The Toba super-eruption: Micro-scale traces of a global

The Toba super-eruption:
Micro-scale traces of a global-scale climate event?
Kim M. Cobb
Stacy Carolin, Jessica Moerman, Ellery Ingall, Luke Chambers, Amelia Longo
Georgia Inst. of Technology
Nele Meckler, Jess F. Adkins
Caltech
Lydia Finney
Argonne National Lab
Victoria Smith
Oxford
Syria Lejau, Jenny Malang, Brian Clark, Alison Pritchard
Gunung Mulu National Park
Andrew Tuen
University Sans Malaysia
Explosive eruptions tear apart the f
producing fine pyroclasts (hot fra
solidified melt (i.e. glass) and crystal
hot, buoyant, pyroclast-bearing g
Plinian eruption column that they
of large explosive eruptions (Wilso
and Blake 2008 this issue). The mi
clasts leaves the volcanic vent and e
near-magmatic temperatures. It imm
heats air, rises buoyantly to heights t
73.88 ± 0.6 kybp
and then spreads laterally as a giant
(Storey etThe
al.,fragments
2012) return t
stratosphere.
ferent ways (Wilson 2008). Some fal
the eruption column
3 DREand blanket th
~3,000km
fall deposits). More energetic and im
however, are the hot and highly flui
can
VEImove
of 8across the ground surfac
dreds of kilometers per hour, coveri
sands of square kilometers with ash
The Toba
super-eruption:
Mulu
IMMEDIACY AND MAGNITU
OF THE THREAT?
The consequences of future super
portrayed in dramatic fashion in th
are discussed by Self and Blake (200
tions are “the ultimate geologic ha
immediate and devastating impact o
our social
Miller and Winfrastructure,
ark, 2008 and with re
climatic effects that will arise from l
with sulfur-rich gases. There is, how
sorts: the global frequency of volca
Likely extent of ash cloud (thick dashed line) and of pyro-
Relative eruption magnitudes
bed in terms of mass
st, can be described
he pre-vesiculation,
nsities range from
somewhat greater
ks, which have den-
ve eruptions) – this
less dense than the
roughly 103 kg m-3)
s surface, commonly
wo to three times
he material.
Miller and Wark, 2008 he mass of erupted
Relative volumes of pyroclastic material erupted at four
young volcanoes, compared with volumes of three
supereruptions discussed in this issue. The two most recent eruptions
shown, from Mount St. Helens and Mount Pinatubo, each caused
FIGURE 1
Did Toba play a role in an observed “bottleneck” in
human mitochondrial genetic diversity?
Maybe. [Ambrose 1998]
No. [Petraglia et al., 2007]
see review by Williams et al., 2012
H. floresiensis Volcanoes
impact AND
global
climate: Instrumental 54data
VOLCANIC ERUPTIONS
CLIMATE
1
100x smaller than Toba
300x smaller than Toba
Kelly et al., 1996 y global-mean surface air temperature profiles for: (a) the Krakatau eruption (1883); (b) PelQ, Soufribre and Santa Maria
g (1963); (d) El Chichon (1982); (e) composite based on events (a-d); and (0 Pinatubo (1991). The data are expressed as
rees Celsius from the appropriate monthly mean for the 5 years preceding month zero, the January of the eruption year. The
dashed horiontal line indicates the 5 percent significance level
Volcanoes impact global climate: Paleoclimate data
rmation from Paleoclimate Archives
Chapt
(a) reconstructed (grey) and simulated (red/blue) NH temperature
Strong
solar
variability
simulations
0.5
0.0
(b)
1000
MCA
Weak
solar
variability
simulations
1200
Volcanic forcing (W m-2)
Volcanic forcing (W m-2)
0
1400
Time (Year CE)
1600
(d)
0.2
0.2
0.0
0.0
1800
20C
2000
IPCC, 2013 data-model comparison
provides critical-0.2constraint on
-0.2
climate sensitivity:
temperature response
to change
-0.4
-0.4
-4in radiative forcing
-0.6
-0.6
-2
-6
-0.8
-1.0
Forcing
Volcanic forcing
coincident with
solar variability
-0.8
-1.0
Forcing
Temperature
)
Temperature
)
(c)
LIA
Solar forcing (W m-2)
-0.5
Forcing
Temperature
)
Temperature anomaly (°C)
1.0
(b)
1000
MCA
1200
(d)
1800
or data?
