an average pace that is deliberate, they are not so slow as to be
beyond our perception.
That additive and subtractive topography- forming processes
are so commensurate is one of Earth’s extraordinary attributes.
The landscapes of other rocky planets and moons look alien
precisely because these worlds lack such a balance in the rates
of creative and destructive topographic agents. On Earth, if
tectonics far outpaced erosion, mountain plateaus would per-
sist longer, creating vast areas of alpine habitat. If erosion out-
stripped tectonics, the continents would be lower but more
rugged, and rivers would carry greater volumes of sediment to
continental shelves, dramatically changing the nature of coastal
regions. In either case, life on land and in the sea would face
different natural selection pressures, and evolution would likely
have followed alternative routes. However, life itself can alter
the processes that shape topography: there is strong evidence
that colonization of land by plants in early Silurian time (ca. 400
million years ago) slowed global erosion rates and led to the
emergence of rivers with well- defined channels.14 (It has taken
humans only a few centuries to reverse that trend; by some
estimates, modern erosion rates— accelerated by deforestation,
agriculture, desertification, and urbanization— are orders of
magnitude higher than geologic averages.15)
80 Ch a pter 3
Remarkably, the pace of biological evolution is well matched
with the rates of tectonic and surface processes over geologic
timescales. This is particularly evident in the Hawaiian Islands,
which have formed in sequence from northwest to southeast,
as the Pacific Plate has passed over a deep- seated “hotspot”
where mantle rock wells up and melts through decompres-
sion. A study of biodiversity on each island over time shows
that adaptive radiations— bursts of evolutionary innovation—
coincided with the growth of each island through volcanism,
then leveled off as erosion gained the upper hand, reducing
an island’s area and elevation range.16 And of course, Darwin’s
original insights about evolution came from the diversity of
species on the equally youthful Galapagos Islands (whose age
was, however, not known to him). One could imagine an alter
ego planet where surface morphology changed too quickly
for evolutionary adaptation of macroscopic life, like a ballet
orchestra that is playing so fast the dancers can’t keep up. For-
tunately, all members of the Earth ensemble— volcanoes, rain-
drops, ferns, and finches— perform in synchrony.
R A I N A N D T E R R A I N
A closer look at how mountains develop reveals even subtler
relationships between tectonics and erosion— and further com-
plicates the Hutton- Dylan riddle. First, rates of erosion depend
on weather and climate, and tectonic topography can change
both. Like air travelers allowed to take only small amounts
of liquid past security checkpoints, air masses are forced by
mountains to drop their moisture as they pass over the crest
line, which creates rain shadows on leeward slopes and leads
to asymmetrical rates of erosion across the mountain belt. In
India, the intensity of the annual monsoon is directly linked
The pace of the earth 81
with the existence of the Himalaya, leading to ferocious ero-
sion in the steep foothills. The Tibetan Plateau, meanwhile,
owes its height in part to the arid conditions the mountains
themselves have created. However, aridity leads to lack of veg-
etation, which makes slopes more vulnerable to gravitational
failure in landslides. As they grow, then, mountains create their
own complex climate systems, which in turn shape their future
evolution.17
Great mountain belts like the Himalaya can even change
global climate. During the Cretaceous Period, before the col-
lision of India and Asia, Earth had a hothouse climate, with
no glaciers or ice caps. An inland sea covered the Great Plains
region of North America and lapped up against western Min-
nesota. Seafloor spreading had been unusually fast for about
40 million years, leading to higher- than- average amounts of
volcanic carbon dioxide (CO2) in the atmosphere. Some dino-
saurs even lived at high arctic latitudes. Starting in the early
Cenozoic Era, at about the same time that the Himalaya began
to rise, Earth’s climate began the long- term cooling trend that
has characterized the last 50 million years. Many geologists be-
lieve there is a causal connection between this cooling and the
creation of high topography in the Himalaya. In particular, the
chemical weathering of rocks by rainwater is, over geologic
timescales, an important mechanism for drawing carbon di-
oxide (CO2), the most abundant greenhouse gas, out of Earth’s
atmosphere (see figure 9 and chapters 4 and 5).
