4.5 — The Earth, Its Weather and Its Rocks

The ground you are standing on is moving, the weather is one machine running on unequal heating, and most of the questions people ask about both have answers that are simpler than expected and completely different from the popular one.

The moving ground

What is actually happening in an earthquake?

Two slabs of rock that have been stuck against each other finally slip.

The Earth's outer shell is broken into about fifteen major plates that move a few centimetres a year — roughly the rate your fingernails grow. Where two plates meet they do not slide smoothly, because rock under pressure has enormous friction. The plates keep moving, the rock at the boundary deforms elastically, and stress accumulates for decades or centuries.

When the friction is finally exceeded, the stored elastic energy releases in seconds. The rock snaps back, and the ground on either side of the fault can jump several metres. That release radiates as waves.

The waves come in a fixed order and the order is useful. P waves (primary) are compressions travelling at 6 to 8 kilometres per second — they arrive first, and they are the sharp jolt. S waves (secondary) shear the rock sideways at about half that speed, and they do most of the shaking. Surface waves arrive last, travel along the ground rather than through it, and cause most of the destruction.

That delay is what early warning systems exploit. A detector near the epicentre catches the P wave, computes the size, and transmits a warning electronically — which travels at light speed and therefore overtakes the damaging S wave. A city 100 kilometres away can get 20 to 30 seconds of notice: enough to stop trains, halt lifts at the nearest floor, and close gas valves.

The magnitude scale is logarithmic, and the two steps involved are worth separating. Each whole number is about 32 times more energy released, and about 10 times more ground motion. So a magnitude 7 releases roughly a thousand times the energy of a magnitude 5. The Richter scale is obsolete for large events and has been replaced by the moment magnitude scale, which is computed from the fault area, the slip distance and the rock's rigidity — the physical size of the event rather than a needle deflection.

Why does India have earthquakes in the Himalayas?

Because the Indian plate is still driving into Asia, and has been for about 50 million years.

India was an island continent that broke off Gondwana and moved north unusually fast — up to 15 centimetres a year, which is quick for a continent. It collided with Asia, and because both are continental crust and neither will sink under the other, the collision zone crumpled upward instead. That is the Himalayas, and the range is still rising by roughly 5 millimetres a year while erosion removes a comparable amount.

The convergence continues at about 4 to 5 centimetres a year, and that motion is absorbed along faults through the Himalayan front. The stress accumulates and releases in large earthquakes — 1934 in Bihar-Nepal, 1950 in Assam, 2005 in Kashmir, 2015 in Nepal. Seismologists regard several segments of the Himalayan front as overdue, in the sense that measured convergence has not been matched by measured slip.

How do we know what is inside the Earth if nobody has been there?

The deepest hole ever drilled, the Kola Superdeep Borehole in Russia, reached about 12.3 kilometres — around 0.2 per cent of the way to the centre.

Everything else is known from earthquake waves, and the method is genuinely elegant. Seismometers all over the world record every large earthquake, and the arrival times and paths of the waves depend on what they travelled through. Waves bend as they cross into material of different density, and the pattern of bending maps the layers.

The decisive observation is a gap. S waves cannot travel through liquid, because a liquid has no resistance to shearing. After a large earthquake there is a zone on the far side of the planet where no S waves arrive at all — the shadow zone. That is direct evidence that the outer core is molten, and its size gives the core's dimensions.

The layers so derived: a thin brittle crust, a hot solid but slowly flowing mantle down to about 2,900 km, a liquid iron-nickel outer core, and a solid inner core about 1,220 km in radius — solid despite being hotter, because the pressure there is over three million atmospheres.

Where does a magnetic compass point, and why does it flip?

To the magnetic pole, which is not the geographic pole and moves.

The field is generated by convection in the liquid outer core: molten iron circulating and carrying electric current, sustaining a magnetic field, in a self-reinforcing arrangement called a geodynamo. It is a fluid system, and it wanders. The magnetic north pole has moved from northern Canada towards Siberia and has been accelerating, moving over 50 kilometres a year in recent decades.

The field also reverses completely, on an irregular schedule averaging a few hundred thousand years, most recently about 780,000 years ago. We know because lava cooling on the seafloor locks in the field direction of the moment, and the seafloor spreading away from mid-ocean ridges records the history as symmetric magnetic stripes on both sides. That discovery in the 1960s is what turned continental drift from a suggestion into plate tectonics.

Weather

Why do we have seasons?

Not because of distance from the sun. The Earth is actually closest to the sun in early January, during the northern winter.

Seasons come from the 23.4-degree tilt of the Earth's axis, which stays pointed in the same direction as the planet goes round the sun. For half the year the northern hemisphere leans towards the sun and for the other half away.

The tilt does two things at once, and both matter. The sun is higher in the sky, so the same beam of sunlight is spread over a smaller patch of ground and delivers more energy per square metre — the same reason a torch shone straight down makes a brighter spot than one shone at a slant. And the days are longer, so the heating runs for more hours.

Southern hemisphere seasons are reversed for the obvious reason, which is why Christmas in Australia is a beach event.

The hottest part of summer arrives well after the longest day, and the coldest part of winter after the shortest, because land and ocean take weeks to warm and cool. That lag is called the seasonal lag and it is why the summer solstice in June is not the hottest day of the year anywhere.

Why do cyclones spin, and in opposite directions?

Because air flowing towards a low-pressure centre is deflected by the Earth's rotation.

Air moves from high pressure to low. If the Earth were still, it would flow straight in and fill the low. But the ground beneath a moving parcel of air is itself rotating, and at different speeds at different latitudes, so from the point of view of somebody standing on the ground the air appears to curve. That apparent deflection is the Coriolis effect.