-0.2
(Mann et al., 2012, 2013;
Anchukaitis et al., 2012)
-0.4
Multi-decadal volcanic activity
-0.6
-40
-20
0
20
40
Year from peak forcing
m1-1-P
M5-SW
M5A-P
O-Mk3L
-ESM-P
CSM1.4
IROC-P
CCSM3
Mk3L-P
dCM3-P
ALS-g1
-E2R-P
SM4-P
-CM3.3
AM5-SS
CHO-G
0
5
10
Year from peak forcing
treecps
a08cpsl
treecps
He07tls
h13pcar
u07cvm
09regm
Lj10cps
min7eivf
a08eivl
05wave
M08ave
-5
(Timmreck et al., 2010)
1.0
-0.8
-1.0
Forcing
-0.0
Volca
coinc
solar
T
-0.0
-0.2
-0.4
-0.6
-40
Multi-decadal solar va
-20
0
Year from peak for
M5-SW
AM5-SS
IROC-P
dCM3-P
-CM3.3
O-Mk3L
ALS-g1
-E2R-P
Individual volcanic events
-0.8
NH temperature anomaly (°C)
-0.4
-0.2
Fact: models tend to cool -0.4
more
-0.6
-0.6
than paleoclimate data suggest
problem with model?
Temperature
NH temperature anomaly (°C)
-0.2
-0.4
-1.0
Forcing
-0.0
-0.2
0.0
O12glac
a08eivl
09regm
a08cpsl
Lj10cps
h13pcar
u07cvm
min7eivf
05wave
-6
0.0
Solar forcing (W m-2)
Volcanic forcing (W m-2)
Volcanic forcing (W m-2)
-4
Temperature
NH temperature anomaly (°C)
1600
0.2
-2
1.0
(c)
1400
Time (Year CE)
Volcanoes impact global climate,
0.2
but how much?
0
-0.6
LIA
solar
variability
simulations
D10107
Modeling Toba’s effects in CCSM3
ROBOCK ET AL.: DID TOBA PRODUCE GLACIATION?
Robock et al., 2009 [6] Another feedback avenue is that the c
conditions after the Toba eruption might have
of the vegetation on the planet, which in mos
dire ffects first 1a0yrs world e
would
havein produced
higher surface a
could have then
led
to further cooling. We
dissipate i
n ~
30yrs hypothesis with a GCM containing a coup
vegetation model that includes this feedback.
[7] It is the SO2 injections into the strat
volcanic eruptions that cause climate change,
ing sulfate aerosol particles remain for a few y
out solar radiation and cooling the surface [R
Jones et al. [2005] assumed that the SO2 e
Toba was 100 times that of Pinatubo, but other
it closer to 300 times Pinatubo [Bekki et al., 1996;
2002]. Would such a larger volcanic forcin
more likely to produce an ice age? To answer
we ran a GCM with a range of forcings: 33, 1
900 times the Pinatubo forcing.
[8] The Toba eruption certainly must hav
cant impacts on stratospheric chemistry, wh
affected ozone and other gases, with a chem
to prolong or enhance the forcing. To test thi
we also conducted an experiment with a m
cludes these chemistry feedbacks, using the
scenario, 300 times Pinatubo.
BEFORE missing memory in snow/ice or vegeta9on? need to look regionally? and at δ18OR +1-­‐4yr Robock et al., 2009 40
Modeling Toba’s effects inC. Timmreck
MPI-ESM
et al. / Quaternary International 258 (201
Scatter plot of maximum cooling (K) versus length of perturbation (years) for
Timmreck et Fig.
al., 11.
2012 the global average and the four specific regions (indicated by different symbols) and
for different volcanoes (indicated by different colors) based on monthly mean values.
