¿ìè¶ÌÊÓƵ

Algernon Charles Swinburne was not known for his interest in earth sciences,
yet one of his more famous lines has caught the eyes of geologists and geomorphologists.
‘Even the weariest river winds somewhere safe to sea,’ he wrote, with more
literal truth than he realised. Every geographer knows that rivers flow
towards the coast. But there are some rivers that appear to break this rule.
And exciting new ideas in earth sciences have come from understanding why
some particularly weary rivers wind to sea thousands of kilometres from
their sources, rather than heading for the nearby coast.

Some major rivers in India, Africa and South America follow this pattern,
according to Keith Cox, of the Department of Earth Sciences at the University
of Oxford: ‘It’s always the same. You get out of the ship, you go up the
hill, and then you find that the rivers are draining out into the ocean
on the other side, or at least going in quite the wrong way.’ And these
areas have other things in common too – they are places where continents
split apart, and the site of some of the biggest volcanic eruptions on Earth.

Cox usually studies the chemistry of volcanic rocks. Now he has gained
a new understanding of the processes responsible for some of the rocks he
studies by looking at the landscape, rather than the rocks which form it.
The unusual drainage pattern of rivers flowing away from the edges of some
continents, coupled with the huge volumes of lava that flooded across those
same continents at the time they split apart, suggested to him a link with
geophysical models about mantle plumes.

Mantle plumes are columns of hot rock that rise through the Earth’s
interior – the mantle, which lies between the crust and the core. As they
approach the surface, they heat the crust above and form an area known as
a hot spot. Geologists first invoked the idea of moving hot spots some 30
years ago to explain chains of volcanic islands in the middle of the oceans.
Once the theory of plate tectonics was established, the chains of islands
became evidence for ocean plates moving over static hot spots in the mantle
below. The Hawaiian islands are one such chain; the active volcanoes are
above the hot spot now, and the inactive volcanic islands and the even older
Emperor chain of submerged volcanoes mark times when their parts of the
plate were over the plume.

Now geologists realise that plumes can rise beneath continents as well
as oceans; the volcanic eruptions in the past few million years at Yellowstone
National Park in Wyoming and the hot springs that bubble away there now
owe their existence to a continental hot spot.

Mantle plumes are part of the convective circulation in the Earth’s
mantle, driven by heat from the core and complicated by heat generated by
radioactive decay of isotopes in the mantle rocks themselves. Researchers
think plumes begin when heat escapes from the Earth’s core, warming material
in the deepest layer of the mantle, so that it becomes lower in density
than its surroundings. The low density layer grows thicker and eventually
becomes unstable. It rises towards the Earth’s crust in the form of a rising
plume of hot, but still solid, rock, travelling at speeds of around a metre
per year. They take several million years to travel the few thousand kilometres
to the crust.

Recent studies of seismic waves generated by earthquakes reveal that
plumes could have a role to play in the rifting of continents and the huge
eruptions of lava that happen at about the same time. Robert White and Dan
McKenzie of the University of Cambridge postulate that when a mantle plume
rises up below a part of the Earth’s crust which is already stretched and
thinned, it heats up the surrounding rock by between 100 and 200 °C.
In geological terms this is only a slight heating, but when combined with
the drop in pressure as the plume nears the surface, it seems to be all
that is needed to melt the lithosphere (the crust and upper part of the
underlying mantle), push the crust into a dome and spark off big volcanic
eruptions.

Geologists can study the behaviour of modern plumes directly in places
such as the Hawaiian islands, where the rock melted by the extra heat of
the plume erupts at the surface as lava. But Hawaii and other ocean islands
lie on relatively thin and easily broken ocean crust, about 7 kilometres
thick on average. The situation is different when a plume comes up below
the stronger continental crust, which is on average around 40 kilometres
thick. This is more difficult to visualise, but current theories suggest
that a plume under continental crust is forced to spread out sideways, and
may form a hot region between 1000 and 2000 kilometres across. The theory
says that the hot rock pushes the overlying crust up into a broad dome,
which can be 2 kilometres high in the middle and 2000 kilometres wide. In
response to this uplift the continent will often show a tendency to rift,
or break apart. The heat of the plume may be responsible for the massive
piles of flood basalts in areas as widespread as the Deccan Traps in northwestern
India, the Karoo in southeastern Africa, and the Parana Basin in southern
Brazil.

Many of the rocks Cox studies are lavas resulting from this type of
volcanism, now forming sequences of basalts up to 18 kilometres thick, that
spread over hundreds of thousands of square kilometres. They are known as
flood basalts because they spread over such large areas and erupted so quickly
– some took as little as 1 million years, which is instantaneous in geological
terms. The ages of these volcanic rocks and their distribution show that
the large plumes thought to create this type of eruption happened at intervals
of roughly 20 or 30 million years at various places over the Earth’s surface
in the past 250 million years.

