Stability of past ocean circulation: examples from past abrupt climate changes
Kerim Hestnes Nisancioglu ([email protected])
Bjerknes Centre for Climate Research, Norway
ACDC 2009, Espegrend
Part 2 - Past Abrupt Climate ChangeAdvanced Courses in the Dynamics of Climate, 2009
Past rapid climate changes as seen in ice cores:
• δ18O records of Greenland and Antarctica.
D-O and H-events of the last glacial cycle:
• Main characteristics and timing of events.
Models of past rapid climate changes:
• Crowley’s temperature seesaw;
• mechanisms behind D-O events.
New conceptual model of D-O events:
• role of the Fennoscandian ice sheet.
Synopsis of Part II
146 V. Masson-Delmotte et al.: Past temperature from ice cores
Fig. 1. Maps of Greenland and Antarctica showing first and most recent deep ice core sites. For each drilling site are indicated the name of
the deep drilling, the year of drilling completion, the final depth and an estimate of the length of the recovered climate and environmental
records.
have been showed to have impacts on the formation of these
deep and salty waters, but also to influence the hemispheric
atmospheric circulation (Dethloff et al., 2006; Yuan andMar-
tinson 2000; Chiang and Bitz, 2005). Atmospheric circula-
tion has been shown to link polar locations with other lati-
tudes (Liu et al., 2004; Marshall et al., 2001) and southern
latitudes (Guo et al., 2004; Yuan, 2004; Turner, 2004). Sev-
eral processes at play in polar regions are involved in climate
feedbacks: dynamical transport effects; atmospheric content
of water vapour (Solomon, 2006); polar cloud cover; surface
snow and ice albedo; sea-ice extent (Holland and Bitz, 2003).
Polar regions north and south are also involved in the car-
bon cycle via the role of the southern ocean and deep ocean
circulation (Kohler et al., 2005) and the soil carbon storage
in permafrost areas (Davidson and Janssens, 2006). Today,
some areas of the polar regions experience the largest ob-
served warming over the instrumental records (IPCC, 2001).
There is therefore a strong interest to obtain past temperature
reconstructions from these polar regions.
The Greenland and Antarctic polar ice sheets represent
!70% of the freshwater on Earth. Because of their volume
and time constants, these ice sheets are the slow components
of the climate system. It is for instance suggested that the av-
erage age of ice inside the Greenland ice sheet is 41±2 kyrs(respectively, 125±5 kyrs and 44±3 kyrs for the East andWest Antarctic ice sheets) (Lhomme et al., 2005). Huge
Clim. Past, 2, 145–165, 2006 www.clim-past.net/2/145/2006/
IPCC, 2007
444
Palaeoclimate Chapter 6
!"#$ %&'()'&*+,-./0&'()'&$1'/)'2)&)-3$$$$$$$$$$$',4$53,'6)(7
6.4.1 Climate Forcings and Responses Over Glacial-Interglacial Cycles
!"#"$%&#'(")'&* +$&%+,-* ,%&.($/)* "* -$0.$/&$* %1* 2#"&'"#3
'/)$+2#"&'"#*&4&#$-*&%5$+'/2*)6$*#"-)*789*:4+*'/*'&$*&%+$-*;<!=>?*
&%((./')4*($(@$+-A*B998CA*"/,*-$5$+"#*('##'%/*4$"+-*'/*,$$D*
%&$"/'&*-$,'($/)-*;E'-'$&:'*"/,*F"4(%A*B99GC*"/,*#%$--*;H'/2*
