Earth processes and mass movements

The layered Earth, the lithosphere and the asthenosphere

You only need an outline of the Earth's interior, but you need the words exactly. At the centre is the inner core, solid iron and nickel held solid by immense pressure; around it the outer core is liquid iron and nickel, and its motion generates the magnetic field. Above that lies the mantle, roughly 2 900 km of hot silicate rock that is solid but deforms and flows extremely slowly, and at the surface the crust, a thin brittle shell.

The crust comes in two kinds. Continental crust is 30 to 70 km thick and made mainly of granite and older metamorphic rocks, so it is relatively light. Oceanic crust is only 5 to 10 km thick and made of basalt and gabbro, so it is denser. Both are cold and brittle, and both sit on the same rigid lid.

That rigid lid is the lithosphere: the crust plus the uppermost part of the mantle, welded to it and behaving as one brittle layer about 100 km thick (thicker under ancient continents). Below it is the asthenosphere, which is hot, close to its melting point and contains a few per cent melt, so it is weak and deforms plastically over long periods. Seismic waves slow down as they pass through it, which is how we know it is there. The plates are slabs of lithosphere, and they slide over the asthenosphere - that is what makes plate movement possible at all.

• Inner core: solid iron and nickel, kept solid by pressure. Outer core: liquid iron and nickel; its flow generates the magnetic field.
• Mantle: about 2 900 km of hot rock that flows very slowly; the source of the magma that reaches the surface.
• Crust: brittle outer shell - 30-70 km of light granite-rich rock under the continents, 5-10 km of dense basalt under the oceans.
• Lithosphere: crust plus rigid uppermost mantle, about 100 km thick; broken into plates.
• Asthenosphere: hot, weak, partly molten layer beneath the lithosphere; the plates move over it.
• Continental plate: thick, low density (about 2.7 g/cm3), old - rocks up to nearly 4 billion years - and buoyant, so it is not easily subducted.
• Oceanic plate: thin, denser (about 3.0 g/cm3), young - nowhere more than about 200 million years old because it is continually created and destroyed - and it subducts readily.

Continental drift and the evidence in the rocks and fossils

In 1912 Alfred Wegener proposed the theory of continental drift: that the continents had once formed a single landmass, which he called Pangaea, surrounded by one ocean. Pangaea began to break up about 250 million years ago into a northern mass, Laurasia, and a southern mass, Gondwanaland, and the fragments then drifted to their present positions. Wegener had no convincing mechanism, which is why the theory was rejected for fifty years, but the evidence he assembled was good and it is what the syllabus asks you to know.

The jigsaw fit is the starting point. The eastern coast of South America and the western coast of Africa fit together remarkably closely, particularly when the fit is made along the edge of the continental shelf rather than the present shoreline. India, Madagascar, Antarctica and Australia slot into the same reconstruction, and the fit is too good to be coincidence.

The geological record supports it. Rock sequences of the same age and type, and fold mountain belts of the same age, appear on opposite sides of the Atlantic: the Appalachian belt of eastern North America continues into the Caledonian belt of Scotland, Norway and Greenland. Coal seams of the same age lie under the Antarctic ice, and glacial deposits are found in present-day India, Brazil and central Africa - all of which are now tropical or subtropical.

The fossil record is the strongest single line of evidence. Glossopteris, a seed fern with heavy seeds that could not have crossed an ocean, is found in South America, Africa, India, Australia and Antarctica. Mesosaurus, a small freshwater reptile, is found only in South America and southern Africa, and Lystrosaurus, a land reptile, in Africa, India and Antarctica. Animals that could not swim an ocean and plants that could not cross one must have lived on land that was once joined.

• Jigsaw fit: South America and Africa match along the edge of the continental shelf, and India, Madagascar, Antarctica and Australia fit the same reconstruction.
• Geological record: matching rock sequences and fold belts of the same age on opposite sides of the Atlantic; Appalachian and Caledonian belts are one mountain chain; coal under Antarctic ice; glacial deposits in now-tropical latitudes.
• Fossil record: Glossopteris (a seed fern), Mesosaurus (a freshwater reptile) and Lystrosaurus (a land reptile) occur on continents now separated by oceans.
• The weakness in the theory: Wegener could not explain what moved the continents, and his suggestion that they ploughed through the ocean floor was shown to be physically impossible.
• The resolution: sea floor spreading, proved from palaeomagnetism and the age of ocean rocks in the 1960s, supplied both the mechanism and the proof that continents move with the plates.