In the absence of human activity, CO2 comes mainly from
volcanic exhalations. When CO2 mixes with water vapor in the
atmosphere, it forms a weak acid (carbonic acid, H2CO3) that is
effective at dissolving rocks over time. Many crustal rocks con-
tain calcium, which is then carried in solution by rivers to the
world’s oceans. In the sea, organisms ranging from corals and
82 Ch a pter 3
F I G U R E 9 . The long- term carbon cycle; the weathering of mountains regulates atmospheric CO2
starfish to single- celled zooplankton use this calcium, together
with bicarbonate (HCO3−) to form shells and exoskeletons
made of calcite (CaCO3). This whole process can be written in
simplified form as a sequence of chemical reactions:
Rock weathering → Ions dissolved in rivers → Formation of limestone
CO2 + H2O + CaSiO3 → Ca2+ + 2HCO3−+ SiO2 → CaCO3 + SiO2 + CO2 + H2O
Combine
Simplified
Calcium
Bicarbonate
Calcite
Silica (used by
to make acid
composition in solution
secreted
other organisms
of igneous rock
by marine such as sponges)
organisms
The pace of the earth 83
But the most critical step, from the perspective of long- term
climate modulation, is that when calcite- secreting organisms
die, their mineral remains rain down to the seafloor to form
limestone, locking up atmospheric carbon dioxide in solid
form, where it remains for tens of millions of years.
This is Earth’s long- term carbon sequestration program— a
greatly underappreciated ecosystem service— and it is more
efficient at times when lots of fresh rock surfaces are being
made available for chemical weathering, such as during the
construction of a Himalayan- scale mountain belt. The growth
of the Himalaya, then, influenced not only local and regional
weather patterns, but climate and even topography at a global
scale, ultimately helping push Earth into the Ice Age, when
&
nbsp; glaciers and ice caps reshaped landscapes all over the world.
P E A K P E R F O R M A N C E S
Another even subtler, and counterintuitive, link between ero-
sion and mountain building involves the way that a mountain
belt interacts with Earth’s mantle. As mountains form owing to
tectonic collision and crustal thickening, the added weight of so
much rock piled up in one place causes the weak (though solid)
upper mantle— called the asthenosphere— to be displaced, like
the water beneath a heavily laden ship. But once a mountain
belt stops growing (as in the case of the young but no longer
tectonically active Alps), erosion gets the upper hand and re-
duces the weight of the crust. This causes the displaced mantle
to flow back into place and the mountains to rise in elevation,
like a ship emptied of cargo (such isostatic rebound also occurs
in areas previously covered by thick sheets of glacial ice18). In
this way, erosion paradoxically helps raise mountains.19
84 Ch a pter 3
Throughout the life of a mountain belt, then, crustal defor-
mation, climate, erosion, and mantle displacement perform a
languid interactive dance in which each player influences the
motions of the others. But on occasion, their slow- motion
choreography is disrupted by sudden jumps and jetés. Charles
Darwin, who experienced a great earthquake in Chile while on
the Beagle expedition, was perhaps the first to speculate that
these destructive events may in fact help build mountains over
time, even though the cause of earthquakes— sudden slip on
faults— was not fully understood at the time. Noting a bed of
“putrid mussel- shells” that had been heaved 3 m (10 ft) above
the high- water mark by the earthquake, Darwin speculated that
older seashells he found at elevations up to 180 m (600 ft) had
been brought there by “successive small uprisings, such as that
which accompanied or caused the earthquake of this year.”20
As usual, Darwin was right.
Unlike most geologic processes, which are difficult to study
because they elapse slowly, earthquakes can be experienced
in real time, but they occur at inaccessible depths. No one has
ever directly witnessed what happens on a fault surface deep
in the crust when an earthquake occurs, but a century of seis-
mologic research integrating elastic wave theory, experimen-
tal rock mechanics, and analysis of modern and ancient fault
zones makes it possible to extract many types of quantitative
inferences from the squiggly lines on a seismogram. The largest
earthquakes are magnitude 9 (M9) megathrust events in sub-
duction zones, like those that took place in Indonesia in 2004
and Japan in 2011. Such events can accomplish in minutes what
would take hundreds of years at background tectonic rates.
In the devastating tsunami- inducing 2004 Sumatra earth-
quake, an astonishing 1100 km (700 mi) of the plate boundary
The pace of the earth 85
was activated.21 The underwater rupture propagated northward
from its origin over a period of 10 hellish minutes at a velocity
of more than 1.6 km/s (1 mi/sec),or 6900 km/h (4300 mi/hr).
All along this distance, the Sunda plate, carrying Indonesia,
lurched an average of 20 m (65 ft) westward, a displacement
equivalent to about 1000 years’ worth of normal plate motion.
As each successive segment of the plate boundary slipped,
power ful seismic waves— the cause of the actual ground-
shaking in an earthquake— were generated, moving outward
in concentric circles like ripples on a pond, at speeds of 3 to
5 km/s (2 to 3 mi/sec). Clocking these rates is of more than
academic interest; while rupture fronts and seismic waves are
fast, electromagnetic waves that transmit digital information
are still faster. In Indonesia, Japan, and other areas of high
seismic risk, cellphone earthquake and tsunami alert systems
have been implemented in the hope that a few critical seconds
of warning may help save lives in future events.