$)*"#IA*B99BCI*J6$*#"-)*8K9*:4+A*L6'&6*"+$*)6$*@$-)*,%&.($/)$,A*
"+$*&6"+"&)$+'-$,*@4*M993:4+*2#"&'"#3'/)$+2#"&'"#*&4&#$-*%1*5$+4*
#"+2$*"(D#').,$A*"-*L$##*"-*#"+2$*&#'(")$*&6"/2$-*&%++$-D%/,'/2*
)%* %)6$+* %+@')"#* D$+'%,-* ;N"4-* $)* "#IA* MO7PQ* R%S* PIMCA* "/,* ")*
('##$//'"#* )'($* -&"#$-* ;T&T"/.-* $)* "#IA* B99BQ* U%+)6VF=!A*
B998CI*?*('/%+*D+%D%+)'%/*;B9W*%/*"5$+"2$C*%1*$"&6*2#"&'"#3
'/)$+2#"&'"#* &4&#$* L"-* -D$/)* '/* )6$* L"+(* '/)$+2#"&'"#* (%,$A*
L6'&6*/%+("##4* #"-)$,*1%+*M9*)%*K9*:4+*;X'2.+$*PIKCI*J6$+$* '-*
$5',$/&$*1%+* #%/2$+* '/)$+2#"&'"#*D$+'%,-*@$)L$$/*8K9*"/,*789*
:"A*@.)*)6$-$*L$+$*"DD"+$/)#4*&%#,$+*)6"/*)6$*)4D'&"#*'/)$+2#"&'"#-*
%1*)6$*#")$-)*Y.")$+/"+4*;<!=>?*&%((./')4*($(@$+-A*B998CI*
J6$*N%#%&$/$A* )6$* #")$-)*%1* )6$-$* '/)$+2#"&'"#-A*$S)$/,-* )%* )6$*
D+$-$/)I*
J6$*'&$*&%+$*+$&%+,*'/,'&")$-*)6")*2+$$/6%.-$*2"-$-*&%35"+'$,*
L')6* "/)"+&)'&* )$(D$+").+$* %5$+* 2#"&'"#3'/)$+2#"&'"#* &4&#$-A*
-.22$-)'/2*"*&#%-$*#'/:*@$)L$$/*/").+"#*")(%-D6$+'&*2+$$/6%.-$*
2"-* 5"+'")'%/-* "/,* )$(D$+").+$* ;R%S* PIBCI*Z"+'")'%/-* '/* >[B*
%5$+* )6$* #"-)*8B9*:4+*@+%",#4* 1%##%L$,*"/)"+&)'&* )$(D$+").+$A*
)4D'&"##4* @4* -$5$+"#* &$/).+'$-* )%* "* ('##$//'.(* ;T.,$#-$$A*
B99MCI*J6$*-$0.$/&$*%1*&#'(")'&*1%+&'/2-*"/,*+$-D%/-$-*,.+'/2*
,$2#"&'")'%/-*;)+"/-')'%/-*1+%(*1.##*2#"&'"#*&%/,')'%/-*)%*L"+(*
'/)$+2#"&'"#-C* "+$* L$##* ,%&.($/)$,I* N'263+$-%#.)'%/* '&$* &%+$*
+$&%+,-* %1* )$(D$+").+$* D+%S'$-* "/,* >[B* ,.+'/2* ,$2#"&'")'%/*
'/,'&")$-*)6")*"/)"+&)'&*)$(D$+").+$*-)"+)-*)%*+'-$*-$5$+"#*6./,+$,*
4$"+-*@$1%+$*>[B* ;T%//'/*$)*"#IA*B99MQ*>"'##%/*$)*"#IA*B99KCI*
H.+'/2*)6$*#"-)*,$2#"&'")'%/A*"/,*#':$#4*"#-%*)6$*)6+$$*D+$5'%.-*
%/$-A*)6$*%/-$)*%1*L"+('/2*")*@%)6*6'26*-%.)6$+/*"/,*/%+)6$+/*
Figure 6.3. Variations of deuterium (!D; black), a proxy for local temperature, and the atmospheric concentrations of the greenhouse gases CO2 (red), CH4 (blue), and nitrous oxide (N2O; green) derived from air trapped within ice cores from Antarctica and from recent atmospheric measurements (Petit et al., 1999; Indermühle et al., 2000; EPICA com-munity members, 2004; Spahni et al., 2005; Siegenthaler et al., 2005a,b). The shading indicates the last interglacial warm periods. Interglacial periods also existed prior to 450 ka, but these were apparently colder than the typical interglacials of the latest Quaternary. The length of the current interglacial is not unusual in the context of the last 650 kyr. The stack of 57 globally distributed benthic !18O marine records (dark grey), a proxy for global ice volume fl uctuations (Lisiecki and Raymo, 2005), is displayed for comparison with the ice core data. Downward trends in the benthic !18O curve refl ect increasing ice volumes on land. Note that the shaded vertical bars are based on the ice core age model (EPICA community members, 2004), and that the marine record is plotted on its original time scale based on tuning to the orbital parameters (Lisiecki and Raymo, 2005). The stars and labels indicate atmospheric concentrations at year 2000.
