Jennifer Francis, Rutgers University

Report
Arctic Sea Ice:
The Arctic Paradox
A Tragic Tale of Loss and Revenge
Jennifer Francis, Rutgers University
or...
Glen Gerberg Weather and Climate Summit
Breckenridge, CO -- 8-13 January 2012
Photo by Janes
In the good old days…
The new normal.
The difference in ice area is ~1,300,000 miles2.
That’s an area covering about 42% of the lower 48.
generated by J. Masters
So, the question is not whether sea-ice loss is
affecting large-scale atmospheric circulation…
…it’s how can it not ?
…but, let’s back up a bit.
How did we get into this mess?
from Bill Chapman’s The Cryosphere Today
The story
fossil goes
fuel era
likebegan
this…
with the industrial
revolution…
CO2 Concentration
Good ol’ days
1990s
Ice thickness (m)
Increasing GHGs and related
feedbacks caused ice to
gradually thin.
Rothrock and Kwok, 2009
…but
Ice
extent
the recent
in the thinner
good ol’ice
days was controlled
Normal conditions
mainly
was
defenseless
by wind variations...
against the
attack of the AO+
during the ‘90s
Arctic Oscillation Index
Positive AO Index
Ice-age Movie
From NOAA’s
ClimateWatch
Ice Age is a Big Deal because… it’s a proxy for ice
thickness.
And THAT’s a big
deal because a
thinner ice cover is
more easily melted,
more easily moved
by the wind, and
more likely to follow
a trajectory of loss
as GHGs continue to
increase.
Thick ice
by Maslanik and Fowler, NSIDC, Arctic Report Card 2011
8
6
5+ years
4
4
3
2
Ice age at the end of summer
2
First year ice
0
1985
1990
Ice extent (millions of km2)
Ice extent (millions of km2)
Ice age at the end of March
1995
8
2000
6
4
2005
2010
5+ years
4
2
3
2
First year ice
0
1985
1990
1995
2000
2005
2010
From Stroeve et al. (2011) Climatic Change
The thinner ice cover is more mobile and more
vulnerable to anomalous wind patterns, like those
generated by a high-amplitude jet stream:
THE ARCTIC DIPOLE (Overland and Wang, 2010)
2007
2008
2009
2010
2011
Thick ice of the good ol’ days was much less
affected by these wind patterns (Wang et al, 2009)
All that new open water absorbs additional
solar radiation during spring and summer…
…that heats the sea surface and adds yet more
fuel to the Arctic fire…
Sea
surface
temps
∆SST
AK
Summary so far…
 GHGs => gradual thinning
 Natural variability: period
of AO+ flushes thick ice out
of Arctic in ‘90s
 Thinner ice cover more
easily pushed by winds and
melted by anomalous heat
fluxes: it can’t recover
 Additional open water
absorbs more sunlight, heats
ocean surface, melts more ice
 What’s up with all that
heat??
Ice extent anomaly
1960 1970 1980 1990 2000
During autumn and winter, energy
- lots more than normal – is being
transferred to the atmosphere as
sensible heat, water vapor, and
infrared radiation.
Which brings us back to the question:
It’s not whether sea-ice loss is affecting largescale atmospheric circulation…
…it’s how can it not ?
And what are the mechanisms ?
Budikova, Global Planet. Change, 2009
Bhatt et al, Geophys. Mono., 2008
Deser et al, J. Climate, 2007, 2010
Francis et al, GRL, 2009
Higgins and Cassano, JGR, 2009
Honda et al, GRL, 2009
Overland and Wang, Tellus, 2010
Petoukhov and Semenov, JGR, 2010
Seierstad and Bader, Clim. Dyn., 2009
Sokolova et al, GRL, 2007
This study focuses on the connections between
Arctic Amplification and extreme weather in
northern hemisphere mid-latitudes
Extreme weather = high-amplitude, slow-moving
upper-level patterns that cause persistent
weather conditions
Coldest days in Tampa
Hottest days in Atlanta
Wettest days in Chicago
500 hPa
500 hPa
500 hPa
Poleward 1000-500 hPa
thickness gradient
North Atlantic
High ice
1000-500 hPa thickness anomaly
during fall 2000 to 2010
Temperature
anomaly at
atanomaly
700 mb
mb
Near-surface
Temperature
temperature
anomaly
850
duringfall
fall2000
2000toto
to2010
2010
during
during
fall
2000
2010
Low ice
OND
9
10 11 12 1
month
North Pacific
2 3
High ice
… and winter
Low ice
from Francis et al, GRL, 2009
9
10 11 12 1
month
2 3
Data obtained from NCEP/NCAR Reanalysis, Kalnay et al.