While we can’t predict exactly when or where great earth-
quakes will occur, we can say with utter certainty there will
be many more. Global instrumental seismic records now span
almost a century and show that, on average, an M9 megathrust
earthquake can be expected along one of Earth’s subduction
zones every few decades. Worldwide, on all types of faults,
there are typically one or two M8 and tens of M7 events each
year.22 Building earthquake- resistant housing in seismically
active regions should be one of the world’s top humanitar-
ian priorities. In the twenty- first century, an M7 earthquake
should not cause 100,000 deaths, as the January 2010 Haiti
quake did. Our surprise and shock when yet another earth-
quake devastates a city and claims thousands of lives is almost
medieval.
86 Ch a pter 3
FA U LT Y L O G I C
For decades, geoscientists thought that faults accommodated
deformation of the Earth’s crust in two distinct modes with
radically different tempos: fast and furious (meters per second)
during earthquakes, but slow and steady (centimeters per year)
the rest of the time. Furthermore, there seemed little common
ground between the physical phenomena that occurred on fault
zones at such different timescales. As a result, seismologists
who study earthquakes, and geologists who study the gradual
tectonic processes that build mountain belts (“structural” geol-
ogists, like me), have traditionally been two distinct academic
clans. More recently, however, the two fields have begun to
converge. In the late 1980s, a distinctive glassy rock type with
the cumbersome name pseuodotachylyte, sometimes found in
ancient fault zones, was shown to be the product of localized
frictional melting, which could have happened only at slip rates
of meters per second— that is, during earthquakes. This dis-
covery has made it possible to observe directly the physical
consequences of seismic slip on rocks that were at the focus of
an earthquake. And since the start of this millennium, a new
generation of seismic arrays, combined with high- resolution
GPS monitoring of ground motion and more powerful data
processing, has led to the discovery that faults have a much
wider spectrum of behaviors than previously thought.
Between long- term “creep” that occurs at background tec-
tonic rates, and conventional earthquakes that occur in seconds
to minutes, geoscientists have now documented intermediate
events called slow earthquakes that elapse over days to weeks,
generating very low frequency tremors that had previously
been dismissed as noise. In contrast with earthquake rupture
speeds of kilometers per second, these events propagate along
The pace of the earth 87
a fault zone at a sedate— even walkable— rate of 16 to 32 km
(10– 20 mi) per day. Oddly, some of them then double back on
themselves, reversing their sense of prop
agation at a slightly
higher velocity than they took on their outward journey,23
like a hiker who quickly retraces her footsteps to pick up a
dropped mitten. Stranger still, slow- slip events on some faults
zones recur at regular, but cryptic, intervals. On the Cascadia
subduction zone off the coast of Washington State and British
Columbia, for example, slow earthquakes follow a 14- month
cycle whose significance is not understood.24
The causes and consequences of slow- motion seismicity are
not yet clear. Many geologists think these episodes could be
related to fluids percolating through deforming rocks, and if so,
mineralized fractures in ancient rocks, called veins— the source
of many metallic ores— may in fact be records of ancient slow
earthquakes. While this concept is intriguing, a more import-
ant question is the relationship between sluggish earthquakes
and sudden, devastating ones. Do slow earthquakes help reduce
stress on faults by relieving it in increments, or do they presage
larger, potentially catastrophic events?25 Studies of fault zones
around the world— in the western United States, New Zealand,
Japan, Central America— suggest that the answer may not be
the same for all depths and fault zones, which is an unsettling
conclusion. It also seems likely that faults have secret habits on
timescales of centuries to millennia that as yet fall outside the
range of our observational abilities.
G O I N G D O W N H I L L
Just as the building of mountains is generally unhurried but
sometimes impulsive, their demolition alternates between
continuous and quantized. We humans think we are glimpsing
88 Ch a pter 3
eternity in rocky alpine landscapes, but in fact they inspire
thoughts of infinity in us just at the point they themselves
sense mortality. Majestic peaks and magnificent palisades are
simply what remains, for now— the provisional results of the
latest cuts by a team of obsessive sculptors: water, ice, and wind
in artistic collaboration with gravity. Yet we are shocked when
a rockfall scars a cherished cliff face in Yosemite or disfigures
the iconic Old Man of the Mountain in New Hampshire. Some
research in the field of geomorphology (the study of landscape
evolution) suggest that in mountainous areas, episodic land-
Timefulness Page 10