward
to
hearing
rand
talking
about
the
walk
of
circulation
in
the
pliocene,
but
I
will
be
starting
things
off,
though,
talking
about
what
the
walker
circulation's
been
up
to
for
the
past
800
years
by
means
of
a
new
multiproxy
maori
method,
reconstruction
that
we're
writing
up
for
publication,
hopefully
in
the
near
future
and
I'll
focus
mainly
on
two
points.
B
B
So
first
things.
First,
what
is
the
walker
circulation?
Well,
it's
the
zonal
component
of
atmospheric
overturning
circulation
over
the
equatorial
pacific,
where
air
rises
over
the
indo-pacific,
whirlpool,
cools
and
sinks
over
the
eastern
equatorial
pacific
with
the
return
flow
forming
the
pacific
trade
winds
and
we
can
quantify
the
strength
of
the
water
circulation
using
the
sea
level.
Pressure
gradient
across
the
tropical
pacific,
so
it's
clamoring
as
you
call
low
pressure
over
the
warm
pool
and
high
pressure
in
the
east.
B
B
For
example,
looking
at
the
diagram
on
the
left
here
we
have
decreased
strength
of
convection
over
the
warm
pool,
increased
convection
and
lower
surface
pressure.
Further
east
weaker
trade
winds-
and
this
is
a
weaker
walker
circulation
or
a
negative
anomaly
in
that
delta.
In
that
sea
level.
Pressure
gradient-
and
this
is
the
breakdown
in
circulation
that
we
get
during
el
nino
events,
then
on
the
right.
This
shows
the
flip
side.
B
So
a
stronger
walker
circulation
like
we
see
during
london
events
and
we're
all
getting
increasingly
familiar
with
those
over
the
last
few
years,
and
these
changes
in
the
strength
of
the
walker
circulation
affect
weather
and
climate
even
far
out
beyond
the
tropical
pacific.
B
Stronger
or
weaker,
with
future
anthropogenic
changes
model
simulations
generally,
although
not
uniformly,
suggest
that
the
net
effect
of
anthropogenic
forcing
should
be
a
weaker
walk
of
circulation.
B
But
we
haven't
really
seen
this
in
our
observations
of
the
walker
circulation
over
the
past
100
years
or
so.
I
have
a
feeling.
Rand
might
be
having
something
to
say
on
that
topic,
but
also
this
is
in
part
due
to
the
walker
circulation's
large
internal
variability,
much
like
enso,
which
makes
it
difficult
to
to
discern
trends
or
changes
within
the
relatively
short
observational
era.
B
Oh
yeah,
I
will
mention
at
this
juncture
that
there's
actually
a
reasonable
consensus
that
the
walker
circulation
has
strengthened
over
the
past
couple
of
decades,
since
the
90s
or
so,
although
we
can't
really
say
whether
or
not
this
is
just
within
the
range
of
natural
decade
or
variability
of
the
work
of
circulation
or.
A
B
So
to
contextualize
all
of
this,
what
we
need,
of
course,
is
a
longer
record
of
the
walker
circulations.
You
know
long
enough
to
contextualize
modern
variability,
and
for
that
we
need
some
variable
with
a
strong
mechanistic
relationship
with
the
worker
circulation,
with
data
that
go
back
some
hundreds
of
years
for
which
we
turn
to
water
isotopes,
which
is
to
say
the
stable
hydrogen
and
oxygen
isotopic
composition
of
precipitation
and
other
meteoric
waters.
B
Now
some
recent
work
from
our
group
demonstrated
in
the
paper
linked
here
that
the
walker
circulation
actually
has
a
strong
imprint
in
global
precipitation
delta
o18
variability,
which
means
that
we
can
leverage
that
relationship
to
reconstruct
the
walker
circulation
back
through
time,
with
some
assistance
from
proxy
records
for
that
variable,
which
is
to
say
that
the
stable
isotopic
composition
of
precipitation
and
other
meteoric
waters
in
the
past.
B
B
B
B
B
All
right
so
here
it
is.
This
is
the
full
reconstruction
ensemble
from
1200
to
2000,
where
the
outermost
ribbon
is
the
90
confidence
interval,
and
so
this
ensemble
or
the
ensemble
members
incorporates
uncertainty
from
all
those
things
I
just
skimmed
through
as
well
as
a
couple
of
others.
I
didn't
mention
all
right
so
now
we
can
use
this
reconstruction
to
look
into
the
response
of
the
walker
circulation
to
the
two
biggest
four
things
of
of
this
period.
So
that's
combined
anthropogenic
forcings
and
volcanic
eruptions.
B
B
So
what
I've
plotted
here
are
two
distributions.
The
in
the
blue
are
all
values
from
the
reconstruction
ensemble
that
fall
before
the
year
1850
and
then
in
pink
are
all
the
values
from
1850
to
2000..
B
However,
the
the
difference
between
these
two
distributions,
it's
no
bigger
than
differences
between
any
other
two
equivalently
long
time
periods.