The length of perturbation indicates the time from the start of the eruption until the
Other para
overestimating
global distribu
on the initial s
time of the YT
by running the
La Niña oscilla
Uncertainti
The assumed d
Certainly, the
duration of the
10, but the mi
pulses would
or if the inject
probably erup
the extreme m
of the erupti
(Herzog and
ignimbrite plu
the formation
vertical tracer
simulations, sh
NBH i.e. the a
Borneo stalagmites as records of regional
climate and environmental history
stalagmite oxygen
isotope records
reproducible
GREENLAND HULU reveal large
millennial-scale
excursions
(Heinrich events)
MULU Carolin et al., Science 2013 18
Fig. 2. Comparison of Borneo stalagmite δ O records to climate forcings and
GS20
largest anomaly
associated with
Toba super-eruption
also captured as
out-sized event
in Hulu stalagmite
oxygen isotopes
40Ar/39Ar
Toba date
18
Figure 2. Comparison of Borneo stalagmite δ O records to climate forcings and records of
18
paleoclimate from key regions. (A) Greenland NGRIP ice core δ O plotted using the
18
GICC05modelext
age model (Wolff et al., 2010). Greenland
Stadialto#20
(GS20)
is indicated
Fig. 2. Comparison of Borneo stalagmite δ
O records
climate
forcings
and by a
18
The handf
records of tropica
through the Toba
wet
-10 Borneo
caves
Carolin eoft-time
t al., 2013 conflictin
-9.5
regional climate e
dry
-9
sediment core fro
shows no sign of
-8.5
wet
associated with th
Hulu/
while it capt
Wang et even
al., 2001 -8 Sanbao
al., 2008 recorded
dry Wang et changes
-7
Similarly, Lane e
40Ar/39Ar((
major climate ano
-6
Toba(date(
Toba ash layer in
ash layer in South China Sea
(located over 700
stalagmites only 1
South
corresponds to initiation of
warm
27.0
Huang emillennial-scale
t al., 2001 China
δ
306
S.-R. Song SST
et al. / Marine Geology 167 (2000) 303–312
cooler reconstructed
26.5 Sea
samples that is re
of dark subhedral biotite and rarer hornblende crystal.
cold
26.0
magnitude (Carol
The gradational contacts of the tephra layer with host
sediments in the top and bottom are a total of 20 cm in
mirrored in sever
thickness. This suggests that the tephra had been
reworked by bottom current and biotubation, which
δ18O records (Wa
likely diffused the tephra within
the host sediments.
NGRIP
warm
(Figure 1). Situat
Two size fractions (63–150
and
!150 !m) were
-40
NGRIP, proxy
2004 record from
separated and sampled for morphological and
geochemical analyses. -42
cold
sharp, prolonged
-44glass shards was studied by
The morphology of the
2cm layer of Tob
the Riguku Scanning Electron Microscopy (SEM) at
-47
the Department of Geology, National Taiwan Univeral., 2001).
warm
-48
71
74
77
80
δ18O (‰)
GS20
δ18O (‰)
SST(°C)
δ18O (‰)
A closer look
68
δ18O (‰)
65
200µm
sity. The major element compositions of the glass
EDML
shards and mineral grains were
determined using a
-49
JEOL JXA-8900R Electron Micro-Probe Analyzer
Earth syst
cold of Earth Sciences, Academia
(EPMA) at the Institute
ice$core$$
EPICA, 2
006 sized eruption wo
Sinica, Taiwan. Operating conditions were sulfate$peaks$
15 kV,
Svensson e
t al., Earth’s
2012 climat
10 nA and were focused ("1 !m) for accelerated
on
71 The74data 77 80
voltage, probe current and65
beam 68
diameter.
Timmreck et al.,
were processed using the ZAF correction
routine.
