Coastal fog belts

Why does a tropical coastline stay cold and fog-bound while the open sea just offshore looks inviting?

Stand on the shore at Lima, or on the Namibian coast, or at Iquique in northern Chile. You are in the tropics or close to it, the sun is fierce wherever it reaches, and the water is painfully cold. Grey stratus sits overhead most mornings of the year. Sail a few hundred kilometres out and the sea is warm and the sky is blue.

The assumption worth questioning is the obvious one: that how warm the sea is at a place is set by how much sun that latitude receives. If that were so, the coast and the water offshore — same latitude, same sun — could not differ. Something other than sunlight is deciding.

Sunlight sets how much heat arrives. It does not set what water is sitting at the surface, because the surface layer is thin and water can be moved — and whatever water is there is what you feel.

So ask what moves it. Along all these coasts the prevailing wind blows roughly along the shore, towards the equator. Wind drags on the surface, but a moving layer on a rotating planet does not go where it is pushed: it is deflected, right in the northern hemisphere and left in the southern, so the net transport through the wind-driven layer ends up roughly at right angles to the wind. Now check the geometry: on the western coast of a continent, an equatorward wind turns that transport away from the shore — and it does so in both hemispheres, because the wind direction and the sense of the deflection flip together. Hence the same pattern off Peru and Namibia and off California and Morocco.

Water leaving the coast has to be replaced. It cannot come from the beach or from the sky at any useful rate, so it comes from below — and a hundred metres or two down, the water is cold. The coast is cold because the wind keeps peeling the sun-warmed skin off it, not because the sun is weak.

Now put air over that. Marine air arriving from offshore is warm and near saturation. Run it across a surface several degrees colder and it is chilled from below to its dew point, and its moisture condenses in place. That is advection fog — fog made by air moving over a colder surface — a different animal from the radiation fog that forms overnight in a still valley.

Two consequences follow. Cooling air from below makes the column stable, so nothing lifts the fog by mixing it upward. And these are the latitudes where the subtropical high sits, air subsiding and warming as it descends — the mechanism desert-latitudes.md works through — laying a warm lid a few hundred metres up. The moist layer is trapped between a cold sea and a warm ceiling: cloud forms readily, rain, which needs deep ascent, cannot. So you get one of the few places on Earth cloudy almost every day and almost never wet. Coastal Peru's rainfall is on the order of a centimetre a year, a figure I am recalling rather than deriving, and nothing below leans on it.

What would refute this? The account makes the wind, not the sun, the driver, so slacken the alongshore wind and the structure should unravel within days: offshore transport stops, upwelling stops, the surface warms, the fog collapses, and with the cold lower boundary gone, rain becomes possible. That is precisely the El Nino signature off Peru and Ecuador — warm coastal water and rain falling on a desert. If a strong warm event arrived with the alongshore wind unchanged and the coast stayed cold and fog-bound, the mechanism given here would be wrong. The test runs at short range too: fog frequency should track wind stress with a lag of days, not the seasonal solar cycle.

Two neighbours need naming. ocean-productivity.md runs the same pump — wind-driven upwelling on an eastern ocean margin — but follows the cold water's nutrients into the food web, where this piece follows the same water's temperature into the atmosphere; the two diverge the moment the water reaches the surface. desert-latitudes.md explains the sinking air that makes these latitudes dry; here that subsidence is not the subject but the lid, and the question it leaves open — why the coastal strip of a hot desert is cold and grey rather than hot and clear — is the one answered here.

Three tempting explanations should go. First, "a cold current from the poles." The Humboldt and Benguela do carry cooler water equatorward, but the cold band is colder than the current's own water, pressed against the shore, and coldest at the beach rather than offshore. No broad current makes a gradient that sharp; upwelling does, because the replacement happens exactly where the water is removed.

Second, "coasts are foggy, that is all." The pattern is specific to eastern ocean margins; at the same latitude on a western margin — Florida, Mozambique, Queensland — the water is warm and the air is clear.

Third, "the water is cold because it is always cloudy." That reverses the causation: the cloud needs a cold surface to condense on, and the warm layer above it is made by subsidence, not shade.

One honest simplification: I have written wind, transport and fog as a tidy chain. Coastline shape, seafloor topography and inversion strength all modulate where the cold band is strongest, and the weight of local upwelling against advection from further along the coast is argued over system by system.

• Why upwelling coasts also carry the world's largest fisheries, and how the same water does both jobs.
• How fog-dependent ecosystems harvest water that never falls as rain.

Every term the collection defines is gathered in the glossary.