B
We
can
also
do
things
like
look
at
just
the
most
recent
values,
so
this
is
the
distribution
of
the
full
ensemble
now
and
then
here
I've
plotted
the
values
just
for
the
most
recent
decade
of
the
reconstruction.
So
that's
some
the
90s,
so
you
can
see
that
in
the
context
of
the
full
800
years,
the
90s
were
a
little
weaker
than
normal.
That
might
have
something
to
do
with
the
pinot
tuba
eruption,
which
I'll
get
to
in
a
moment,
but
still
well
within
the
range
of
variability.
B
Actually,
I'd
argue
that
probably
this
lack
of
a
significant
post-industrial
mean
change
is
down
to
at
least
two
things.
I
guess
I
mean
first,
the
large
natural
intrinsic
variability,
which
you
can
see
quite
clearly
here
in
the
reconstruction,
which
means
that
it
takes
quite
a
large
mean
state
change
to
show
up,
and
secondly,
perhaps
more
importantly,
is
that
recent
research
from
a
few
different
groups
around
the
world
suggest
that
various
different
anthropogenic
climate
changes
actually
push
the
walker
circulation
in
different
directions.
B
For
example,
I
might
talk
about
this
a
bit,
but
the
thermodynamic
effect
of
global
warming
is
is
to
drive
a
weaker
walk
of
circulation,
but
the
associated
amounts
slow
down
and
also
anthropogenic
aerosol
emissions.
Both
drive
a
stronger
walker
circulation.
B
So
maybe
at
the
moment
we're
just
in
a
sort
of
a
more
transient
state
where
the
various
influences
are
cancelling
out.
This
is,
of
course,
a
question
for
climate
models.
My
group's
working
on
this
at
the
moment,
that's
largely
being
led
by
kelly
flager
who's,
a
grad
student
with
sam
stevenson
at
ucsb,
looking
at
the
walker
circulation
in
the
cesm
last
millennium,
ensemble.
B
Right
so,
just
before
I
move
on
to
volcanoes,
I'm
going
to
muddy
the
waters
slightly
remember.
I
said
in
the
intro
that
there's
a
reasonable
consensus
in
the
literature
that
the
walker
circulation
has
strengthened
over
the
past
couple
of
decades.
So
that's
a
trend
to
more
la
nina,
like
conditions
and
one
of
the
first
papers
to
mention
this
look
at
the
trend
from
1992
to
2011..
B
So
that's
a
trend
over
20
years.
So
what
I've
done
here
is
to
calculate
and
plot
the
distribution
of
all
possible
20-year
trends
in
the
full
reconstruction
ensemble.
B
So
there
are
three
histograms
here,
because
I've
split
the
ensemble
by
the
gridded
sealable
pressure
product
that
was
used
to
train
the
reconstruction.
So
the
era
20th
century
reanalysis
had
slp
or
ico
ads,
and
then
I've
highlighted
the
high
and
low
tails
now
the
reconstructions.
Then
they
go
to
the
year
2000.
But
of
course
observational
data
continues.
So
we
can
calculate
the
magnitude
of
that
recent
trend
over
the
20
years
from
92
to
2011.
B
and
show
that
on
these
distributions
in
the
red
bars
here
and
as
you
can
see,
that
recent
strengthening
trend
is
actually
quite
unusual.
B
So
in
one
case
it
is
in
the
upper
tail
when
we
calculate
that
trans-specific
sea
level,
pressure
gradient
using
heterozyg
and
in
the
other
two
cases
where
I've
calculated
the
gradient
using
ico,
coeds
or
the
era
20th
century
analysis,
it's
on
the
high
end
of
the
distribution,
but
not
quite
in
the
tail.
B
B
B
So
to
test
this,
I've
used
an
approach
called
superposed
epoch
analysis
where
we
take
one
reconstruction,
ensemble
member
and
also
a
reconstruction
of
volcanic
eruptions
over
this
period,
which
I've
shown
in
the
red
lines.
Here
then,
from
the
reconstruction
we
snip
out
10
year
segments
covering
each
eruption
so
with
a
few
years
prior
and
six
years
following
and
then
composite
all
these
segments
up
into
a
single
10-year
time
series
so
take
the
mean
and
finally,
test
weather
variability
after
eruptions
is
significant
compared
with
a
null
hypothesis
design.
B
Specifically
for
this
analysis,
those
dotted
lines
and
the
point
of
this
being
that
if
there
is
any
consistent
walker
circulation
response
to
volcanic
forcing
that
will
be
concentrated
by
this
analysis.
B
Now,
in
this
case,
so
with
this
particular
combination
of
eruptions
and
this
particular
ensemble
member,
you
can
see
that
there
is
not
quite
a
consistent
walker
response
to
volcanic
eruptions.
We
haven't
exceeded
our
confidence
intervals,
so
in
this
case
the
variability
is
to
just
cancel
it
out
to
give
us
a
relatively
flat
line.
