Transcript What caused Glacial-Interglacial CO2 Change?

What caused Glacial-Interglacial
CO2 Change?
Douglas L. Love
Meto 658A
Spring 2006
Suggested papers:
Reviews:
Newer ideas:
Broecker and Peng
Henderson
1998 1998
Archer et al., 2000
Zeng 2003
Toggweiler et al. 2005
Paillard and Perenin 2005
Archer et al, 2000
David Archer Arne Winguth
U Chicago
U Wisconsin
David Lea
UCSB
Natalie Mahowald
NCAR
Archer et al, 2005
• Glacial pCO2 was 80-90 μatm lower than interglacial
• Radiative forcing from CO2 accounts for
half of climate change
• Tight repeatable
corellation between
 pCO2
Ice volume
 Temperature records
Archer et al, 2005
• Glacial pCO2 was 80-90 μatm lower than in the interglacial
• Radiative forcing from CO2 accounts for
half of climate change
 “The terrestrial biosphere and soil carbon reservoirs
would have to be approximately double in size to deplete
pCO2 by 80 μatm.”
 “δ13C from deep-sea CaCO3, more 12C rich during
glacial time, tells us that if anything, the terrestrial
biosphere released carbon during glacial time
[Shackleton, 1977]”
Archer et al, 2000
Glacial cycles:
• Advances and retreats of ice sheets
• Documented by isotopic composition of seawater
Oxygen in CaCO3:
 16O is selectively sequestered in glacial ice.
 Oceans become enriched in 18O
Archer et al, 2005
• Clear physical link between Northern Hemisphere
summer heating and ice sheets
• No easy link from orbital variations to pCO2.
• pCO2 rise clearly precedes the 18O of the atmosphere
by several thousand years
 (an indicator of melted ice sheets)
 Implies that pCO2 is a primary driver of melting.
• Alternatively, pCO2 could be driven by changes in
meteorological forcing:
 dust delivery of trace metals to the ocean surface
 an acausal correlation between Northern
Hemisphere summer insolation and ice volume
Archer et al, 2005
“Because CO2 is more soluble in colder water, colder sea
surface temperatures could lower pCO2.
However, the magnitude of the glacial cooling can account
for only a small fraction of the observed pCO2 drawdown.”
Archer et al, 2005
A New model of Ocean and Sediment Geochemistry Mechanisms to lower glacial pCO2:
1. Increase biological activity at surface
so that Carbon sinks to deep sea
sediments as particles
 Increase Ocean Inventory of PO43- and NO3 Change the ratio of nutrient to C in phytoplankton
 Iron limitation of biological production at surface
indicates a Southern Ocean Biological Pump could have
intensified in a dustier, more iron-rich environment.
 Glacial dust could stimulate the rate of Nitrogen
fixation, increasing the ocean pool of NO3-
Archer et al, 2005
A New model of Ocean and Sediment Geochemistry Mechanisms to lower glacial pCO2:
2. Change the pH of the whole ocean
 Convert seawater CO2 into HCO3- and CO3=,
which can’t evaporate in the atmosphere.
 pH is regulated by balance between influx of dissolved
CaCO3 and removal by burial of CaCO3 sediments.
 Timescale of 5-10 kyears is within observed timescales.
Archer et al, 2005
2. Change the pH of the whole ocean
 Conditions under which it could occur:
1) Glacial rate of weathering is higher
2) CaCO3 deposition shifts to deep sea
3) Rate of CaCO3 production decreased
4) CaCO3 compensation may also affect pCO3 response to
the biological pump in #1.
 Results:
 burial efficiency would increase
 the Ocean would become more basic
 degradation of biological C in sediments would promote
Calcite dissolution, further increasing Ocean pH.
Archer et al, 2005
A New model of Ocean and Sediment Geochemistry Mechanisms to lower glacial pCO2:
Two Caveats:
“The ocean carbon cycle is a complicated system, controlled
by biological processes we are only beginning to understand.
Thus the formulation of the model is not completely constrained by our understanding of the underlying processes.
Furthermore, we use the model to predict…conditions which
we are unable to observe except indirectly via clues preserved
in the sedimentary record.”