Palaeomagnetism, the age of the ocean floor and the forces that move plates

Palaeomagnetism is the record of the Earth's magnetic field locked into rocks. When basaltic lava cools below about 580 °C, iron-rich minerals such as magnetite align themselves with the magnetic field as it is at that moment, and once the rock has solidified the record is frozen. The field is not constant: it reverses polarity every few hundred thousand years, so rocks record either normal polarity, as now, or reversed polarity.

In the 1950s and 1960s magnetometers towed behind research ships across the ocean floor found long magnetic stripes of alternating normal and reversed polarity running parallel to the mid-ocean ridges, and arranged in mirror-image patterns on either side of the ridge axis. The explanation, proposed by Vine and Matthews in 1963, is that new basalt erupts at the ridge, cools and records the field of the day, and is then carried away sideways in both directions as spreading continues. Each new band of crust inherits the polarity of its own moment of formation, so the pattern repeats symmetrically.

The age of the sea floor rocks confirms it. Samples dated from deep-sea drilling show that the crust gets steadily older with distance from the ridge, in exactly the same pattern as the stripes, and that the sediment cover gets thicker away from the ridge because it has had longer to accumulate. The oldest ocean floor is about 200 million years old, while continental rocks reach nearly 4 billion years, which makes sense only if ocean crust is continually created at ridges and destroyed at trenches.

Matching the stripes against the dated reversal record also gives the spreading rate: usually 1 to 10 cm a year, and up to about 15 cm a year on parts of the East Pacific Rise. Satellite GPS now measures plate motion directly, confirming both the direction and the speed.

Three forces are named in the syllabus. Convection currents in the mantle, driven by heat from the core and from radioactive decay, circulate slowly and drag the base of the plates. Ridge push is the gravitational sliding of the lithosphere away from the elevated mid-ocean ridge, where the new crust is hot, buoyant and stands high.

Slab pull is the downward pull of a cold, dense, subducting slab on the rest of its plate. Slab pull is now regarded as the strongest of the three, which fits the observation that the fastest-moving plates are the ones with long subducting edges, such as the Pacific and Nazca plates.

The global distribution of the four plate boundaries

A plate boundary is where two plates meet, and the syllabus names four types, defined by the relative direction of movement. Almost every landform and every tectonic hazard in this topic can be predicted from which type of boundary you are looking at.

Divergent (constructive) boundaries are where plates move apart and new crust is created. They occur in the middle of oceans, as along the Mid-Atlantic Ridge and the East Pacific Rise, and on land where a continent is splitting, as in the East African Rift. Convergent (destructive) boundaries are where an oceanic plate sinks beneath another plate, producing the deep trenches and volcanic arcs that ring the Pacific - the Japan Trench, the Mariana Trench, the Peru-Chile Trench - and the Lesser Antilles in the Caribbean.

Convergent (collision) boundaries are where two continental plates meet and neither will sink. They form the Alpine-Himalayan belt, running from the Pyrenees and the Alps through Turkey and Iran to the Himalayas and on into Southeast Asia. Conservative (transform) boundaries are where plates slide past each other, and because they neither create nor destroy crust they produce no volcanoes at all: the San Andreas Fault in California, the North Anatolian Fault in Turkey and the Alpine Fault in New Zealand are the standard examples.

Put the four types on a world map and the pattern of landforms falls out of it: trenches and island arcs outline the Pacific, a ridge runs the length of the Atlantic, fold mountains and destructive earthquakes trace the Alpine-Himalayan belt from the Mediterranean to Indonesia, and long faults such as the San Andreas carry earthquakes with no volcanism at all.

Tectonic processes: faulting, rifting, spreading, subduction and thrusting

Rock at a plate boundary is being pushed or pulled, and it responds in one of three ways. Tension pulls it apart and produces normal faults, where the hanging wall slides down. Compression squeezes it and produces reverse or thrust faults, where the hanging wall is pushed up and over the footwall. Shear makes the two sides slide past one another and produces strike-slip faults, where movement is horizontal.