B
But
of
course
I've
done
this
for
all
5,
000
or
so
ensemble
members
and
also
subsetting
the
eruption
database
dataset
in
various
ways.
We've
wound
up
with
a
couple
of
really
key
findings.
B
What
proportion
of
ensemble
members
give
a
significant
response
in
that
year
and
whether
it's
a
positive
or
a
negative
anomaly,
so
you
can
see
that
only
a
very
small
proportion
of
ensemble
members
give
a
significant
response
about
10
in
a
composite
made
up
of
all
25
of
these
eruptions,
and
now
we
can
step
through
this
list
of
eruption
years
progressively
removing
the
weaker
eruptions
until
we're
running
the
analysis,
with
only
the
six
largest
volcanic
eruptions
of
the
past
800
years.
B
B
B
B
I
guess
to
the
point
of
erasure
by
chronological
uncertainty
in
the
reconstruction
as
well
as
some
other
things
that
I
I
won't
get
into
unless
anybody
asks
so
when
we
do
mostly
avoid
the
the
muddying
influence
of
chronological
uncertainty
by
only
looking
at
relatively
recent
eruptions,
you
can
see
that
there
is
indeed
a
clear
el
nino-like
weakening
of
the
walker
circulation
in
the
years
following
a
volcanic
eruption.
B
B
So
what
I'm
showing
oh
good,
I'm
almost
done
actually
yeah.
So
what
I'm
showing
here
are
the
results
of
that
same
super
post,
epoch
analysis
performed
on
all
fully
forced
members
of
the
cesm
last
millennium
ensemble
in
this
case,
using
dates
from
the
volcanic
forcing
fire
which
is
galwaneau
2008.
B
So
here
I've
plotted
the
composite
response
for
the
19
strongest
eruptions
from
that
volcanic,
forcing
data
set
going
through
to
only
the
two
strongest
eruptions,
and
you
can
see
that
it's
broadly
the
same
result
as
in
our
reconstructions,
with
an
el
nino-like
weakening
of
the
walk
of
circulation
in
the
years
following
a
volcanic
eruption
in
pretty
much
all
ensemble
members.
B
B
D
Thanks
so
much
georgie,
so
we
have
a
few
minutes.
We
can
spend
on
questions
for
georgie
and
then
we'll
move
on
to
the
next
talk
and
after
that
we'll
have
a
larger
discussion
following
both
talks.
So
if
you
have
a
question,
feel
free
to
you
can
either
type,
I
have
a
question
in
the
chat
and
then
unmute
and
ask
it
or
you
can
raise
your
hand
or
gesticulate
in
some
other
way,
so
that
we
know
you
have
a
question.
A
B
Period
actually,
the
relationship
between
the
walker
circulation
and
global
temperature
is
not
particularly
strong
right.
So
it's
negative,
as
you
would
expect.
The
correlation
in
the
annual
mean,
is
something
like
minus
0.3
or
thereabouts.
B
So
I
I
did
use
the
I
used.
There's
a
reconstruction
of
global
mean
surface
temperature
across
the
common
era,
so
I
compared
that
with
my
reconstruction
and
similarly
the
correlation
is
negative,
but
very
weak.
The
the
mean
correlation
is
around
minus
0.1,
and
this
doesn't,
I
think,
it's
a
chance.
B
I
should
put
them
in,
but
yeah
and
they'll
say
it
doesn't
change
that
much
on
on
different
on
shorter
time
intervals,
either,
which
is
maybe
surprising
and
something
that
we
do
plan
to
look
into
more
using
model
simulations
yeah,
so
a
weak
negative
correlation
as
you're.
E
I
have
a
quick
follow-up
question.
Sorry,
I
can't
figure
out
how
to
raise
hands
georgie.
I
know
this
in
this
time
series.
Actually,
this
is
a
perfect
place
for
the
question.
There
is
an
increase
in
variability
later
on.
In
the
time
series
I
was
just
wondering.
Is
it
because
of
the
more
data
getting
incorporated
in
reconstructing
the
delta
slp
or
is
it
it's?
I
guess
the
question
is:
is
there
a
reason
for
that.
B
Yeah,
so
you
mean
this:
this
increase
in
the
the
confidence
intervals
right
at
the
top
like
from
the
1970s
or
so
is
that,
where
you're.
E
Sorry
I
actually
make
the
increasing
variability.
The
ups
and
downs
seems
like
there's
some.
The
frequency
is
getting
higher
like
starting
from
the
1600
compared
to
the
previous
time.
Almost
look
like
the
the
ups
and
downs
the
cycles
are
getting
shorter
or
it's
getting
more
variable
standard,
yeah.
B
I
mean
yeah
so
part,
yes,
part.
I
mean
part
of
that.
This
particular
graphic
is
slightly
misleading
in
that
sense,
because
the
variability
in
the
median
does
get
larger,
but
that's
just
because
the
confidence
intervals
are
actually
narrower.
So
you
know
there's
less
smearing
out
of
the
of
the
central
tendency
of
the
reconstruction.