Archer et al, 2005
A New model of Ocean and Sediment Geochemistry Mechanisms to lower glacial pCO2 - CO2 pump scenarios:
1. Fe fertilization of existing NO3 or PO4 pools
 attains glacial pCO2 values in box models
 But not in circulation models
2. Increase NO3- by 50%
 Attains glacial pCO2 for a few thousand years until CaCO3
compensation lowers Ocean pH.
 Requires a change in the Redfield Ratio.
Archer et al, 2005
A New model of Ocean and Sediment Geochemistry Mechanisms to lower glacial pCO2 - CO3= pump scenarios:
• Coral reef hypothesis: lowered sea level causes a decrease
in shallow CaCO3 deposition, which drives increased
deposition in the deep sea
 Increased pH would lower pCO2
 Not backed up by deep-sea cores
• Rain ratio hypothesis: decrease in CaCO3 production or increase in organic carbon production could shift Ocean pH.
 A doubling of H4SiO4 could explain it, but can’t be
rationalized.
 Predicted distribution of CaCO3 on seafloor is a poor fit.
Archer et al, 2005
A New model of Ocean and Sediment Geochemistry –
Procedures and summaries:
Present Day Ocean simulation
 pCO2 within 2 μatm of observed values
 Distribution of CaCO3 a poor fit:
Present-day CaCO3
distribution on seafloor
Modeled Present-day CaCO3
distribution on seafloor
Archer et al, 2005
A New model of Ocean and Sediment Geochemistry:
The Glacial Ocean model description:
 High Lat. Air temperatures 10°-15° C colder than now
 Tropical cooling 1°-2° C cooler from plankton and O
isotope ratios
 Glacial flow field estimated from best “second guess”
velocities
 Atlantic overturning shallower and 30% slower than now
 δ13C tracer says Southern Ocean was high-nutrient, low
Oxygen, contradicting Cd data.
Archer et al, 2005
A New model of Ocean and Sediment Geochemistry:
The Glacial Ocean model results:
 Iron flux to sea surface increases by 2.5 goes to regions that
already receive sufficient iron.
 NO3- decreases from 110 x 1012 mol to 80 x 1012 mol.
 pCO2 lowered by 8 μatm.
 CO3= and H4SiO4 tweaked until burial rates of CaCO3 and
SiO2 are those of present day.
17% H4SiO4 decrease yields a 70% SiO2 burial increase.
Organic C production increased from 0.198 to 0.210
Acidification of ocean overwhelms iron fertilization,
increasing pCO2 to 280 μatm.
Archer et al, 2005
A New model of Ocean and Sediment Geochemistry:
Collapse of the terrestrial biosphere:
13C/12C
ratio in deep sea CaCO3 was .4%o lower,
indicating that an isotopically-depleted carbon reservoir
released 40 x 1015 mol C, raising the Ocean-Atmosphere
inventory by 1%
Possible sources:
 Terrestrial biomass: 40 x 1015 mol C
 Soil organic carbon: 120 x 1015 mol C
 Sedimentary C on continental shelves
Archer et al, 2005
A New model of Ocean and Sediment Geochemistry:
Collapse of the terrestrial biosphere:
Reconstructions call for 2-3 x this δ13C value.
Initially raises pC02 to 305 μatm.
 Reaction with CaCO3 will neutralize the added CO2
 Lowering to 297 μatm predicts a lowering of 17
μatm in the future.
 After compensation, pCO2 is 295 μatm.
Archer et al, 2005
A New model of Ocean and Sediment Geochemistry:
Tropical Temperatures
• Lowering Tropical Sea Surface Temperature by 4°C
decreases pCO2 by 5 μatm.
• Biological production is altered
 Stratification decreases, organic Carbon increases.
 SiO2 decreases as H4SiO4 recycling decreases.
 Small increase in pCO2.
Archer et al, 2005
Constraints on the cause of glacial/
interglacial atmospheric pCO2
• Deglacial increase leads ice volume, eliminating sea-leveldriven explanations such as submersion of continental shelves
• Deglacial transition was slow: 6-14 kyears.
The pCO2 response is much faster.
• Glacial rates of weathering and burial
were not much different than today.
• Isotopic signatures of C, N, B, Cd, Ba
• Distribution of CaCO3 and SiO2 on sea floor
Archer et al, 2005
Solution:
challenge one or more of the basic
assumptions of chemical oceanography!