Recognising the stress tells you the boundary type, and the boundary type tells you the landform.

At a divergent boundary the lithosphere is stretched and thinned. Normal faults develop in parallel sets, and the block between two faults drops down relative to its neighbours, producing a rift valley with steep fault scarps on either side. The East African Rift is the classic example on land, and the same structure exists underwater as the axial rift along the crest of the Mid-Atlantic Ridge. As the crust thins, the pressure on the hot mantle beneath falls, which allows it to melt without any added heat - decompression melting - and the basalt that erupts builds volcanoes and long fissures along the rift floor.

Sea floor spreading is what happens next. New basaltic crust is added at the ridge axis a few centimetres at a time, and because it is hot and therefore less dense, it stands higher than the older crust further away. As it moves away from the ridge it cools, contracts, becomes denser and subsides, which is why the ridge is a broad swell and why the ocean gets deeper towards its edges. The ridge crest is cut by a rift valley and is offset at intervals by transform faults, which are the conservative segments that link one spreading ridge to the next.

At a destructive boundary the oceanic plate is old, cold and dense, so it sinks. Its bending creates a narrow, very deep ocean trench on the seaward side of the boundary, and as the slab descends it releases water and other volatiles that lower the melting point of the mantle above it, producing magma that rises to build a line of volcanoes set back from the trench. Sediment and fragments of ocean crust scraped off the descending plate pile up against the margin as an accretionary wedge. Earthquakes occur along the whole length of the descending slab, from the shallow thrust faults at the trench to depths of several hundred kilometres where the slab is still cold enough to be brittle.

At a collision boundary the same compression operates, but the crust on both sides is continental and buoyant, so instead of subducting it is folded and stacked. The dominant structures are thrust faults, along which great sheets of rock are pushed over one another, and tight folds; the crust is thickened to as much as 70 km beneath Tibet, which is why the Himalayas stand so high. Movement is not purely head-on: large strike-slip faults, such as the Karakoram Fault, allow rock to escape sideways as India pushes north. At a conservative boundary there is only shearing, so the faulting is strike-slip, the plates lock and release in shallow earthquakes, and no magma is produced anywhere along it.

• Tension -> normal faults (hanging wall down) -> rift valleys and fault scarps at divergent boundaries.
• Compression -> reverse and thrust faults (hanging wall up) -> fold mountains and crustal thickening at collision and destructive boundaries.
• Shear -> strike-slip faults (horizontal movement) -> offset streams, fences and roads at conservative boundaries.
• Sea floor spreading: new basalt at the ridge axis, cooling and subsiding away from it, offset by transform faults.
• Subduction: trench, accretionary wedge, water-driven melting of the mantle wedge, volcanic arc, earthquakes down the slab.
• Rifting: parallel normal faults, a dropped central block, thinning crust and decompression melting along the rift floor.

The landforms made at plate boundaries

Fold mountains are formed where a thick pile of sedimentary rock is compressed. The layers buckle into upfolds (anticlines) and downfolds (synclines), and where the compression continues the folds tighten, overturn and are cut by thrust faults that stack one sheet of rock on another. Collision produces the highest range on Earth, the Himalayas, where the Indo-Australian plate meets the Eurasian plate; destructive boundaries produce fold mountains too, as along the Andes, but there the range also carries a chain of volcanoes because magma is being generated beneath it.

Volcanoes are the surface expression of magma reaching the top of the lithosphere. At a destructive boundary the magma is produced by melting of the mantle wedge above the slab; it is sticky and gas-rich, so eruptions are explosive and the cone is built of alternating ash and lava. At a divergent boundary the magma is basaltic and runny, so eruptions are quieter and the landform is a broad shield or a line of fissures along the rift. In both cases the volcano is a constructional landform: it is built up by what comes out of it.

Mid-ocean ridges are the longest mountain chains on Earth, although almost all of the ridge lies underwater. They stand 2 to 3 km above the surrounding ocean floor and are hundreds of kilometres wide, with a rift valley along the crest where the plates are pulling apart; the Mid-Atlantic Ridge reaches the surface in Iceland, which is why the island is volcanically active. Ocean trenches are the opposite: narrow, steep-sided troughs formed where a slab bends downwards, and they include the deepest points on the planet, such as the Mariana Trench in the western Pacific at about 11 km below sea level.