B
As
far
as
changes
in
the
actual
spectral
densities
go
there.
There
is
well
actually
so
there's
there's
really
not
a
lot
of
change
or
barely
any.
If
you
know
splitting
the
reconstruction
in
into
various
parts,
the
only
thing
that
really
stands
out,
interestingly,
in
terms
of
sort
of
frequencies,
is
that
actually
there
seems
to
be
a
slight
shift
in
spectral
power
in
the
post-industrial
to
slightly
lower
frequencies.
B
So
you
know
you
can
see
the
the
the
power
spectral
density
is
very
similar
to
answer.
In
the
most
spectral,
power
is
concentrated
in
the
sort
of
two
to
nine
year
band
when
we
restrict
it
to
just
the
post-industrial
we
sort
of
lose
a
bit
of
power
in
the
two
to
three
year
bands.
It
goes
more
to
the
sort
of
four
to
ten
but
yeah.
I
would
say
that
that
what
looks
like
larger
amplitude
variability
in
the
median
in
the
reconstruction
is
is
really
just
due
to
decreased
uncertainty.
In
that
case,
yeah.
E
F
E
C
C
E
E
Okay
still
see
the
notes,
yep.
E
There
you
go:
okay,
apologies
for
that!
All
right,
hello!
Everybody!
Thanks
for
that
wonderful
introduction
about
the
walker
circulation
by
georgie.
Today,
I
would
like
to
talk
about
the
possibility
to
use
payload
climate
records,
especially
past
pleistocene,
the
late
specifically
the
late
pricing
to
test
the
sensitivity
of
pacific
walker
circulation
to
co2
forcing
and
this
work
is
done
by
my
master's
student,
teddy
mayer
and
in
collaboration
with
tripty
biotrans
group,
as
their
qs
applies
in
climate
variability
group,
as
well
as
on
card
group.
E
E
E
This
prediction
has
been
corroborated
by
many
generations
of
atmosphere,
ocean
couple
models
to
the
bottom
right
here,
focusing
on
the
grain
markers
those
are
the
predicted
late,
20th
and
and
the
25
early
21st
century
trends
of
the
walker
circulation
strengths,
measured
by
two
different
metrics,
yielding
simulations
contributed
to
on
the
climate
model
in
the
comparison
project,
five.
So
in
the
top
left
quadrant
that
shows
regardless
which
metric
you
use.
There
are
many
many
models
that
shows
the
weakening
of
the
walker
circulation.
E
It
has
been
really
difficult,
as
mentioned
by
georgie,
to
test
this
prediction,
largely
because
of
the
internal
variability,
which
is
very
strong
recorded
in
our
instrumental
records,
as
well
as
in
the
records
that
georgie
has
shown
before.
This
is
the
same
figure
by
tron
at
all.
But
this
time
I
want
you
to
focus
on
the
yellow
marker
here.
So
those
are
the
late
20th
century
trends
of
walker
circulation,
strengths,
simulated
by
community
earth
system
model,
but
initialized
with
different
initial
conditions.
E
It's
apparent
that,
depending
on
how
the
model
is
initialized
at
multi-decadal
time
scale,
there
might
be
strengthening
awakening
of
the
water
circulation
that
is
enforced
by
co2.
However,
our
instrumental
record
only
lasts
for
past
couple
decades.
So
the
argument
is
that
maybe
they're
observing
the
internal
variability
instead
of
forced
trends.
E
In
addition,
cooling
due
to
20th
century
aerosol,
forcing
in
the
northern
hemisphere,
perhaps
also
a
change
to
the
strength
of
the
walker
circulation.
So
in
this
chart
here
across
the
render
models.
Northern
hemisphere
cooling,
due
to
aerosol
forcing
showing
those
green
tiny
grain
bars
here,
tends
to
strengthen
the
water
circulation,
which
compensates
the
force.
The
weakening
of
the
water
circulation
by
co2
increase,
showing
the
red
bars
on
the
chart.
E
E
You
can
see
that
there
is
a
sign
switch
in
the
direction
of
feedback
in
the
eastern
side
of
the
equatorial
pacific,
from
negative
feedback
which
was
restraining
the
surface
warming
to
a
positive
feedback
which
actually
would
enlarge
the
surface
warming,
and
this
result
actually
indicates.
Perhaps
our
instrumental
records
are
not
long
enough
to
see
the
weakening
of
the
walker
circulation,
which
may
become
more
obvious
as
the
response
time
increases.
E
Here
we're
looking
at
the
possibility
of
using
late
pricing
proxy
records
to
test
this
prediction
of
weakening
water
circulation
in
response
to
writing
co2.
The
reason
for
using
leg
pricing
are
two
folds.
First,
this
time
interval
had
relatively
small
changes
in
boundary
conditions
compared
to
many
other
past
warm
intervals.
E
It
also
has
well
constrained
co2
estimates,
as
shown
in
this
figure
here
estimated
mid
pricing.
