Appearance
2.2 — Why the Wind Blows
Look at a world map of deserts. The Sahara, the Arabian, the Thar, the Iranian, the Sonoran and Chihuahuan in Mexico, the Mojave. Now the southern hemisphere: the Kalahari and Namib, the Atacama, the Australian interior.
Nearly every one sits between 20 and 35 degrees of latitude, north or south. Different continents, different rocks, different histories, same band. That is not a coincidence and it is not about distance from the sea. It is a direct consequence of the fact that the Earth is a sphere, that it spins, and that the Sun heats the equator harder than the poles.
This page derives that band from those three facts, and gets the trade winds, the westerlies, the jet stream and the doldrums along the way.
Step one: heating is uneven, and that makes pressure differences
Sunlight hits the equator nearly straight on and the poles at a glancing angle. The same beam of sunlight spread over a slanted surface covers more area, so each square metre gets less. The beam also passes through more atmosphere on the way. Result: the tropics receive far more energy per square metre than the polar regions.
More heating means the air expands and rises. Rising air removes weight from the column beneath it, so the surface pressure falls. Somewhere else, air descends and piles up, and surface pressure rises.
Air then flows from high pressure to low pressure, because a pressure difference is a force. That is the whole of wind, before rotation gets involved. The steeper the pressure change over distance — the pressure gradient — the stronger the wind. On a weather map, isobars close together mean it is blowing hard.
The simple picture you would predict is therefore one giant loop per hemisphere: air rises at the equator, travels aloft to the pole, sinks there, and returns along the surface. That was proposed by Edmond Halley in 1686, and it is wrong. The Earth spins, and that breaks the loop into three.
Step two: the Coriolis effect, done properly
This is the most misexplained idea in geography, so here it is from the beginning.
The Earth is a sphere turning once a day. A point on the equator is carried eastward at about 1,670 km/h. A point at 45 degrees latitude is carried at about 1,180 km/h, because it is on a smaller circle. A point at the pole does not move sideways at all; it just pivots.
Now send a parcel of air from the equator toward the north pole. It leaves carrying eastward motion of 1,670 km/h. As it moves north, the ground beneath it is moving east more slowly. The air keeps its eastward speed, so relative to the ground it drifts east — it appears to be deflected to the right.
Send a parcel from the pole toward the equator. It carries almost no eastward motion, arriving over ground that is racing east beneath it. Relative to the ground it falls behind — it appears to be deflected to the west, which from the point of view of something travelling south is again to the right.
In the northern hemisphere every freely moving thing is deflected to the right of its direction of travel. In the southern hemisphere, to the left. The deflection is zero at the equator and strongest at the poles.
Three things this is not. It is not a force — nothing pushes the air; the ground turns underneath it, and the apparent curve is a consequence of describing motion from a spinning platform. It does not affect water going down a drain, where the direction is set by the basin's shape and by however the water was moving when the plug came out; the effect is far too weak at that scale. And it does not act on things over short distances or short times — a thrown ball or a fired arrow is unaffected in any way you could notice, although long-range artillery does correct for it.
The scale where it matters is hundreds of kilometres and many hours, which is exactly the scale of weather systems and ocean currents.
Step three: why the single loop breaks into three
Halley's single loop cannot survive because air moving polewards aloft is deflected further and further to the right, until by about 30 degrees latitude it is moving almost due east and has stopped going north at all. It has to come down. So the tropical cell closes there.
That gives three cells per hemisphere, and each has a name and consequences.
The Hadley cell — 0 to about 30 degrees. Air rises at the equator, moves poleward at height, and sinks around 30 degrees. Named for George Hadley, who added rotation to Halley's picture in 1735.
The Ferrel cell — about 30 to 60 degrees. An indirect cell, driven mechanically by the two cells on either side rather than by heating of its own. Surface air moves poleward here.
The polar cell — 60 degrees to the pole. Cold dense air sinks at the pole and flows outward at the surface.