Volcanic island arcs form where an oceanic plate is subducted beneath another oceanic plate. The magma produced above the slab builds a curved chain of volcanic islands parallel to the trench and set back from it, such as the Aleutian Islands, the Mariana Islands, the Lesser Antilles and Japan. Fault scarps are the small-scale landform of faulting: a step or cliff in the ground where one side of a fault has moved up or down relative to the other, fresh and sharp when the fault has just moved and rounded and degraded with time. Rift valleys are the large-scale version, a long trough bounded by normal faults, with a flat or lake-strewn floor: the East African Rift, whose floor holds lakes such as Tanganyika and Malawi, and the Rhine Rift Valley in Europe.

Mass movement: heaves, flows, slides and falls

Mass movement is the downslope transfer of rock, soil and debris under the influence of gravity, without the material being carried by a river, a glacier or the wind. It happens when the shear stress acting on a slope - the weight of the material pulling it downslope - exceeds the shear strength of the material holding it in place. Anything that adds weight, steepens the slope or reduces strength moves a slope closer to failure. Mass movements are classified by the type of movement, by how fast it happens and by how much water is involved.

Heaves are the slowest and least dramatic. In frost heave, water in the soil freezes, expands and lifts particles at right angles to the slope surface; when it thaws the particles fall back vertically, so each freeze-thaw cycle moves them a little further downslope, producing the slow downslope creep of the whole soil layer. Wetting and drying does the same thing through expansion and contraction. The movement is measured in millimetres or centimetres a year, but over centuries it produces distinctive small steps and terraces on grassy slopes, and it gradually tilts and buries fences, walls and posts.

Flows are movements in which the material deforms and moves like a fluid. A mudflow is dominated by fine material - silt and clay - with a high water content, and it usually travels down a channel as a thick, fast slurry that can move faster than a person can run. A debris flow is a mixture of coarse rock, gravel, soil and water that can carry boulders several metres across; it is typically triggered by intense rainfall on a steep, loose slope, often after a wildfire or deforestation has removed the vegetation, and it spreads out as a lobe where the valley floor flattens.

Slides are movements in which a mass of material moves along a recognisable failure surface. A translational landslide slides along a plane - a bedding plane, a joint, a fault or the boundary with the underlying rock - and the moving mass stays largely intact. A rotational slide, or slump, fails along a curved surface: the mass rotates backwards as it moves, leaving a steep concave scar at the top and a bulging toe at the bottom, with the slipped blocks tilted back towards the slope. Slumps are common in thick clay and in artificial embankments.

Falls are the free fall of material from a steep face. Rockfalls occur where a cliff is undercut, where joints have been opened by freeze-thaw or unloading, or where an earthquake shakes blocks loose; the fragments fall, bounce and accumulate at the base as a talus or scree slope. Rockfalls are the fastest mass movement, they give no warning, and they are the reason rock faces beside roads and railways are netted and bolted.

Physical and human causes of mass movement

Geology sets the baseline. Weak rocks such as clay, shale and mudstone have low shear strength, especially when saturated. A permeable layer such as sandstone or chalk resting on an impermeable clay collects water at the junction, so the clay beneath becomes saturated and slippery. Joints, fissures and bedding planes give water a route in and give the rock a surface to slide along, and where bedding or a fault dips in the same direction as the slope, the whole hillside is predisposed to fail.

Topography decides how much stress the material has to bear. The steeper the slope, the greater the downslope component of gravity; a concave profile concentrates water at the base, and a slope that has been undercut by a river, a wave or a glacier has lost the support at its toe that was holding it up. Precipitation is the most common trigger: it is the intensity of the rain rather than the annual total that matters, because a short violent storm delivers water faster than the soil can drain it. The water adds weight, lubricates the failure surface and, most importantly, raises pore water pressure, which pushes the grains apart and reduces the effective stress holding them together.