1. Ocean circulation models are more diffusive than the
modern ocean, underestimating the pCO2 sensitivity
to the biological pump
2. Increase the glacial NO3- inventory beyond the PO43limitation, assuming the Redfield N/P number was
different in glacial time.
3. Double the inventory of H4SiO4 in the ocean, raising
the pH of the deep ocean.
USMAI (all campuses)
Number of hits Request permutation (No Adjacency)
0
Words= Greenhouse Puzzles Part II
Prince George’s Memorial Library System:
Keyword Search: ti:(Greenhouse Puzzles, Part II)
0 record(s) found.
Greenhouse puzzles Part 2
Secondary sources:
• A silicon-induced “alkalinity pump” hypothesis, Marine Inorganic
Chemistry/ Department of Chemical Oceanography, The Ocean Research Institute ORI,
University of Tokyo, Japan http://www.ori.u-tokyo.ac.jp/en/special/topics_4/topics-e.htm
(refers to Broecker and Peng, Part 2 1994 version as “Archer’s World”. Also references
Martin, J.H., “The Iron Hypothesis”)
• Field-based Atmospheric Oxygen Measurements and the Ocean Carbon
Cycle, PHD Thesis by Britton Bruce Stephens, Chapter 6, The Influence of Antarctic
Sea Ice on Glacial-Interglacial CO2 Variations
•Modeling of marine biogeochemical cycles with an emphasis on vertical
particle fluxes, PhD Theis by Regina Usbeck,
http://www.awi-bremerhaven.de/GEO/Publ/PhDs/RUsbeck/RUsbeck.html
•Zeng, Ning, Glacial-Interglacial Atmospheric CO2 Change - the Glacial
Burial Hypothesis. http://www.atmos.umd.edu/~zeng
Greenhouse puzzles Part 2
Secondary sources:
• ORI:
biological pump model of atmospheric CO2 variability
• Stephens:
Harvardton-Bear index:
Actual atmospheric CO2 change /
potential change due to cooling of low-latitude surface box
• Usbeck:
compares others’ works with recent estimates of total Corg
accumulation
• Zeng:
Ocean δ13C, .35%o,
land-carbon difference (Holocene - LGM) 460
Substitute or correct paper?
The sequence of events surrounding Termination II
and their implication for the cause of
glacial-interglacial CO2 changes
Wallace S. Broecker and Gideon M. Henderson, Paleoceanography, V
13 , No 4, PP. 352-364, August 1998
Wallace Broecker, Lamont-Doherty
Gideon Henderson, now at Oxford
Broecker and Henderson, 1998
Clues from the Vostok ice core:
• Antarctic Temperature and atmospheric CO2 increased
together for 8000 years, bounded by
A drop in dust flux at the onset
A drop in δ18O at the finish
• A similar lag between dust flux and foraminiferal δ18O in
the Southern Ocean indicates that the δ18O in Vostok ice is
a valid proxy for ice volume.
• Synchronous changes in CO2 and Southern Hemisphere
temperatures preceded melting of Northern Hemisphere ice
• Nutrient reorganization in North Atlantic occurs with or
after the sea level rise
Broecker and Henderson, 1998
Clues from the Vostok ice core:
• The previous observations eliminate many scenarios
proposed to explain the CO2 rise
 Those which rely on sea level change
 Conveyor-related nutrient redistribution
 North Atlantic cooling
• Southern Ocean scenarios become the front runners.
• The most popular, Iron fertilization, has 2 problems:
 Much of the dust demise occurs prior to the change in CO2,
so there must be a threshold value above which it does not
increase.
The CO2 rise continues for 4-5 kyr after the dust flux has
fallen to zero.
Broecker and Henderson, 1998
Clues from the Vostok ice core:
• Problems with iron fertilization causing the rise in CO2
may be solved if the increased iron supply in dust caused
higher rates of nitrogen fixation during Glacial periods.
 In this case, residence time of oceanic nitrate of a few
thousand years would enable decreasing productivity to be a
global rather than a local phenomenon
 This would explain the slow rampup of atmospheric CO2.
Broecker and Henderson, 1998
Timing is everything for Broecker and
Henderson. More comfortable than
their predecessors with relating time
markers, their whole theoretical setup
is based on these time relationships.
O2 created by photosynthesis has the
Isotopic composition of surface
seawater, which is controlled by
global ice volume. Turnover time is
1-2 kyears.