The rising and sinking branches decide where the rain is, and this is the payoff.
Rising air cools, and cool air cannot hold as much water vapour, so it rains. That is why the equatorial belt has the world's rainforests — the Amazon, the Congo, Indonesia — and why sailors called that belt the doldrums: air is going up, not sideways, so there is little surface wind and sailing ships could sit becalmed for weeks.
Sinking air warms as it is compressed, and warm air can hold more vapour, so it does not rain — it dries the ground instead. That is the desert belt at 30 degrees, on every continent, in both hemispheres. The Sahara, Arabian, Thar, Kalahari, Atacama and Australian deserts are the descending branch of the Hadley cell made visible. Sailors called these latitudes the horse latitudes, and the ships becalmed there reportedly threw horses overboard when water ran short.
At about 60 degrees, cold polar air meets warmer mid-latitude air and the warm air is forced up over it, which is why northern Europe, the northwestern United States and southern Chile are wet, and why the boundary — the polar front — is where mid-latitude storms are born.
The surface wind belts, and why exploration went the way it did
Apply the deflection rule to each surface flow and the world's named winds fall out.
| Belt | Latitude | Surface flow | Deflected to | Called |
|---|---|---|---|---|
| Tropics | 0–30° N | Toward equator (southward) | Right, so from the northeast | Northeast trade winds |
| Tropics | 0–30° S | Toward equator (northward) | Left, so from the southeast | Southeast trade winds |
| Mid-latitudes | 30–60° | Poleward | Right (N) / left (S), so from the west | Westerlies |
| Polar | 60–90° | Toward equator | Right / left, so from the east | Polar easterlies |
"Trade winds" does not come from commerce. It comes from an older English use of "trade" meaning a fixed track or course — a wind that blows steadily in one direction. The commercial usefulness came later and reinforced the name.
And the pattern dictated three centuries of sailing. Columbus went west on the trade winds from the Canaries and came home on the westerlies further north; every Spanish fleet afterwards used the same two-lane route. The Manila galleons crossed the Pacific westward on the trades and returned via the northern westerlies. The southern westerlies between 40 and 60 degrees south, uninterrupted by any continent, blow so hard and so consistently that sailors named them the Roaring Forties and Furious Fifties, and the fastest clipper routes to Australia and the fastest way home round Cape Horn both used them. Chapter 8.4 shows how much of the shape of European expansion was set by this map.
The jet streams
At the boundaries between cells, where cold air meets warm air, the temperature contrast is sharp. A sharp horizontal temperature contrast produces a strong wind high up, for a reason that is worth spelling out: cold air is denser, so pressure falls off faster with height in cold air than in warm air. Go up a few kilometres and you find a large pressure difference between the cold side and the warm side even if there was none at the surface. That difference drives a fast wind, and the Coriolis effect turns it to run along the temperature boundary rather than across it.
The result is a ribbon of air moving west to east at 100 to 400 km/h, near the tropopause, a few hundred kilometres wide and a few kilometres deep. There are two in each hemisphere: the polar jet at around 60 degrees, and the subtropical jet at around 30.
The jets do not run straight. They meander in large waves — Rossby waves — that shift north and south. Those meanders are what steer weather systems, and a stuck meander is what produces a heatwave that will not break or a rain event that sits over one basin for a week.
Three practical consequences.
Flight times. A flight from Delhi to London against the jet takes noticeably longer than the return, which rides it. Airlines route to catch it eastbound and avoid it westbound; the difference can exceed an hour.
Indian winters. The subtropical jet is split by the Tibetan Plateau in winter, and its southern branch runs across northern India. Disturbances embedded in it, travelling from the Mediterranean and called western disturbances, are what bring winter rain and snow to Punjab, Haryana, Uttarakhand, Himachal and Kashmir. That rain is what the rabi wheat crop lives on, so a winter with few western disturbances is a bad wheat year.
The monsoon's trigger. The northward jump of that same jet in late spring, when the plateau heats up, is one of the events that unlocks the monsoon — Chapter 2.5.