Volcanism on this scale is hundreds or thousands of times bigger than
anything seen at active mid-ocean ridges or on land in recorded history.
Cox cites the example of the Deccan Traps in northwest India, sheets of
lava erupted when the Seychelles split off from western India 65 million
years ago. ‘One single lava flow from something like the Deccan province
in India is capable of covering an area the size of Ireland with a sheet
of lava at 1200 °C in probably a week or two,’ he says. ‘That’s the
sort of thing that would make Mount St Helens look completely trivial.’

Dating the domes

Catastrophic eruptions of this type must have had a major effect on
the climate and ecological stability of life on Earth. Some scientists,
such as Vincent Courtillot, of the Paris Institute of Earth Physics, speculate
that the volcanism 66 million years ago, which resulted in the Deccan Traps,
may have been a cause of the roughly simultaneous extinction of many groups
of organisms, including the dinosaurs.

Surprisingly, when the continent moves away from the plume the domes
still remain. Cox explains this by analogy with icebergs; the crust of the
Earth floats on the mantle, like an iceberg on water. Just as a thick iceberg
stands higher above the water than a thin one, so the thicker parts of the
Earth’s crust rise higher. Geologists believe that because of the sheer
volume of melting associated with a plume, continental crust near its axis
is thickened by the solidification of molten rock on its base. This means
that although the crust rises in the first place because of the thermal
and dynamic effects of the plume of hot rock, the crust stays high after
it has moved off the plume because it is thicker. Only when the crust has
returned to a normal thickness can the surface subside.

Because the domes remain long after the crust has moved away from the
plume, the characteristic river drainage patterns have time to develop.
The rivers cut down into the surface of the Earth out from the high point
of the bulge and expose the older lava flows. By dating these rocks using,
for example, the rate of decay of radioactive isotopes in the minerals,
geologists can find out when the lava formed, and so the oldest time when
the rivers could have begun to cut through them. This type of work has demonstrated
to Cox that in some areas, southeastern Africa, for example, the drainage
patterns are almost 200 million years old.

Cox reasoned that if plumes had pushed the land surface into a dome,
river patterns seen today might preserve the radial drainage pattern. To
test his idea, Cox abandoned his more familiar tools of hammer, microscope
and mass spectrometer in favour of topographic maps and satellite images,
and set to work. ‘What I have been doing is simply going around and looking
at rivers and superimposing the drainage pattern on the maps of plumes provided
by geophysicists,’ he says. ‘I’ve been looking to see whether the rivers
seem to know about the geophysical model: do they flow in the right direction
off the dome? The answer is yes, in all the cases I’ve looked at there is
very good correspondence, although in older provinces like the 190-million-year-old
Karoo in southeastern Africa, the pattern is only partly preserved.’

The drainage patterns Cox studied reflect landscapes that developed
after the plume had done its work generating huge piles of volcanic rocks
and splitting continents in two. Rather than seeing a complete circular
pattern, Cox has found half domes, split by the new ocean. Ray Kent of the
Department of Geology at the University of Leicester believes that it may
also be possible to use drainage patterns to recognise uplift caused by
an underlying plume as early as 150 million years before volcanism began
and the continent split apart, even though that uplift is no longer preserved
in the topography.

Whereas Cox can study modern drainage patterns to test his ideas, Kent
uses sedimentary rocks to look at ancient drainage patterns. He looked at
the directions in which water flowed in braided rivers from early Permian
to early Cretaceous times, between 260 and 145 million years ago, in northeastern
India. His information on the direction in which the rivers flowed came
from the distribution of pebbles and fine debris carried by the rivers and
preserved in the sedimentary rocks. In such rivers today, finer fragments
of rock tend to travel further along the river, whereas coarse grains and
stones stay upstream. On a smaller scale, Kent used the structures formed
in the river bed and bars to infer the local flow patterns. The slope he
has identified in sediments deposited before the volcanism began developed
early in the life of the plume that broke India apart from Australia and
Antarctica around 120 million years ago.

White and McKenzie believe that the upward pressure of the rising plume
also forces the continent to break apart, creating a new ocean. Africa and
South America were probably split by just such a mechanism 130 million years
ago. The plume that caused the split is now sitting below the island of
Tristan da Cunha in the middle of the South Atlantic.

