Geomorphology I: Endogenetic and Exogenetic Forces, Continental Drift, Plate Tectonics and Earth Movements

Alfred Wegener, a German meteorologist, presented continental drift in 1912 and set it out in Die Entstehung der Kontinente und Ozeane (The Origin of Continents and Oceans, 1915). He argued that the continents were once joined in a single supercontinent, Pangaea, surrounded by a universal ocean, Panthalassa, and that Pangaea began to split in the Mesozoic — first into a northern Laurasia and a southern Gondwanaland separated by the Tethys Sea, and then into the present continents. (F.B. Taylor had proposed a similar drift of continents in 1910, to explain the arcs of Tertiary mountains.) Wegener's evidence was of five kinds.

• Jigsaw fit of opposite coasts, above all South America and Africa; Bullard's computer fit of 1965, matched along the continental slope rather than the shoreline, made it far closer.
• Geological matching: rocks of the same age and structure on both sides of the Atlantic, and the Appalachian and Caledonian fold belts continuing across it.
• Fossils: the freshwater reptile Mesosaurus in both Brazil and southern Africa, and the Glossopteris flora across India, Africa, South America, Australia and Antarctica.
• Palaeoclimate: tillites of the Permo-Carboniferous glaciation in now-tropical India, Africa and South America, which make sense only if these lands once lay together near the South Pole.
• Placer deposits: gold in the Ghana coast with no source rock locally, its parent lodes lying in Brazil.

The hypothesis was rejected for three decades because of its mechanism, not its evidence. Wegener proposed a pole-fleeing force (Polflucht) arising from the Earth's rotation and a westward tidal force from the Sun and Moon. Physicists showed that both are orders of magnitude too weak to push continents through the rigid ocean floor. Arthur Holmes (1928–31) proposed thermal convection currents in the mantle as a driver, which came much closer to the eventual answer but lacked evidence at the time.

Post-war mapping of the ocean floor supplied what Wegener lacked. Harry Hess (1962) proposed sea-floor spreading: magma rises at mid-ocean ridges, forms new oceanic crust and moves away on either side, and old crust descends into the mantle at deep-sea trenches. Palaeomagnetism confirmed it. As basalt cools it records the direction of the Earth's magnetic field, which reverses at irregular intervals; Vine and Matthews (1963, and independently Morley) explained the symmetrical magnetic stripes on either side of ridges as a tape recording of spreading. Ocean-floor rocks are youngest at the ridges and grow older towards the trenches, and nowhere is oceanic crust older than about 200 million years (Jurassic). J. Tuzo Wilson (1965) identified transform faults, and between 1967 and 1968 McKenzie and Parker, Morgan and Le Pichon assembled the theory of plate tectonics.

The rigid lithosphere (crust plus uppermost mantle, roughly 100 km thick) is broken into about seven major plates — Pacific, North American, South American, Eurasian, African, Indo-Australian and Antarctic — and many minor ones such as the Nazca, Cocos, Philippine, Caribbean, Arabian and Juan de Fuca plates. They move over the weak, partly molten asthenosphere. The current view of the driving force gives the largest share to slab pull (the weight of cold, dense lithosphere sinking at subduction zones), with ridge push at spreading centres and mantle convection as the deeper engine.

The dipping plane of earthquake foci beneath a subduction zone is the Wadati–Benioff zone. Volcanism far from plate edges is explained by hotspots — Wilson (1963) for the Hawaiian chain, W.J. Morgan (1971) for deep mantle plumes: a plate moving over a fixed plume leaves a line of volcanoes that grow older away from the active one. The Deccan Traps (about 66 million years ago) are linked to the plume now under Réunion. The Indian plate separated from Madagascar in the Late Cretaceous, travelled north unusually fast, and began colliding with Eurasia around 50 million years ago; it still converges at about 4–5 cm a year, which is why the Himalaya is seismically active.

Folding is the bending of strata under compression. The up-arched limb is an anticline and the down-folded trough a syncline; as compression increases, folds become asymmetrical, overturned, recumbent (lying on their side) and, when a recumbent fold is sheared and carried forward over a thrust plane, a nappe — the structure of the Alps. A monocline has one steep limb; in an isoclinal fold both limbs dip at the same angle. A large anticline carrying many minor folds is an anticlinorium, and its counterpart a synclinorium. Faulting is fracture with displacement. Tension produces normal faults; compression produces reverse and, where the fault plane is low-angled, thrust faults; horizontal shear produces strike-slip (tear) faults. A block let down between two normal faults is a graben or rift valley, and a block left standing between them is a horst or block mountain — the Rhine rift between the Vosges and the Black Forest is the classic pair, and the Narmada flows westward through a rift between the Vindhyas and the Satpuras.

An earthquake is the sudden release of strain along a fault. It begins at the focus (hypocentre); the point on the surface above it is the epicentre. P waves are longitudinal, the fastest, and travel through solids, liquids and gases; S waves are transverse and cannot pass through liquids; surface waves (Love and Rayleigh) arrive last and do most damage. Because S waves do not cross the liquid outer core, there is an S-wave shadow zone beyond about 103° from the epicentre, and refraction of P waves at the core boundary creates a P-wave shadow zone between about 103° and 142°. Size is measured two ways: magnitude (Richter, 1935, now usually moment magnitude Mw) is a logarithmic measure of energy released — each whole number is a tenfold increase in recorded amplitude and roughly 32 times the energy — while intensity (Modified Mercalli scale, I to XII) describes observed effects at a place. About four-fifths of the largest earthquakes occur around the circum-Pacific belt, and most of the rest along the Alpine–Himalayan belt, with a third, shallow belt along the mid-ocean ridges.