Local winds, and why the coast breathes
The global picture is the average. Locally, the same principle — uneven heating makes pressure differences makes wind — operates on scales of kilometres.
Sea breeze and land breeze. Land heats and cools far faster than water, because water has a high heat capacity, mixes, and is partly transparent so the heat is spread through a depth. By afternoon the land is much warmer than the sea, so air rises over the land and cooler sea air flows in to replace it. That is the afternoon sea breeze that every coastal city on Earth gets. At night it reverses, weakly, as the land cools faster than the sea.
Mountain and valley winds. Slopes catch the morning sun before the valley floor, so warm air runs up the slopes during the day and cold dense air drains down them at night. The night-time drainage is why valley floors get frost when the slopes above them do not, and why orchards in hill country are planted on slopes rather than in hollows.
Föhn or rain-shadow winds. Air forced over a mountain range rains on the windward side, then descends the other side. It is now dry, and descending air warms by compression — and because it has already dumped its moisture, it warms faster on the way down than it cooled on the way up. So it arrives hot and dry. This produces the loo of northern India in May and June, the Santa Ana winds that fan Californian wildfires, and the extraordinary dryness of the Tibetan Plateau and Ladakh, which sit in the Himalaya's rain shadow.
What the equations actually say
The full description of atmospheric motion is a set of coupled equations expressing conservation of momentum, mass and energy for a fluid on a rotating sphere — the Navier–Stokes equations with rotation added. They cannot be solved exactly for any realistic case, which is why forecasting is done numerically, and Chapter 2.9 deals with that.
But one simplification explains almost everything on a weather map. Away from the surface, where friction is small, the pressure-gradient force and the Coriolis effect come into balance:
f v = \frac{1}{\rho}\frac{\Delta P}{\Delta n}
Here f is the Coriolis parameter (which grows with latitude), v the wind speed, \rho the air density, and \Delta P / \Delta n the rate at which pressure changes across the flow. In words: the faster the wind, the steeper the pressure change it can balance.
The consequence is startling and it is why weather maps work. In this balance the wind blows along the isobars, not across them — parallel to the lines of equal pressure, with low pressure on the left in the northern hemisphere. So air does not flow straight into a low-pressure centre; it spirals around it. Northern hemisphere: anticlockwise around a low, clockwise around a high. Southern hemisphere: the reverse. Every satellite image of a cyclone shows it. Near the ground, friction slows the wind, weakening the Coriolis deflection, so the air does spiral inward — which is what feeds a storm.
Where this shows up in your life
Why your city has the climate it has. Mumbai and Chennai sit at similar latitudes but get their rain in different seasons, because one faces the summer monsoon flow and the other catches the retreating northeast monsoon. Delhi is hot and dry in May because it sits under descending subtropical air with the loo blowing across it. Bengaluru is mild for its latitude because it is at 900 metres.
Wind power siting. Turbines go where the belts and local effects concentrate wind: Tamil Nadu's Palghat gap, where the monsoon flow is funnelled through a break in the Western Ghats; Gujarat's coast; the North Sea, in the heart of the westerlies.
Air pollution. Delhi's winter smog is a wind problem as much as an emissions problem. In winter, cold heavy air settles near the surface and a warmer layer sits above it — a temperature inversion — which puts a lid on convection and stops pollutants dispersing upward, while the winds are weak. The same emissions in June, with strong heating and a deep mixing layer, produce far lower concentrations.
And the fact that the deserts are where they are has shaped where humans farmed, where they could not, which civilisations grew along rivers running through deserts, and where the trade routes had to go. Part 4 begins in exactly that band.
What the next page covers
Wind moves air, but weather is mostly about what the air is carrying. Chapter 2.3 covers water in the atmosphere — what humidity actually measures, why air cools when it rises even though nothing takes heat away from it, how a cloud forms and why there are so many kinds, why raindrops need dust to exist at all, and where lightning gets its charge.