Therefore, δ18Oatm should have risen
1.4%o with δ18Oocean.
Broecker and Henderson, 1998
The first assumption is that
variation in ocean surface δ18O
is the only contributor to
changes in δ18Oatm.
They then claim that the Dole
Effect, where the atmosphere is
enriched in 18O by 23.5%o over
the ocean, keeps it steady.
They then present similar
offsets between events as
indicating a good correlation.
Broecker and Henderson, 1998
Broecker’s Bipolar Seesaw concept is also an important
consideration, where deepwater formation alternates between the
North and South Atlantic. This eliminates mechanisms that occur
only in the North Atlantic.
• Cooling in the Southern Ocean at the same time as CO2 is falling is
considered as a cause, but is nowhere strong enough to cause the
observed drop.
• Changing the productivity or alkalinity is also suggested as a
control of Oceanic CO2. Observations indicate that these changes
moved in the opposite direction.
• Nitrogen fixation by iron fertilization is considered, but the
residence time for NO3 is too long to keep it locally confined.
Broecker and Henderson, 1998
Tentative conclusions:
δ18O constrains the rise in atmospheric CO2 to have preceded the
melting of the North American ice sheets.
This eliminates seal level change, North Atlantic Nutrient
redistribution, and North Atlantic cooling as causes.
Iron fertilization can’t explain
 Southern Ocean paleoproductivity
 the long duration of the CO2 rise
Increased dust flux in the glacials caused more nitrogen fixation,
which allowed a greater CO2 drawdown in surface waters.
Long residence time of NO3 in ocean explains how CO2 can
continue to increase after the dust flux ix zero, and means
productivity changes can be global.
Newer ideas 1:
Zeng, Ning,
Glacial-Interglacial Atmospheric CO2 Change the Glacial Burial Hypothesis
Readily available from
http://www.atmos.umd.edu/~zeng
Newer ideas 1: Zeng, N
• Advancing ice sheets buried vegetation and
soil carbon accumulated during warm periods.
• Simulation over 2 cycles found a 547 Gt
carbon release, resulting in a 30 ppmv
increase in atmospheric CO2, the remainder
absorbed by the Ocean.
• Atmospheric δ13C drops by .3%o at
deglaciation, followed by a rapid rise to a high
interglacial value, in response to oceanic
warming and regrowth on land.
• With other ocean-based mechanisms, offers
a full explanation of the observed atmospheric
CO2 change.
Newer ideas 1: Zeng, N
Fig. 8. Modeled atmospheric
CO2 (a) and land carbon
storage (b) from the control run
and 5 sensitivity experiments
described in the text: control is
in black line, SST in green,
CO2v120 in yellow, SoilD5h in
red, Soil D20k in blue, and
WarmGlac in purple. The
largest change of a 55 ppmv
deglacial CO2 increase is due
to a cooler glacial ocean in
addition to the land carbon
release (green) and a 40 ppmv
increase due to a long delayed
regrowth (blue).
Newer ideas 1: Zeng, N
Data from Table 1:
Land carbon difference, Holocene - LGM
Newer ideas 1: Zeng, N
Look for it:
• On the ground
• Back in time
• In the models
• In comparisons
Newer Ideas 2: Paillard and Perenin
Newer Ideas 2: Paillard and Perenin
Glacial bottom waters
 were possibly much more saline
 May have an unsuspected large density
 Glacial deep stratification could account for the difference.
 Ice formation around Antarctica involves
 Brine rejection over the Continental Shelves
 Is directly linked to changes in
Sea Ice Formation
Antarctic ice-sheet extent
Kitchen Experiment: mixing saline water
Cold 30% salt water
They mixed
immediately.
Warm 10% salt water
Newer Ideas 2: Paillard and Perenin
From Wikipedia:
Newer ideas 3:
Toggweiler, JR, GFDL,
Climate change from below
Quaternary Science Reviews 24 (2005) 511-512
Newer ideas 3: Toggweiler, Climate Change from below
• Adkins et al, 2002, showed bottom waters around Antarctica are significantly saltier than the rest of the ocean, apparently from accumulation of brine during sea ice production.
• This shows that the glacial deep ocean was more stably
stratified than it is today.
• Geothermal heat would have slowly warmed it from below,
destabilizing it, like a discharging capacitor.