(1996), NOAA/ESRL Physical Sciences Division, Boulder CO
from their web site at http://www.esrl.noaa.gov/psd
Connecting the dots (focus on fall and winter):
Thickness increases are larger in high latitudes than in
mid-latitudes => expect 2 main effects:
First effect: Weaker poleward temperature gradient
=> weaker zonal wind speeds. Do we see that? Yep.
N. America and N. Atlantic
OND
JFM
1000-500 hPa thickness
difference between
80-60oN and 50-30oN
14
JFM
OND
12
JAS
10
AMJ
8
Zonal mean wind at
500 hPa, 40-60oN
~ 20%
less
Weaker zonal wind speeds favor slower
moving Rossby waves, which leads to
more persistent “stuck” weather patterns.
Sound familiar?
Second effect:
Larger warming at high latitudes causes peaks of
ridges to elongate
500 hPa isopleth
 Wave amplitude
increases
 Higher-amplitude
waves progress more
slowly
 More persistent
weather patterns
Is this really happening? Let’s dig deeper:
 Focus on 500 hPa heights – integrates effects of
heating in lower troposphere.
 Select narrow height range that captures trajectory
 Analyze temporal and spatial behavior
All data for this work are from the NCEP/NCAR Reanalysis, Kalnay et al. (1996), obtained from the NOAA/ESRL
Physical Sciences Division, Boulder CO at http://www.esrl.noaa.gov/psd
Is wave amplitude really increasing?
Trends (OND)
 Wave amplitude
measured as seasonalmean difference in latitude
between ridges and troughs
at each longitude
 Amplitude is increasing
almost everywhere
How has the spatial and
temporal distribution of 500 hPa
heights changed during autumn?
Autumn (OND)
 Increased ridging
north of 50oN
 Decreased troughs
 Has wave amplitude
increased or has whole
pattern shifted
northward?
Fractional anomalies in number of gridpoints
with selected 500 hPa height
Autumn (OND)
Sept ice area
 Maximum latitude of
ridges increasing
 Bottoms of troughs
steady since ~1980
r = -0.8
r = -0.1
 Amplitude increasing
steadily since ~1980
 High correl’n with ice
r = -0.8
Where are northward elongations occurring?
Autumn
(OND)
Trends (OND)
 Ridge peaks located
mainly over western N.
America and eastern N.
Atlantic
 Number of ridge points
north of 50oN increasing
west of Greenland
Could this be
contributing to
increasing max and min
fall temperatures in
U.S. since mid-1990s??
1920
1940
1960
1980
2000
T-MAX
T-MIN
from NCDC/NOAA Climate Extremes Index
1920
1940
1960
1980
2000
Now let’s take a look at winter:
Winter (JFM)
Fractional anomalies in number of gridpoints
with selected 500 hPa height
 Increased ridging
north of 40oN
 Decreased troughs
 Increased wave
amplitude ?
Winter (JFM)
AO Index
 Maximum latitude of
ridges increasing
 Bottoms of troughs
shifting northward
r = 0.3
r = 0.7
 Amplitude increasing
steadily since late 1980s
 Weak correl’n with AO
r = 0.1
from NOAA/CPC
Sunspots
Trends (JFM)
Winter (JFM)
 Ridge peaks located
mainly over western N.
America and eastern N.
Atlantic
 Number of ridge points
north of 40oN increasing,
especially over N. America
Winter
Trends
(JFM)
(JFM)
troughs
 Troughs consolidating
along US east coast
 Fewer (weaker) troughs
over west/central US and
eastern N. Atlantic
What about summer?
 Snow is melting
earlier over highlatitude land
 Soil is exposed to
sunlight earlier, so it
dries and warms
earlier
 Further Arctic
amplification
Summer
surface
T anoms
from Rutgers Snow Lab
Summer (JAS)
May snow area
r = -0.9
 Ridge peaks and
troughs shifting
northward
 Amplitude increasing
 High correlations with
May snow area
r = -0.9
r = -0.7
Summer Trends
(JAS)(JAS)
 Preferential ridging over
western N. America
 Ridging increasing
generally, especially in
recent years and in western
N. Atlantic (=> Greenland?)
Could this be
contributing to
increasing max and min
summer temperatures
in U.S.?
1920
1940
1960
1980
2000
T-MAX
T-MIN
from NCDC/NOAA Climate Extremes Index
1920
1940
1960
1980
2000
J.A. Francis – Rutgers Univ.
Weather and Climate Summit, 2012
Summary
Arctic Amplification
High latitudes warming more than mid-latitudes,
especially in fall and winter, but also in summer over land
=> Poleward thickness gradient weakening
Weaker upper-level,
zonal-mean flow,
reduced phase speed
Peaks of upper-level ridges
elongate northward, wave
amplitude increases
 Rossby waves progress more slowly
 Weather conditions more persistent
 Increased probability of extremes: cold spells, heat
waves, flooding, prolonged snowfall, and drought
Northern Hemisphere, OND
Northern Hemisphere, JFM (5400)

similar documents