The deflection is to the right in the northern hemisphere and to the left in the southern. Air spiralling into a low therefore rotates anticlockwise in the north and clockwise in the south — which is why a satellite image is enough to tell you which hemisphere you are looking at.

Two consequences follow. Cyclones cannot form within about 5 degrees of the equator, because the Coriolis effect goes to zero there and nothing can start the rotation. And the effect is far too weak to influence a bathtub or a toilet, where the geometry of the basin and the way the water was poured dominate completely. The draining-sink claim is false, and the demonstrations sold to tourists on the equator are done by hand.

What is the monsoon, actually?

A seasonal reversal of wind direction driven by the different rates at which land and sea change temperature.

Land heats up much faster than water in summer, because water mixes and has a large capacity for absorbing heat with little temperature change. So by May the Indian landmass and the Tibetan plateau are far hotter than the Indian Ocean. Hot land heats the air above it, that air rises, and lower pressure forms over the land. Moist air over the ocean flows in to replace it, deflected by the Coriolis effect into the familiar south-westerly flow, and dumps its water as it rises over the land and the mountains.

In winter the reverse: the land cools faster than the sea, pressure over land rises, and the wind blows offshore, dry.

The Himalayas are essential rather than incidental. They block cold central Asian air from entering in summer, and they force the moist air upward, which is why the wettest places on Earth — Mawsynram and Cherrapunji in Meghalaya — sit on a slope facing the incoming wind.

The monsoon supplies around 70 to 80 per cent of India's annual rainfall in about four months, which is why its timing and total are followed as economic news rather than weather.

Why is a raindrop not shaped like a teardrop?

Because nothing gives it that shape. A falling raindrop is round, or a flattened round.

A very small drop is spherical, because surface tension pulls it into the shape with the least surface area for its volume. As drops get larger, air resistance pushes on the underside and flattens it, so a large falling drop looks more like a bun or a hamburger than a tear.

Above about 4 millimetres across, the flattening becomes unstable, the bottom hollows into a parachute shape, and the drop breaks into smaller ones. That sets an upper limit on raindrop size, which is why rain does not arrive in buckets however hard it is falling.

The teardrop shape comes from drops hanging and detaching from a surface — a tap, or a leaf — where the neck stretches before it lets go.

Why is snow white when ice is clear?

Same reason sand is not transparent, in 1.2.

A snowflake is transparent ice. A snowdrift is billions of ice crystals with air between them, and every surface between ice and air scatters light. Light entering the drift bounces from surface to surface and comes back out in every direction, and since ice scatters all visible wavelengths about equally, the mixture that comes out is white.

Deep snow and glacial ice look blue for a different reason: over several metres, water ice absorbs red light slightly more than blue, so what emerges from a deep hole is blue-shifted. It is the same absorption that makes a swimming pool blue.

Rocks and time

How are diamonds actually made, and are they rare?

Carbon crystallised under pressures above about 45,000 atmospheres, at 900 to 1,300 °C, at depths of 150 kilometres or more in the mantle. Most gem diamonds are between one and three billion years old.

They reach the surface by an unusual and violent route. A rare kind of eruption, from a magma called kimberlite, travels up from mantle depth at high speed — fast enough that the diamonds do not have time to convert to graphite on the way. The eruption leaves a carrot-shaped pipe, and diamond mining is essentially the mining of those pipes.

Diamonds are not particularly rare as gemstones go — ruby, emerald and several others are scarcer. The price was maintained through most of the twentieth century by tightly controlled supply, principally by De Beers, which at its peak handled the large majority of the world's rough diamonds and released them at a managed rate. The idea that a diamond is the correct engagement gift, and that it should cost a set fraction of your salary, comes from an advertising campaign begun in 1938 — the slogan "A Diamond Is Forever" was written in 1947 by a copywriter named Frances Gerety.

Laboratory-grown diamonds are now chemically and optically identical, distinguishable only by trace features, and cost a fraction as much.

Why is soil brown and how long does it take to make?

The brown is mostly humus, decomposed organic matter, along with iron oxides from weathered rock. Colour reads as information: dark brown or black means high organic content and generally fertile; red means well-drained and iron-rich; grey or blue-grey means waterlogged, where the lack of oxygen leaves iron in a different chemical state.

Soil forms from rock broken down physically by frost and roots and chemically by acidic water, then colonised by organisms whose remains supply the organic fraction. The rate is slow enough to matter: something in the region of 100 to 1,000 years to form a centimetre of topsoil, depending heavily on climate and parent rock.

Cultivated land in many parts of the world loses topsoil to erosion far faster than that. Which is the whole reason soil conservation is treated as an agricultural emergency rather than a gardening topic.

What is the difference between weather and climate, in one usable sentence?

Weather is what the atmosphere is doing; climate is the statistics of what it usually does.

The useful analogy is that weather is a single roll of a die and climate is the distribution. You cannot predict one roll beyond a few days — the atmosphere is chaotic, and small errors in the starting conditions grow exponentially, which is exactly what Edward Lorenz discovered in 1961 when a rounded-off number produced a completely different forecast.

But you can say a great deal about the distribution. Loading the die changes the frequency of outcomes without letting you predict any single roll, which is why a cold winter is not evidence against a warming climate, and why the meaningful climate statements are about shifting probabilities of extremes rather than about individual events.

What comes next

The last page of this Part is about the units themselves — where the metre came from, why a kilogram was a lump of metal in a vault in Paris until 2019, what a decibel measures, and why a "1 TB" hard disk shows up as less than a terabyte.