Co2
is
around
400
ppm
following
the
mid
pricing
towards
the
ply
sitting,
as
shown
in
the
top
panel
here,
there
is
a
long-term
decline
of
co2,
accompanied
by
global
cooling,
which
is
shown
in
the
bensick
delta
180
record
in
the
bottom
panel.
E
E
Second,
what
are
the
meaningful
proxy
signals
that
can
be
used
to
test
the
water
circulation
sensitivity
to
co2
changes?
It's
frequently
argued
that
we
can
just
use.
Look
at
eastern
east
west
sst
contrast
across
secretary
pacific
to
to
say
whether
circulation
might
be
stronger
might
be
weaker.
However,
the
east-west
sst
gradient
is
affected
by
many
other
aspects
of
ocean
dynamical
adjustments
right,
such
as
changes
in
upwelling
strengths,
changes
in
subtropical
tropical
overturning
circulation.
E
So
in
this
study,
we're
particularly
interested
in
finding
proxy
signals
that
are
more
direct
leading
to
worker
circulation,
strengths
itself,.
E
I
think
georgie
published
on
this
to
answer
those
two
questions,
so
we
also
used
to
the
what
I
still
tracking
enable
the
community.
Our
system
model
version
1.2.
E
This
model
tracks
the
transport
of
water,
isotopes
in
energy
ocean
sea
ice
and
tracks,
the
mass
dependent
fractionation
during
phase
change
of
water.
As
a
result,
it
can
stimulate
isotopic
imprints
of
hydrological
cycle
as
well
as
circulation
changes
in
amsterdam
ocean.
We
carried
out
two
sets
of
simulations
using
this
model
framework.
E
E
In
addition,
we
also
used
a
pricing
a
pre-industrial
simulation
which
has
been
published
always
the
same
model
previously.
E
So
in
those
simulations
we
essentially
prescribed
ocean
dynamics,
so
ocean
is
reduced
to
a
thermodynamic
component
with
mixed
layer,
ocean
heat
flux
in
equilibrium,
with
changes
in
the
atmosphere,
the
simulations
featured
pre-industrial,
400
ppm,
two
times
pre-industrial
and
three
times
pre-industrial
levels
of
co2.
So
this
really
allows
us
to
test
the
sensitivity
of
circulation
changes,
isotopic,
imprints
to
a
range
of
co2
levels.
E
Those
simulations
are
run
with
two
set
of
boundary
conditions.
One
is
pre-industrial,
the
other
one
is
placing
as
well
as
heat
fluxes
generated
by
corresponding
couple
models.
E
In
those
simulations
we
viewed
the
water
tagging
function,
which
is
showing
this
little
diagram
here
we
tag,
let's
see
if
we
have
this
place,
called
shire.
If
you're
a
lot
of
lord
of
the
rings
fan
and
tonight
the
tv
series,
a
lot
of
rain
is
gonna,
be
on
amazon,
prime,
but
anyway,
so
in
this
place
called
shire.
E
For
example,
we
tag
the
evaporation
of
water
vapor
in
the
model
and
at
any
other
location
with
ink
or
outside
shire,
when
precipitation
forms,
the
model
can
tell
us
exactly
how
much
precipitation
is
from
shire
is
from
the
vapor
that
evaporated
from
shire
and
what's
the
deuterum
or
delta
18-hour
concentration
of
precipitation
that
are
sourced
from
moisture
from
shire
right.
I
will
show
you
a
second
why
this
is
useful.
E
Right
without
simulations,
I
will
first
answer
the
question
of
do
boundary
conditions
during
the
mid
pricing,
alter
the
strengths
and
the
sensitivity
of
water
circulation
to
co2
changes.
E
We
use
this
zoono
mass
overturning
to
quantify
the
strength
of
water
circulation
in
order
to
avoid
potential
shift
in
circulation
center.
The
bottom
diagram
shows
the
calculated
zoonomat
stream
function
with
positive
values,
showing
the
clockwise
circulations.
As
you
can
see,
the
stream
function
depicts
a
circulation
cell
with
riding
motion
in
the
western
side
of
decreator,
pacific
and
descending
motion
in
the
eastern
side
of
the
equatorial
pacific.
Those
vertical
motions
are
connected
with
horizontal
weights.
E
E
Similar
results
applies
to
the
spatial
pattern
of
precipitation.
Deuterium
isotope
responds
to
co2.
This
map
here
shows
the
response
pattern
of
precipitation
delta
d
to
co2
increase,
derived
with
singularity
decomposition
using
the
mixed
layer,
ocean
experiments
with
the
range
of
co2
values
so
to
the
first
order.
Many
areas
across
the
land
and
ocean
feature
isotopic
enragement
as
a
result
of
more
enraged
vapor
due
to
surface
warming,
and
this
is
very
classical
temperature
effect
in
precipitation
isotopes.