 Just 2°C is enough to destabilize it.
 This would take 10,000 years.
 This matches:
 Heinrich Events in the North Atlantic.
 Bond cycles in Greenland Ice Cores
 Bi-polar seesaw between Greenland and Antarctica
 He gives several examples separated by 7000 years.
Newer ideas 3: Toggweiler, Climate Change from below
• This injects salt into the upper North Atlantic, kickstarting thermohaline circulation.
• Reinvigorated circulation warms up Greenland and the
North Atlantic.
• This confirms the finding that the warmest intervals in
Greenland occur during the interstadials that follow
Heinrich and Antarctic Intervals.
• This contradicts the prevailing view
– these events are caused by fresh water input
– Explains why interstadials after Heinrich events are longer
and warmer than others.
Newer ideas 3: Toggweiler, Climate Change from below
• Short paper.
• Is it the right one?
• So I wrote and asked!
Woah! Preprint!
Newest ideas yet
Newest ideas yet: Toggweiler et al
Another new idealized general circulation model explains:
 tight correlation between atmospheric CO2 and Antarctic temp
 lead of Antarctic temp over CO2 at terminations
 Shift of ocean’s δ13C minimum from N. Pacific to Atlantic sector of
the southern Ocean
Changes occur at transitions between on and off states of the
southern overturning circulation.
Newest ideas yet: Toggweiler et al
Newest ideas yet: Toggweiler et al
Proposal: overturnings occur in nature through a positive
feedback that involves mid-latitude westerly winds.
 Glacial climates seem to have
equatorward-shifted westerlies
which allow more respired CO2
to accumulate in the deep ocean.
 Warm climates like the
present have poleward shifted
westerlies that flush respired
CO2 out of the deep ocean.
Newest ideas yet: Toggweiler et al
Contains 6 pages of good references at the end.
But wait – there’s MORE!
A silicon-induced “alkalinity pump” hypothesis
The Ocean Research Institute, University of Tokyo
In order to maintain atmospheric
CO2 at 190-200 ppm, alkalinity
and pH in the surface ocean must
be higher by ~85 μeq/L and ~.14
units, respectively.
Proposal:
species change in phytoplankton
produced only 20-25% Carbonate
plus Opal during ice ages vs 90%
now.
Quotes Broecker and Peng
and more!
Carbon Storage on exposed continental shelves
during the glacial-interglacial transition
Montenegro et al
•Up to ~10,000 years before present,
time-dependent estimate of inundated
carbon is in good agreement with the
increase in the atmospheric reservoir.
•Carbon stock of the LGM exposed
shelves cannot be ignored and merits
more detailed attention from modelling
and reconstructions.
•Quotes Zeng (2003)
and Still more!
A movable trigger: Fossil fuel CO2
and the onset of the next glaciation
Archer and Ganopolski
Uses models of how much CO2 and cooling is required to start an
ice age to predict how soon the next ice age will come, considering
how much CO2 we have/will put in the atmosphere.
“A carbon release from fossil fuels… of 500 Gton C could prevent
glaciation for the next 500,000 years….The duration and intensity
of the projected interglacial period are longer than have been seen
in the last 2.6 million years.”
Quotes Archer et al, Broecker and Henderson,
and Paillard’s PhD thesis
one more…
Glacial Hiccups
Didier Paillard
•
The climate instability of glacial
times probably resulted frm abrupt
switches in ocean circulation.
•
Figure shows Climate (temperature)
stability as a function of freshwater
input at high latitudes in the North
Atlantic.
a.
b.
c.
d.
•
unperturbed present-day state
Last Glacial Maximum
An intermediate situation 50,000 years
ago. Dansgaard-Oeschger events
Figure shows Climate
Quotes himself.
And finally,
Aric Global Climate Change Student Guide
Palaeoclimatic Change: CO2 Feedbacks
Reviews the many hypotheses of the causes of CO2 changes, and the
phase relationships of CO2, ice volume and termperature, that have
passed through many stages over the last decade or so.
Gives a review of all factors involved, with equations.
Ocean ΣCO2 profile
Ocean δ13C profile
Conclusions:
• There are many partial solutions
• This problem is a hard nut to crack.
• Truth, however, is elusive prey
- Sandra Collins, Pittsburgh Post-Gazette, March 26 2005