Timefulness
Page 17
answer lies in the many mechanisms of positive feedback—
self- amplifying processes— in the Earth’s climate system. For
example, during the long cooling periods of the Pleistocene,
areas beyond the margins of the ice sheets would have hosted
tundra ecosystems of slow- growing lichens, moss, and small
vascular plants, as in Svalbard today. When this vegetation died,
the cold temperatures would have inhibited decomposition
(largely accomplished by microbial activity, which gets sluggish
in the cold), and so organic matter would simply have accu-
mulated over the millennia in thick piles of peat. This fact was
scorched into my mind one summer in Svalbard when a col-
league and I thought we would clean up an ugly heap of plastic
142 Ch a pter 5
containers and rotting rope that had washed ashore from ships,
which often use the ocean as their rubbish bin. We set a fire
on the beach and were glad to find that the nasty trash burned
well. Then we noticed that a strip of mossy tundra a little fur-
ther inland was no longer moist and green but instead dry and
brown— and seemed to be smoking. In a sickening instant, we
realized that our fire had ignited a hidden layer of peat under
the beach cobbles. Fortunately, after several frantic minutes of
rushing back and forth with cooking kettles between the slow-
moving front and the sea, we had doused the smoldering fire.
Fire is dramatic evidence of rapid oxidation; decomposition
accomplishes the same thing invisibly, in slow motion. During
times in the Pleistocene when Milankovitch cycles caused tem-
peratures to warm even a little, tundra microbes would wake
up and get back to work, munching away at the plentiful peat
and releasing its long- sequestered carbon as carbon dioxide (or
methane where oxygen was scarce). This in turn warmed the
planet more, further accelerating the microbial feeding frenzy,
which released still more greenhouse gases, and so on, in a
classic positive feedback circle.
Other positive feedbacks in the climate system include the
albedo, or reflectivity, effect, which had played a powerful role
in sending the cooling planet into a “snowball” state at the end
of the Proterozoic. But the albedo effect works both ways: once
melting begins, the darker color of dirty ice, bare land, or open
sea water causes greater absorption of the sun’s heat, leading
to more warming, more melting, and the expansion of dark
surfaces. This accelerating warming can then amplify carbon-
cycle feedbacks even further.
Positive feedback processes can accentuate cooling— for
example, windier conditions during glacial times fertilized
iron- starved phytoplankton in the oceans with nutritious dust,
Great acceler ations 143
and when some fraction of their biomass sank to the seafloor
without decomposing, atmospheric CO2 was gradually drawn
down. But the sawtooth pattern that is so prominent in ice
and sediment cores underscores an inescapable asymmetry in
Earth’s climate system: it takes a lot longer to cool the planet
than to warm it up.
C S I C K N E S S
For us in the Anthropocene, the urgent questions are, How fast,
exactly, did past warming episodes happen, and How high were
greenhouse gas levels at those times? The last glacial maximum,
when great ice lobes left the Wisconsinan deposits of Chamber-
lin, occurred 18,000 years ago. At that time, atmospheric carbon
dioxide concentrations stood at 180 parts per million (ppm).
After that deep- winter state, orbital factors began to favor milder
conditions again, and CO2 levels rose, too. Earth then entered a
period of steady warming, interrupted by a temporary cold snap
between about 12,800 and 11,700 years ago (the Younger Dryas
interval), which is thought to have been caused by disruption
of the Gulf Stream, which conveys warm tropical waters to the
North Atlantic, as fresh water from melting ice sheets flooded
the North Atlantic. By this point, CO2 levels had risen to about
255 ppm over 6300 years, at an average rate of 0.01 ppm/yr.
When the Gulf Stream reestablished itself, it was as if the Earth
had made a New Epoch’s resolution to adopt an entirely differ-
ent mode of behavior. In a matter of just decades, around 11,700
years ago (the golden spike for the Holocene), global average
temperatures vaulted suddenly to their Holocene values, and
Earth left the bungee- jumping days of the Pleistocene behind.
But the transition into the Holocene was a time of mas-
sive geographic readjustment. The ice caps shrank, and their
144 Ch a pter 5
meltwaters ponded in vast lakes. Some of these lakes, bounded
by fragile ice barriers, drained catastrophically; the peculiar
landscape of the Channeled Scablands in eastern Washington
State records unimaginably cataclysmic flooding when an ice
dam that had impounded a volume of water equivalent to Lake
Michigan suddenly failed (sorry, Mr. Lyell). New river systems
set to work organizing drainage networks in the lumpy, deglaci-
ated landscapes. In North America, the main tributaries to the
Mississippi, the Missouri and Ohio Rivers, mark the edges of
the last ice sheet, where the greatest volumes of meltwater had
to be processed. As glacial meltwaters found their way back to
the oceans, sea level rose hundreds of feet in a few thousand
years, flooding coastal areas and changing old river valleys into
estuaries. The land bridge that had connected Asia with North
America was drowned. Britain became separated from the rest
of Europe as the Channel filled. Eventually, though, coastlines
stabilized. Weather patterns became regular and predictable.