E
We
built
the
water
tagging
functionality
to
understand
this
contrasting
precipitation
delta
d
changes,
so
in
response
to
increased
co2
from
one
times
to
two
times
turns
out
the
isotopic
signatures
of
precipitation,
either
sourced
locally
as
a
source
remotely,
which
is
in
red
in
this
diagram
or
locally,
which
is
in
blue
in
this
diagram
in
western
equatorial
pacific
or
eastern
equatorial
pacific,
had
a
similar
increase
of
precipitation
delta
d,
regardless
of
source.
So
this
could
not
be
the
reason
for
the
contrasting
delta
d
changes
between
west
and
the
east.
E
However,
we
noticed
a
much
stronger
increase
in
remote
moisture
source
in
the
eastern
equatorial
pacific
at
12
percent
compared
to
that
at
the
western
side
of
equatorial
pacific,
which
is
only
three
percent
so
due
to
the
fact
that
remote
moisture
source
is
much
more
depleted
in
deuterium
isotope
compared
to
the
local
source.
We
think
this
relative
increase
in
remote,
moisture
convergence
in
the
eastern
side
of
equatorial
pacific
explains
the
muted
or
even
negative,
precipitation
precipitation
delta
d
changes
in
this
region
in
response
to
co2
increase.
E
The
enhanced
moisture
convergence
in
the
eastern
side
of
the
equatorial
pacific
is
a
signature,
pacific,
walker,
circulation
weakening,
and
indeed
there
is
a
linear,
clear,
linear
dependency
of
east-west
to
durham
precipitation
to
term.
I
studied
contrast
and
worker
circulation
strength
across
the
range
of
co2
levels
so
to
the
bottom
right
to
the
left.
E
That's
the
annual
delta
d
climatology
simulated
by
the
model
and
to
the
right
I'm
showing
the
east-west
delta
d
contrast
as
a
function
of
mass
overturning
strengths,
and
you
can
see
that
there
is
a
linear
relationship
with
mass
overturning
getting
weaker
and
weaker
in
response
to
increasing
co2.
E
So
this
is
testable
as
far
as
we
have
records
and
thanks
to
our
proxy
collaborators,
the
tripty
butter
trial
and
david
storage
are
generating
records
generating
leaf,
wax
precipitation
delta
d
records
from
a
western
and
eastern
side
of
equatorial
pacific
site.
So
hopefully
nothing
not
too
far
in
the
future.
We
can
actually
start
testing
this
model.
Behavior.
E
So,
besides
precipitation
isotopes,
there
are
many
responses
in
the
climate
system
that
corresponds
to
the
weakening
of
the
water
circulation.
One
of
the
most
obvious
ones
is
probably
weakened.
Equatorial
wing
stress
okay,
so
the
top
panel
here
in
this
figure
shows
the
climatological
zoono
wing
stress
in
the
tropical
pacific.
E
The
bottom
shows
the
difference
of
mid-pricing
simulations
between
400
ppm
and
the
pre-industrial
level
of
co2.
So,
despite
there
is
a
enhanced
moisture
convergence
associated
with
enhance
the
easterly
wind
stress
off
the
equator
across
the
equator,
you
do
see
a
reduction
or
weakened
zonal
wing
stress.
E
In
response
to
the
weekend,
equatorial
easily
wind
stress,
equatorial
current
also
can
highlight
it
in
the
red
box
here,
comparing
the
top
panel,
which
is
climatology
to
the
bottom
panel,
which
is
the
difference.
E
And
an
interesting
feature
coming
out
of
the
weakened
current
is
perhaps
it
it
actually
ordered
the
seawater
delta,
18o
contrast
between
east
and
western
side
of
the
equatorial
pacific.
E
The
colors
in
the
top
panel
shows
the
stimulated
sea
water
sea,
surface
water
delta
contrast,
and
you
can
see
to
the
west,
it's
generally
negative
and
to
the
east,
to
the
central
and
in
the
central,
pacific
and
eastern
pacific
is
generally
positive,
and
this
pattern
is
further
enhanced.
E
E
The
correlation
map
featuring
contrasting
correlation
across
the
equator,
and
this
suggests
that
when
xenon
current
is
weaker,
remember
that
zonal
current
the
mean
state
of
zonal
current
is
negative
because
it
goes
towards
the
west.
E
Yeah,
I'm
almost
done
so.
This
suggests
that
the
seawater
delta
ato
reconstructed
from
paleo
oceanography
record
could
provide
a
proxy
for
wing-driven
equatorial
current
strengths,
and
we
did
look
into
this
thanks
to
heather
ford
at
university
of
london
who
compiled
the
planktonic,
carbonyl
delta,
ato
and
sea
surface
temperature,
partly
also
from
tripty's
work.
E
We
can
convert
the
planktonic
delta
eighteen
hole
into
seawater
delta,
eating
a
waste
independently
estimated
alkanol
sea
surface
temperature,
so
at
two
locations
of
western
and
eastern
side
of
the
equatorial
pacific,
we
observed
relatively
low
values
of
seawater
deleting
out
to
the
west
during
the
mid
pricing
and
relatively
high
values
of
seawater
delta
eating
up
to
the
east.