Climate change is altering these controls. More intense convective rainfall and longer dry spells followed by heavy storms increase the frequency of flows and slides; the thaw of permafrost removes the ice that cemented high mountain rock, and the retreat of glaciers removes the buttress that supported valley sides, so both increase rockfalls and rock avalanches in mountain regions; rising sea level and more energetic storms accelerate coastal cliff erosion and undercutting. The syllabus asks about the potential impact, so describe the mechanism rather than claiming a precise increase in events.

Human causes work by doing the same things faster. Building and road construction cut into the toe of a slope or load the crest with fill, buildings and traffic, and blasting sends vibrations through the rock; a road cutting through a hillside is one of the most reliable ways to trigger a slide. Land-use change removes deep-rooted vegetation that would otherwise bind the soil and take up water: deforestation, ploughing downslope, terracing without drainage, and the spread of informal settlements onto steep ground all reduce stability. Water management is the third: leaking pipes, sewers and soakaways, irrigation of fields on a slope, blocked or badly designed drains discharging onto a hillside, and the rapid drawdown of a reservoir all raise pore water pressure or add water where it does not belong.

It is worth separating a cause from a trigger. The cause is what made the slope vulnerable over years - a clay layer, a steep cut, deforestation, a leaking main. The trigger is the event that finally pushed it over - a storm, an earthquake, a thaw. Examiners reward answers that name both and link them.

Managing mass movement hazards: prediction, impacts, slope stabilisation and evaluation

Prediction and risk identification come first, because a hazard that is mapped can be avoided. Hazard maps combine geology, slope angle, past events, drainage and land use in a GIS to zone the land from stable to highly unstable, and they are used to steer development, to plan evacuation and to prioritise spending. Monitoring looks for the signs of movement: crack meters and extensometers across cracks, inclinometers in boreholes to detect shear at depth, piezometers to measure pore water pressure, survey marks, GPS and satellite InSAR to detect millimetre-scale ground movement, and rain gauges to compare rainfall against a known threshold. The weakness is timing: monitoring tells you a slope is moving, not when it will fail, and instruments only cover the slopes someone has decided to instrument.

The impacts on people are deaths and injuries, lost homes, farms, roads, railways, pipelines and water supplies, communities cut off, and the cost of evacuation and rebuilding. The impacts on the environment are lost soil and vegetation, damaged habitats, sediment choking rivers and reservoirs, and rivers temporarily dammed by a landslide, which can then break and flood downstream.

Slope drainage is usually the most effective single measure, because water is the commonest trigger: surface drains and cut-off ditches keep running water off the slope, deep horizontal drains and drainage galleries lower the water table inside it, and sealing cracks stops water getting in. Slope grading reshapes the slope - reducing its angle, benching it into steps, or removing material from the crest to unload it and placing it at the toe as a counterweight. Pinning uses rock bolts, soil nails and ground anchors to tie the unstable mass back into solid rock, and netting - draped mesh or bolted mesh - holds a rock face together and catches falling blocks before they reach the road or the houses below.

Afforestation increases strength with deep roots, intercepts rainfall and takes up water, and it is cheap and environmentally attractive, but it takes years to work and cannot hold a deep-seated slide. Planning restrictions - refusing permission to build on the most unstable ground, requiring engineering reports, and keeping housing off the valley floors where flows run out - are the cheapest measure of all and the only one that prevents new risk, but they are hard to enforce where land is scarce and people are poor. Emergency responses include evacuation, search and rescue, closing roads and railways, and emergency drainage and netting; recovery covers rehousing, compensation, rebuilding to better standards and repairing infrastructure.

Evaluating the strategies means weighing cost against effect and asking who is protected. Hard engineering - pinning, netting, drainage and grading - works immediately, protects specific assets such as a road or a power station, and can be designed for a known slope, but it is expensive, needs maintenance and usually protects only the site it was built for. Soft measures - afforestation and planning restrictions - are cheap and address the cause, but they act slowly and depend on enforcement. Monitoring and hazard mapping do not stop anything, but they are what allow people to be moved in time, and they are the only practical option for very large, deep-seated slides.

The defensible judgement is that the best package is drainage plus planning control plus monitoring, and that the weakest link is usually enforcement: a hazard map that is not acted on protects nobody.