Reconstructed Antarctic temperature and atmospheric greenhouse gas concentrations follow global sealevel.
Antarctic Ice Cores
Greenland Ice Cores & δ18O-ratio
Greenland temperatures as recorded in the δ18O-ratio of ice show evidence of millennial scale abrupt climate changes during the last glacial period.
warm
cold-46
-44
-42
-40
-38
-36
-34
-32
020,00040,00060,00080,000100,000120,000
NGRI
P δ1
8 O (
‰)
NGRIP Age (years BP)
δ18O of Ice Explained
The δ18O value recorded in ice cores vary with changes in local temperature of precipitated snow, changes in source area of water vapor, and δ18O in the source regions.
The effect of global ice volume on δ18O in ice is relatively small.
Ruddiman, Earth’s Climate, 2001
Historical PerspectiveWilly Dansgaard discovered link between stable isotopic composition of precipitation and atmospheric temperature by accident in 1952.
“An investigation will be undertaken as soon as the opportunity offers” - Dansgaard (1954)
Analyzed first Greenland ice core at Camp Century in 1969 drilled by Chester Langway.
Historical PerspectiveWilly Dansgaard discovered link between stable isotopic composition of precipitation and atmospheric temperature by accident in 1952.
“An investigation will be undertaken as soon as the opportunity offers” - Dansgaard (1954)
Analyzed first Greenland ice core at Camp Century in 1969 drilled by Chester Langway.
Millenial scale abrupt (“violent”) events were first discovered in 1981 after drilling Dye-3.
Hans Oeschger found “drastic climatic changes” in glacial lake deposits outside Bern matching the millennial scale events in Dye-3.
The glacial period on Greenland is punctuated by a series of rapid warming events (D-O events) with temperature changes on Greenland on the order of 10-15°C.
D-O events are most pronounced during MIS3.
Rahmstorf, Nature, 2002
Dansgaard-Oeschger Events
The glacial period on Greenland is punctuated by a series of rapid warming events (D-O events) with temperature changes on Greenland on the order of 10-15°C.
D-O events are most pronounced during MIS3.
Rahmstorf, Nature, 2002
10-15°C
Dansgaard-Oeschger Events
Time (years since start of event)
δ18O
(‰
)
Te
mp
era
ture
(oC
)
Dansgaard-Oeschger Events
Each DO-event is characterised by a rapid shift from a cold stadial climate to a warm interstadial climate.
The return from a warm interstadial to a cold stadial climate is slower.
Ganopolski & Rahmstorf, Nature, 2001
δ18
O (
‰)
SS
T (
°C)
GISP2 Age (Ka BP)
Time (years since start of event)
δ18O
(‰
)
Te
mp
era
ture
(oC
)
Dansgaard-Oeschger Events
Each DO-event is characterised by a rapid shift from a cold stadial climate to a warm interstadial climate.
The return from a warm interstadial to a cold stadial climate is slower.
Ganopolski & Rahmstorf, Nature, 2001
δ18
O (
‰)
SS
T (
°C)
GISP2 Age (Ka BP)
Greenland Summit
NW Indian Ocean Monsoon:stronger/Earlier onset during warm stadial
Carriaco Basin:increased Precip/runoff during warm stadial
Santa Barbara Basin:more ocean ventilation during warm stadial China Caves:
lower δ18O during warm stadial
Global Imprint of D-O events
Heinrich EventsThe marine geologist Hartmut Heinrich (1988) was the first to discover major episodes of ice rafting (IRD) in the North Atlantic separated by 5000 years or more.