E
So
those
contrasting
changes
does
seem
to
be
consistent
with
the
weakened
equatorial
current
and
we
cut
water
circulation
and
there's
opposite
trend
from
the
mid
ply
scene
towards
the
place
sitting
in
the
seawater
delta.
You
know
between
the
west
and
the
east,
and
this
opposite
trend
seems
to
suggesting
long-term
enhancement
in
the
equatorial
current
strengths
and
perhaps
long-term
strengthening
of
walker
circulation
as
well.
E
Here
are
some
conclusions,
so
mid-pricing
geographic
changes
don't
seem
to
affect
the
sensitivity
of
walker
circulation,
strength
to
co2
changes,
which
is
good
news
for
proxy
reconstructions,
and
we
find
two
possible
ways
to
give
us
some
estimate
of
worker
circulation
strength.
One
is
precipitation
delta
d
contrasts
between
east
and
west
and
seawater.
The
other
one
is
seawater
delta,
ignore
contrast.
They
separately
offers
us
a
view
of
worker
circulation
in
the
past,
from
the
atmosphere
and
from
the
ocean
right.
Thank
you
hope.
You
have
a
good
semester.
D
Thank
you
so
much
for
that
great
talk.
Let's
take
a
couple
of
minutes
for
questions
for
iran
and
then,
after
that
we
will
have
questions
for
both
of
our
speakers.
B
I
have
a
quick
question
so
yeah
that
that
sort
of
potential
proxy
of
the
seawater
delta-
eighteen
contrast
across
the
pacific
for
wind
strength,
is
really
cool.
I
mean:
do
you
have
a
reader?
What
sort
of
time
scales
over
which
you
would
expect
that
to
become
detectable
outside
of
the
sort
of
internal
variability
or
what
sort
of
change
in
co2
concentration
you'd
need
for
that?
To
sort
of
be
bigger
that
that
force
change.
E
Yeah
georgie,
that's
a
that's
a
good
question,
so
I
want
to
do
something
that
I
should
have
said
at
the
beginning.
So
those
are
equilibrium,
simulations
meaning
that
we
have
run
those
simulations
for
hundreds
and
hundreds
of
years
and
they
reach
the
equilibrium.
That's
a
good
question.
I
do
not!
I
I
don't
have
a
good
handle
now.
How
long
does
it
reach
equilibrium
and
what's
the
transient
response
is
going
to
look
like?
E
So
I
don't
know
that
yet
to
the
second
part
of
the
question,
how
much
co2
we
used
in
those
simulations.
It
was
just
compared
400
ppm
to
pre-industrial.
A
Okay,
go
for
it
yeah.
I
have
a
question
regarding
the
potential
use
of
sst.
So,
if
you
compare
say
delta
180
in
precipitation,
double
html
of
seawater
and
also
sst
say
we.
If
we
have
three
candidates,
maybe
which
one
has
the
highest
signal-to-noise
ratio,
getting
our
understanding
of
the
uncertainty.
E
That's
a
good
question.
I
I
think
sst
is
well
the
uncertainty
in
sst
and
seawater
awaiting
and
carbonate
awaiting
are
all
related
right.
Okay,
so
it's
hard
to
detangle
them
and
I
haven't
tried.
We
could
try
that
that's
a
good
question.
E
My
problem
was
using
the
sst
to
reconstruct
essentially
use
sst,
to
say:
oh,
it's
because
the
the
ss
changes
is
because
of
the
walker
circulation
is
weakening
just
because
the
worker
circulation,
the
sst
contrasts
they're,
not
one
to
one
they're,
not
completely
the
same
right
so
for
ssd
contrast.
You
also
have
linked
to
like
how
responsible
your
upwelling
is
to
the
changes
in
the
equatorial
current
and
the
wind
stress
and
how
responsive
the
extratropical
subtropical.
E
Tropical
overturning
is
right,
but
walker
circulation
has
its
own
degree
of
freedom,
which
is
actually
linked
to
the
tropical
growth
moist
static
stability
that
is
not
completely
tied
to
the
sst
changes,
but
more
towards
the
like
global,
the
the
vertical
structure
of
the
tropical
woman.
E
F
Hi,
john:
let's
talk,
I
have
a
quick
question.
I
was
wondering
whether
you
have
a
look
at
the
precipitation
change
in
east
and
west
side
of
the
tropical
pacific
and
whether,
if
you
have
already
cut
that
whether
that
follows
the
change
of
the
student's
worker
circulation.
E
Yes,
the
short
answer
is
yes,
so
the
in
the
toronto
paper
that
I
cited-
and
they
already
showed
that-
and
there
are
previous
studies-
also
showed
that
so
you
could,
you
can
construct
another
in
the
index
for
the
walker
circulation
or
just
looking
at
the
precipitation
contrast
between
the
warm
pool
and
the
cotton.