Other geologists, including Courtillot, Mark Richards of the University
of Oregon and Robert Duncan, of Oregon State University in the US, are also
convinced that mantle plumes are responsible for huge piles of solidified
lava in some parts of the world. But they do not believe that it is necessary
for a plume to rise up below a thinned or rifted part of the crust in order
to melt rocks in the overlying crust and produce the massive volcanism.
Instead, they hold that the volcanism is triggered by heat alone, in the
form of the hot ‘head’ of a plume, which is probably 250 kilometres across.
In contrast, they think that the cooler ‘tail’ of a plume is responsible
for the much smaller scale volcanism which form chains of ocean islands.

In support of their idea that huge amounts of volcanism can occur in
the absence of continental rifting, Courtillot and his colleagues point
out that there are places, such as Siberia in the USSR and the Columbia
River plateau in the northwest of the US, where huge piles of volcanic rocks
are not associated with continental break-up. They also believe that some
of the major volcanic episodes linked to rifting by White and McKenzie happened
before the continents split up. For example, in the Karoo in southeast Africa,
the major volcanism happened several tens of millions of years before Africa
separated from Antarctic, around 178 million years ago.

The question of whether continental rifting is a necessary part of a
plume reaching the surface, or whether the plume itself causes a continent
to split up is still open. Geologists and geophysicists cite examples from
around the world to support both ideas: rifting without flood basalts, and
flood basalts without continents breaking apart. But whatever the outcome
of this geological chicken and egg argument, Cox and Kent have a vital piece
of the evidence. The topographic traces of ancient plumes, preserved in
the present-day drainage patterns, will help to clarify the history of the
moving continents, as well as provide information about the origin and movement
of plumes.

With his tongue firmly in his cheek, Cox comments: ‘The drainage patterns
certainly seem to suggest that the geophysicists have placed the plumes
in the right place, and got things the right size.’ Perhaps geomorphology,
with its emphasis on processes at the surface of the Earth, will in the
end provide the vital clues needed to understand the deep-seated processes
that shaped our planet.

Nina Morgan is a writer specialising in earth and physical sciences.

* * *

What the geomorphologists think

Geologists studying present day drainage patterns believe they can provide
a quick and easy way to study ancient landscapes, locate ancient mantle
plumes and find sites of continental rifting. But some geomorphologists,
including Mike Summerfield of the Department of Geography at the University
of Edinburgh, who specialises in looking at large-scale landforms, think
this idea is a bit too simple.

In the first place, Summerfield points out that patterns of drainage
directed inland can arise simply as a result of upwarping of continental
margins that haverifted. The edges of continents can rise through various
mechanisms; mantle plumes are not necessarily the cause. Furthermore, because
the domes associated with rising mantle plumes are essentially topographic
features, Summerfield believes that the best way to find out more about
the effects of mantle plumes is to study the topography directly from maps
and digital data.

River drainage patterns can provide helpful additional data, but geomorphol-ogists
do not consider them to be the best source of information about ancient
landscapes. This is because the main features reflected in present day drainage
patterns are recent ones; features of ancient drainage patterns are difficult
to recognise with certainty.

Summerfield thinks that the best way to study the effects of mantle
plumes on the landscape is to try to reconstruct the effects of erosion
on the resulting dome. One way this can be done is to examine the evidence
preserved in ancient sediments, particularly that relating to the sea level
in ancient times. ‘The presence of sediments first deposited at or around
sea level provides a datum to determine absolute changes in surface elevation,’
says Summerfield. For example, if sediments first deposited in a shallow
sea are now high above sea level, the land level must have risen in the
intervening years.

In addition, the orientation and arrangement of the grains in sedimentary
rocks can provide valuable clues about how they were deposited – the direction
in which a river flowed, for example, or whether the river was meandering
on a flood plain or tumbling down a steeper slope. In turn, this information
relates to ancient drainage patterns. Sequences of sedimentary rocks can
include ancient erosion surfaces, which can aid in the understanding of
the form of ancient landscapes.

Another useful technique is fission track analysis. This involves studying
the ‘tracks’ left behind within crystals by alpha particles formed in the
radioactive decay of elements such as uranium. When the mineral grains are
hot, the damage caused by each alpha particle heals relatively quickly.
Only when the rocks are cooler than a certain threshold temperature, which
varies from mineral to mineral, are the tracks preserved.

The radioactive decay that gives rise to the tracks happens at a known
rate for each isotope, so by counting the number of tracks in a given volume
of mineral, geologists can say when the grain last passed its threshold
temperature. And where the main cause of temperature differences is depth
of burial, this method indicates when the rock was last at a certain depth.
By finding a rock containing minerals with threshold temperatures that span
a range of depths, counting fission tracks can give a detailed uplift history
for an area.

Research using fission tracks to study the erosion of the dome in southwest
Africa created by the mantle plume associated with the opening of the South
Atlantic is currently underway in a joint project between the Geography
Department at the University of Edinburgh and the Geology Department at
La Trobe University in Melbourne, Australia.