Humans could get down to the business of raising crops and
building civilizations.
By about 1800, just before we began to consume ancient
carbon fuels in significant quantities, the concentration of at-
mospheric CO2 had risen to about 280 ppm, only 35 ppm higher
than at the start of the Holocene. This suggests that over the
course of 11,000 years, Earth’s carbon cycle had settled into an
equilibrium state in which the carbon exhaled by volcanoes and
released from decaying organic matter was about balanced by
the carbon inhaled by photosynthesizers and sequestered as
limestone. Now and then, small imbalances in the carbon bud-
get threw human societies into periods of famine and conflict.
In the decades after the industrial revolution, we, like over-
grown microbes devouring peat, began to gorge on long-
stored carbon— first coal, then petroleum and natural gas.
Great acceler ations 145
Photosynthesis and limestone precipitation could no longer
keep up. An unjust fact about carbon emissions is that while
one part of the world— the United States and western Europe—
was responsible for a disproportionate share of the twentieth-
century output, the whole world suffers the consequences.
This is
because the mixing time for the troposphere (the lower
atmosphere)— the time it takes for turbulent stirring by winds
and weather to homogenize the air on a global scale— is rel-
atively short (1 year) compared with the residence time of
carbon in the atmosphere (hundreds of years). If the mixing
time were long compared with the residence time, then carbon
emissions would hover close to the places where they were
released— like garbage piling up when trash haulers strike— and
might motivate action to curb them. But because our individ-
ual emissions are not only invisible but conveniently dispersed
around the world, we feel little incentive to curtail them.18
By 1960, the level of global atmospheric CO2 had reached
315 ppm— rising as much in 160 years as it had over the pre-
vious 11,000— at a rate of 0.22 ppm/yr, more than 20 times
the rate in the Late Pleistocene, when Earth began to heat up
significantly. In 1990, we breezed passed the 350 ppm mark,
which many climatologists consider the upper threshold for
maintaining Holocene climate stability— the point at which
the juggernaut of positive feedbacks was likely to be set off.
By 2000, the CO2 level had reached 370 ppm, rising at a rate of
2 ppm/yr. As I write, we have broken the 400 ppm ceiling, and
the rate of increase is still increasing.
In all the yo- yoing of Pleistocene climate, CO2 levels never
exceeded 400 ppm. The last time CO2 concentrations were this
high was Pliocene time, more than 4 million years ago. And
there is certainly no Pleistocene precedent for the rate at which
carbon dioxide levels are increasing. The closest analog is a
146 Ch a pter 5
climate crisis 55 million years ago, at the boundary between the
two earliest epochs in the early Cenozoic Era: the Paleocene-
Eocene Thermal Maximum, known by its less unwieldy acro-
nym, the PETM.
A D I S TA N T M I R R O R
Like eye- witness reports of an earthquake, sea- sediment cores
at dozens of sites around the globe provide vivid accounts of
the PETM. The cores all tell of a sharp shock: a sudden 5°– 8°
spike in temperature, as recorded by oxygen isotope ratios in
microfossils; a simultaneous jump in ocean acidity, marked by
a crash in the amount of calcitic shell material; and a huge in-
flux of carbon from some biogenic source, as indicated by its
unusually high enrichment in 12C relative to 13C.19 The fossil
record speaks of an ocean ecosystem in disarray: many spe-
cies of plankton suffered serious reductions in numbers, and an
extinction in bottom- dwelling microorganisms called benthic
foraminifera indicates that even the deep waters of the ocean
were affected. These changes in turn triggered a major reorga-
nization of the marine food chain. On land, hotter and more
arid conditions forced dramatic migrations of mammal species,
while one- fifth of plant species, unable to move fast enough,
went extinct. Marine and land- based records of the PETM in-
dicate that it took the oceans and biosphere 200,000 years to
achieve a new equilibrium.20
The size of the shift in carbon isotope ratios during the
PETM allows estimates of the amount of carbon that must
have been released; most calculations fall in the range of 2000
to 6000 billion metric tons, or gigatons (Gt), of carbon. ( Note:
Sometimes carbon emissions are reported as Gt of CO2, not just
C; in this case, values are greater by a factor of 3.7, reflecting
Great acceler ations 147
the higher molecular mass of CO2). The 2000– 6000 Gt figure
is hard to understand until one realizes that total cumulative
anthropogenic carbon emissions to date are around 500 Gt,
and a quarter of that has been released since the year 2000.