E
I
haven't
looked
at
the
precipitation
month
much,
but
my
understanding
is
perhaps
the
leaf
wax
delta
d
would
be
better,
since
there
is
another
conversion
from
the
2x
delta
d
to
either
precipitation
amount
or
some
kind
of
moisture
source
effect,
but
in
the
model
we
can
actually
simulate
the
flux
delta
d.
So
that's
the
only
signal
we
have
to
match
them.
That
would
be
a
cleaner
data
model.
Comparison.
F
Yeah,
I
think
that
is
also
related
to
my
next
question,
which,
which
is
more
general,
I
think,
both
for
georgie
and
ryan.
Do
you
guys
think
the
the
astro
composition
of
precipitation
is
a
better
indicator
for
the
water
circulation
compared
to
the
precipitation
amount
like
across
the
east
and
west
of
the
tropical
pacific?.
B
My
short
answer
is:
is
yes,
I
think
much
like
the
big
thumbs
up.
We
published
a
paper
on
this
just
last
year
and
much
like
the
comment
that
chippy
just
put
in
the
chat
I
mean
so
precipitation
amount.
I
mean
it's
a
relative
generally,
it's
it's
a
relatively
local
signal
compared
with
the
isotopic
composition
of
precipitation,
which
incorporates
information
from
other,
often
more
remote
processes,
with
more
information
about
the
sort
of
circulation,
dynamical
processes
that
brought
the
precipitation
to
the
location
and
it's
those
processes
that
generally
hold
more.
B
E
I
think
I
think
I'd
agree
with
you
that
I
think
the
isotopes
contains
more
of
a
circulation
signal
in
our
simulations.
It
looks
like
the
isotopic
signals
is
largely
driven
by
the
moisture
relative
changes
in
the
moisture
convergence,
so
there's
major
change
in
moisture
convergence
in
the
eastern
side
of
the
equatorial
pacific
in
response
to
the
weakening
of
the
water
circulation,
but
there's
not
much
in
the
west,
which
is
primarily
just
a
thermodynamic
response.
E
So
I
think
that's
something
perhaps
I
still
would
be
more
sensitive
to
capture.
However,
if
you
have
prop,
if
you
have
precipitation
record,
ideally
they
match
right,
the
more
record.
You
have
the
less
uncertainty
you
can
narrow
down
the
uncertainty
of
your
reconstructions,
because
all
the
records
have
uncertainties.
So
I
think
they
would
both
be
helpful.
B
I
mean
it's
a
good
point
you
made,
though,
and
I
mean
not
always,
but
but
sometimes
often
a
precipitation
amount
record
might
be
ultimately
derived
from
a
leaflex
delta
d
or
whatever.
So
there's
sort
of
more
interpretive
steps
there,
whereas
delta
d
of
precipitation
is
a
relatively
straightforward
connection
to
the
thing
we're
interested
in
the
water
circulation.
E
G
I
guess
I
had
a
kind
of
naive
outsider
view
of
where
things
are
headed
with
reconstructing
the
walker
circulation
in
response
to
volcanic
eruptions
and,
if
there's
really
much
hope
going
forward,
because
you
you
seem
to
have,
like
you
know,
state-of-the-art
reconstructions
with
really
high
temper
resolution
and
you're
suggesting
that
you
know
going
much
below
that
makes
it
extremely
difficult
to
actually
pull
it
out.
B
Yeah,
it's
a
great
question.
I
think
that
looking
at
the
response
to
individual
eruptions,
in
that
case,
I
I
would
hope,
slash
assume
that
for
it
like
summer
last,
for
example,
their
response
would
show
up.
However,
I
will
I
mean
the
response.
Generally,
you
know
of
the
the
walk
of
circulation,
for
example,
to
a
volcanic
corruption.
Much
like,
and
so
is
usually
it's
quite
short.
B
You
know,
we
see
something
happen
in
the
year
of
the
year
following
maybe
a
couple
of
years
later
and
then
and
then
essentially,
the
signal
is
merged
back
into
the
variability
of
the
climate
system,
and
so
I
I
I'm
getting
the
feeling
that
you
really
do
require
a
quite
highly
resolved
record
to
catch
that
signal
and
and
the
chronological
uncertainty
thing
is
a
real-
I
mean
it
it.
It
makes
things
much
more
difficult.
So
you
know,
for
example,
if
I
can
share
my
screen
for
a
second,
this
is
gonna.
B
Can
you
see
a
bunch
of
wiggles
yeah,
so
this
is
looking
in
so
my
reconstruction,
which
is
annually
resolved
to
the
individual
eruptions
throughout
this
reconstruction
period.
You
can
see
that,
for
example,
in
the
top
row,
the
second
from
the
left
1257
sabotage.
That
was
a
really
big
eruption.
B
I
think
the
biggest
out
of
all
of
these,
and
yet
the
composite
signal
in
the
reconstruction
ensemble
is
a
flat
line,
because
by
that
point
we've
got
maybe
chronological
uncertainty
of
plus
or
minus
five
years
or
so,
which
means
that
the
signal
is
really
hard
to
detect.