Ruddiman, Earth’s Climate, 2001
Ruddiman’s IRD belt
What causes DO-events?
(1) Natural oscillations inherent to the internal behavior of NH ice sheets.
What causes DO-events?
(1) Natural oscillations inherent to the internal behavior of NH ice sheets.
(2) A response to solar variations external to the climate system.
What causes DO-events?
(1) Natural oscillations inherent to the internal behavior of NH ice sheets.
(2) A response to solar variations external to the climate system.
➡ A possible pacemaker?
What causes DO-events?
(1) Natural oscillations inherent to the internal behavior of NH ice sheets.
(2) A response to solar variations external to the climate system.
➡ A possible pacemaker?
(3) The result of internal interactions among several parts of the climate system (seaice, AMOC, deep ocean).
What causes DO-events?
(1) Natural oscillations inherent to the internal behavior of NH ice sheets.
(2) A response to solar variations external to the climate system.
➡ A possible pacemaker?
(3) The result of internal interactions among several parts of the climate system (seaice, AMOC, deep ocean).
➡ Most (all) theories involve freshwater perturbations to the North Atlantic.
What causes DO-events?
(1) Natural oscillations of ice margins
A: ice sheets have very long response times, however the margins are thin and capable of fast changes.
B: small amounts of heat can destabilize the ice margin causing the ice to surge and release icebergs.
C: as the ice margins thicken, the bedrock is depressed, and the ice is lifted and floated off by sea-water.
Ruddiman, Earth’s Climate, 2001
(1) Natural oscillations of ice margins
A: ice sheets have very long response times, however the margins are thin and capable of fast changes.
B: small amounts of heat can destabilize the ice margin causing the ice to surge and release icebergs.
C: as the ice margins thicken, the bedrock is depressed, and the ice is lifted and floated off by sea-water.
Ruddiman, Earth’s Climate, 2001
(2) External variations in Insolation
Changes in the production of 14C can be used as a record of the Sun’s strength in the past. The major cycle found is at 420 year. There is no significant cycle near 1500-years, and the link to climate is faint.
Ruddiman, Earth’s Climate, 2001
Broecker et al., A salt oscillator in the Glacial Atlantic?, Paleoceanography, 1990:
“we suggest that the Greenland climate changes are driven by oscillations in the salinity of the Atlantic Ocean which modulate the strength of the Atlantic's conveyor circulation.”
(3) Interactions within the climate system
Broecker et al., A salt oscillator in the Glacial Atlantic?, Paleoceanography, 1990:
“we suggest that the Greenland climate changes are driven by oscillations in the salinity of the Atlantic Ocean which modulate the strength of the Atlantic's conveyor circulation.”
Problems:
• most climate models require a large freshwater perturbation (0.1 - 1.0 Sv).
• cold stadial conditions correspond to complete AMOC shutdown.
• unknown source of freshwater on D-O timescales.
(3) Interactions within the climate system
Hysteresis of the AMOC
Response of AMOC to freshwater in simple models:e.g. Stommel (1961); Bryan (1986); Manabe & Stouffer (1988); Stocker & Wright (1991); Rahmstorf (1995).
Rahmstorf, 2002
Hysteresis of the AMOC
Response of AMOC to freshwater in simple models:e.g. Stommel (1961); Bryan (1986); Manabe & Stouffer (1988); Stocker & Wright (1991); Rahmstorf (1995).
Rahmstorf, 2002
Hysteresis of the AMOC
Response of AMOC to freshwater in simple models:e.g. Stommel (1961); Bryan (1986); Manabe & Stouffer (1988); Stocker & Wright (1991); Rahmstorf (1995).