With rates of emissions still climbing, we are likely to reach
or exceed many of the estimates of the PETM carbon spike by
the year 2100.
An important but unresolved question is how so much bio-
genic carbon could have been released in the PETM, long be-
fore humans got into the habit of burning fossil fuels. The two
primary candidates are (1) ignition of coal or peat by magmatic
activity during the opening of the North Atlantic ocean (akin
to the long- burning underground fires that have smoldered
for 50 years beneath Centralia, Pennsylvania); and (2) sudden
vapor ization of a form of methane caged in ice— clathrate or gas
hydrate— from sediments on the seafloor. This frozen methane,
produced by microbes happily gobbling up organic matter, is
stable under only a limited range of temperature and pressure
conditions. If seawater warms, or if a submarine landslide sud-
denly uncovers a layer rich in gas hydrates, the frozen meth-
ane can become unstable and erupt from the seafloor in great
oceanic belches. Gas hydrates weren’t even known until the
1980s; before that, sediment cores commonly came up with
large voids in them, indicating that something had been lost
on ascent— the strange ices had vaporized before scientists
could even look at the cores. More efficient core recovery fi-
nally revealed what had occupied the empty spaces: ice that
could be burned. Estimates of the mass of gas hydrate currently
stockpiled in marine sediments vary from 1000 to 10,000 Gt.
Like tundra peat, these carbon stores could become unstable
as climate warms; their sudden volatilization would trigger a
nightmarish runaway greenhouse effect.
148 Ch a pter 5
But the sedimentary record of the PETM, with a resolution
no better than a few millennia, does not allow us to distinguish
between an essentially instantaneous release of carbon from a
belching ocean and a longer- term (1000- year) combustion of
coal or peat. This distinction is not of merely academic interest.
If the denominator for the rate of carbon output in the PETM
is one year, we can still cling to the idea that our emissions
are not completely unprecedented. But if the denominator is
thousands of years, our Anthropocene carbon spewing is a truly
extreme geologic outlier.
A N E W L E A F
These days, we humans are emitting more than 10 Gt of carbon
every year— mainly through fossil fuel burning, but also cement
production (which roasts limestone) and deforestation— easily
out- gassing the world’s volcanoes by a factor of 100. But could
we mimic biogeochemical cycles and find ways to take the
carbon we emit back out of the atmosphere? There are many
possible strategies, ranging from cutting- edge engineering to
direct replication of natural processes. So far, the high- tech
approaches are too expensive to be feasible and the low- tech
ones are too slow; the thing about geologic processes is that
they tend to take their own sweet geologic time.
For years, the U.S. coal industry has been pushing the oxy-
moronic idea of “clean coal,” based on the unlikely scenario
that carbon capture and sequestration (CCS) sy
stems would be
installed in power plants around the country. The technological
capability for CCS exists; it involves containing the CO2 emitted
from coal combustion, compressing the gas at high pressure,
and injecting it into porous rocks deep underground, ideally
on or near the site of the power plant (if the local geology is
Great acceler ations 149
suitable). For power plants near coastlines, some CCS schemes
have imagined disposal of CO2 in deep- ocean water, but this
would be self- defeating, since ocean acidification is one of the
effects of elevated atmospheric CO2 that the sequestration pro-
cess is trying to mitigate in the first place.
For a time in the early 2000s, it seemed possible that with
sufficient economic incentives— such as a carbon tax, or a cap-
and- trade carbon emissions market— that CCS technologies
might be implemented on a broad scale, but this was quashed
by the emergence of “unconventional” natural gas production
from shales through horizontal drilling and hydrofracturing, or
“fracking.” Energy prices fell dramatically, and the lower net
CO2 output from combustion of natural gas, compared with
coal, drained the momentum from the nascent movement to-
ward CCS. (While it is true that natural gas emits about 50%
less CO2 than coal per heat unit produced, the gas industry’s
claims about natural gas as a low- CO2 fuel are partly negated
by “fugitive” methane leaking from poorly sealed wells and
badly maintained pipelines.)21 Gas- fueled power plants could
also employ carbon capture systems, but at a steep price: con-
struction costs for new plants would be almost doubled, and
the cost of CO2 captured— which sets the lower limit for an
effective carbon tax or market value— is estimated at about $70/
ton, excluding transport and storage.22 In the present economic
and political climate, CCS seems unlikely to be the solution to
the miasma of carbon we have created.
Even if carbon capture technologies were economically vi-
able, they are not necessarily a panacea. Direct CO2 emissions
from power plants can be reduced by 80%– 90%, but signifi-
cant amounts of energy are required for the CCS process itself.
And if sequestration cannot be done on- site, transport of CO2
creates additional energy demands. Finally, the injection of