Age (kybp)
Trace elements and REE were determined by Inducsize, duration, an
Figure 1.Spectrometry
Potential (ICP-MS)
signatures of the Toba
tively Coupled Plasma-Mass
18
climatic impacts
super-eruption
in: Borneo
using a Perkin–Elmer
Elan-6000 spectrometer
at the stalagmite δ O
68
71
74
-10 Borneo
caves
-9.5
-9
-8.5
δ18O (‰)
Hulu/
-8 Sanbao
-7
27.0
26.5
South
China
Sea
26.0
NGRIP
-40
-42
-44
GS20
multiple sulfate peaks in
Antarctic ice core tied
to sulfate peaks in Greenland
à  relationship to Toba
eruption(s)?
δ18O (‰)
SST(°C)
-6
warm
EDML
The handf
records of tropica
through the Toba
wet
Carolin eoft-time
t al., 2013 conflictin
regional climate e
dry
sediment core fro
shows no sign of
wet
associated with th
while it capt
Wang et even
al., 2001 al., 2008 recorded
dry Wang et changes
Similarly, Lane e
40Ar/39Ar((
major climate ano
Toba(date(
Toba ash layer in
(located over 700
stalagmites only 1
warm
Huang emillennial-scale
t al., 2001 δ
samples that is re
cold
magnitude (Carol
mirrored in sever
δ18O records (Wa
warm
(Figure 1). Situat
NGRIP, proxy
2004 record from
cold
sharp, prolonged
2cm layer of Tob
-47
al., 2001).
-48
77
80
-49
cold
ice$core$$
sulfate$peaks$
65
68
71
74
77
80
δ18O (‰)
A closer look
δ18O (‰)
65
Earth syst
EPICA, 2006 sized eruption wo
Svensson et on
al., Earth’s
2012 climat
Age (kybp)
Figure 1. Potential signatures of the Toba
18
super-eruption in: Borneo stalagmite δ O
Timmreck et al.,
size, duration, an
climatic impacts
Research questions:
1)  How many times did Toba erupt ~74,000yrs ago?
And with what relative sizes?
2)  What were the regional climate impacts of the
eruption(s)?
3)  Did the Toba-related climate effects in Borneo
occur before, during, or after the initiation of
pronounced regional drying?
Approach:
A
Find the volcanic markers in the Borneo
stalagmites.
Compare to oxygen isotope-based
climate record in same stalagmites.
B
Available stalagmites
18
re 4. Photos of candidate stalagmite samples for the reconstruction of high-resolution δ O and
ed geochemical markers. High-precision U/Th dates are denoted in yellow, in kybp (individual a
Preliminary synchrotron Fe data
U/Th%date%
73,636±305yrs%
Toba%ash%40Ar/39Ar%age%=%
73,880±640yrs%%
(Storey%et%al.,%2012)%
ect/)
s
n
a
r
)t
g
n
li
samp
on)
3
c
e
ir
)d
h
t
w
gro
Fe%layer%
(Toba?)%
U/Th%date%
80,758±208yrs%
Fe%layers%
(known%hiatus)%
Preliminary SIMS Sulfur profile
~550yrs
initiation of
steep
δ18O shift in
bulk data
10µm beam = ~2.2yrs
20µm sampling res = 4yrs
data from Stacy Carolin Next steps:
Geochemically fingerprint
i)  Toba ash
ii)  Mulu clays
concentrations of iron. Raw counts in th
background signal of the stalagmite. Ligh
change in the crystallographic orientation
perfectly aligned with the growth/samplin
dating samples, with resulting dates, are
the sample from the material removed for
indicated by green arrows. The radiometr
Compare to synchrotron/SIMS scans
across Toba horizon
(Fe, Si, S, Br, K, Al)
other
volcanic
ashes
K2O%wt%%%
*see review by Frisia et al., 2012
TOBA ASH
CLAYS
Lane et al., 2013 Mermut and Cano, 2001 <1% K2O
SiO2%wt%%%
Figure 7. Plot illustrating the
Fig.
glas
YTT
plat
elon
glas
(28.
dist
pos
tinc
(26)
noe
tion
gen
cisio
wit
as
“f
ge
pr
cla
70
(F
(C
an
fro
cla