Rahmstorf, 2002
Model Response to Freshwater
Coupled models with simplified oceans (zonal mean or idealized bathymetry)
Coupled models with 3D oceans
Rahmstorf et al., GRL, 2005
Bipolar Temperature SeesawT. J. Crowley, North Atlantic deep water cools the southern hemisphere, Paleoceanography, 1992 (see also Broecker, 1998; Stocker, 1998):
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Perturbations in northern high latitudes trigger transitions in the state of the AMOC. This gives a large signal in the North Atlantic as well as a weak opposite signal in the south due to changes in PHT.
Vellinga & Wood, Climatic Change, 2002:
HadCM3 perturbed with freshwater input to the North Atlantic between 50N and 90N.
Pure white noise freshwater forcing with a standard deviation of 0.03 Sv.
White noise freshwater forcing with periodic signal of amplitude 0.01Sv.
Simulated temperature response.
Simulated temperature response.
Stochastic forcing and DO-events
Ganopolski & Rahmstorf, Physical Review Letters, 2002
Stochastic forcing
Pure white noise freshwater forcing with a standard deviation of 0.03 Sv.
White noise freshwater forcing with periodic signal of amplitude 0.01Sv.
Simulated temperature response.
Simulated temperature response.
Stochastic forcing and DO-events
Ganopolski & Rahmstorf, Physical Review Letters, 2002
Stochastic forcing + weak periodic forcing
Questioning the Role of the Ocean in DO-events
generates detectable signals at great distances in latitude andlongitude. The extent to which it is generated in the oceanictropics, as opposed, for example, to being largely governed bystochastic westerly wind bursts of continental origin, is thesubject of great debate (e.g., Neelin et al., 1998). Thus for D–Oevents, even the detectability of a signal remotely need not leadto a deduction of its local importance, nor to inferences aboutgeneration—a subject taken up next.
North Atlantic circulation control
Our focus now changes to the separate issue of the cause (or‘‘trigger’’) of the rapid changes seen in central Greenland.Consider the widely accepted scenario that Greenland D–O
events are a direct consequence of a major shift in the NorthAtlantic meridional overturning circulation. Some confusionoccurs at the outset because of a failure to specify whichelements of that circulation are supposed to generate the climatechange (Wunsch, 2002). Often, the focus is on the mass fluxassociated with the meridional overturning, and the mass flux isindeed central to the dynamics of the ocean. But in terms of theimpact on the climate system, it is the oceanic poleward heat fluxthat has the most immediate consequences for the atmosphere.Alternatively, sea-surface temperatures are most often used todetermine how the ocean is affecting the atmospheric state,although these will be in large part a consequence of the heat fluxdivergence (exchange of enthalpy with the atmosphere).
Figure 10 displays the estimated net meridional transport ofheat by the combined ocean–atmosphere system, as well asseparate estimates of the oceanic and atmospheric contributions(Wunsch, 2005). A number of features stand out in this figure.First, the oceanic Northern Hemisphere contribution polewardof about 25-N falls very rapidly as heat is transferred to theatmosphere through the intense cyclogenesis in the mid-latitudestorm belts, and as the relative oceanic area rapidly diminishes.By 40-N, the oceanic contribution is less than 25% of theatmospheric contribution. Of this 25%, most is in the NorthAtlantic (e.g., Ganachaud and Wunsch, 2002). The assumptionthat a fractional change in this comparatively minor contributionto the global heat flux is the prime mover of abrupt climatechange is not very appealing, if there is any alternativepossibility (one will be proposed below). Furthermore, air masstrajectories circling the globe at high latitudes are in contact withthe North Atlantic Ocean for only a very short time compared tothe North Pacific, Asia and North America. The oceanic tail maynot necessarily be wagging the meteorological dog.
Figure 10. Total meridional heat transport of the combined ocean/atmosphere
system estimated from Earth Radiation Budget Experiment (ERBE) satellites
(solid), direct ocean measurements (dashed), and atmospheric contribution as a
residual (dash-dot line). Taken from Wunsch (2005). Standard error bars are
shown.
Figure 9. Power densities (a, b) and coherences (c– f) for the GRIP CH4 and y18O records. Both records were normalized to unit variance before computation so that
the power densities are dimensionless power/cycle/kyr. Coherence is plotted on both linear and logarithmic frequency scales. An approximate 95% confidence level
of no significance for the coherence magnitude is shown as a dashed line. Coherence magnitudes are above the level-of-no-significance only at low frequencies.
C. Wunsch / Quaternary Research 65 (2006) 191–203196
Carl Wunsch, Quat. Res., 2006:
Oceanic PHT decreases rapidly poleward of 25°N: at 40°N the oceanic contribution (mostly N. Atl.) is less than 25% of that of the atmosphere.
“The oceanic tail may not necessarily be wagging the meteorological dog.”
Note: changes in the AMOC and in the mass flux of the ocean do not correspond to changes in PHT and climate.
The total PHT is governed by the combined atmosphere and ocean transports which is anti-symmetric about the equator.
If this anti-symmetry is maintained as the climate system shifts, a reduction of the North Atlantic PHT would be compensated by an increase in atmospheric PHT (Stone, 1978).
Unresolved Issues with DO-events
• Discharge of ice (H-events) is focused towards the end of the cold stadials and does not occur for every DO-event.
• Unclear if changes in the AMOC have a significant impact on the total PHT at high latitudes.
• Warming event at the start of a DO is extremely rapid (cannot be explained as a direct response to AMOC recovery).
Proxy record from the Norwegian Seas
Trond Dokken et al., in prep, 2009
Figure 3
IS8 IS7 IS6 IS5a)
b)
-38
-40
-42
-44
c)
8
6
4
2
d)
e)
0.0
-0.5
-1.0
-1.5
-2.0
40 39 38 37 36 35 34 33 32 3141
5.0
4.5
4.0
3.5
3.0
4000
3000
2000
1000
IS9
40 39 38 37 36 35 34 33 32 3141
No data
Figure 1
Greenland
temperature
Subsurface
ocean temp.
time (GICC05 b2k)
Subsurface
ocean salinity
Ice Rafted
Detritus (IRD)
Brine release
due to sea ice
formation
NG
RIP
!1
8O
Te
mp
. (°
C)
Pla
nk
ton
ic
!1
8O
(s
mo
w)
IRD
> 0
.5m
m/g
Be
nth
ic !
18O
A Conceptual Model of D-O Events
Greenland temperature
Nordic Seas sea ice
Nordic Seas subsurface temperature
Fennoscandia freshwater discharge
time
A
B
C
D
1 32 4
time
A Conceptual Model of D-O Events
stadial
time
A Conceptual Model of D-O Events
stadialinterstadial
Response to Changes in Sea Ice
Camille Li et al., GRL, 2005
• Reducing winter sea ice cover in the North Atlantic has a large impact on temperatures in the region.
• Simulated warming on Greenland Summit is close to 7°C.
Changes in the Jet Streamcold state
warm state
0 10 20eddy heat flux 850mb (Km/s)
Li & Battisti, JoC, 2008
Under glacial conditions, the atmosphere can exhibit different circulation regimes
Contour: DJMF winds at 200mb, Colours: DJFM heat flux by storms at 850mb (Km/s)
When do we expect D-O events?
D-O events do not occur when sea level is below the shelf break:
• less efficient brine formation. It is not possible to form sufficient sea ice in the Nordic Seas and the halocline is not maintained.
• atmospheric jet stream is more zonal and steadier during LGM (less precipitation in the Nordic Seas)
figure by Camille Li, BCCR
Time (Ka)
Summary• The δ18O-ratio of ice in Greenland and Antarctic ice cores records
past temperatures over the past 650Ka.
• On Greenland there is evidence for millennial scale rapid changes in climate called Dansgaard-Oeschger events during the last Glacial.
• D-O events occur roughly every 1500 years and are characterized by an abrupt warming of up to 15°C, followed by a slower cooling.
• Based on work with box models and simple climate models it was believed that these D-O events where caused by a hysteresis in the AMOC in response to freshwater discharges form the Laurentide ice sheet.
• Recent work suggests that changes in sea ice and atmospheric circulation are a more likely explanation of D-O events.