Rotation and storm size
A Socratic walk-through of rotation and storm size — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why does Earth's rotation steer a hurricane but not the water leaving a sink?
Everyone has heard that the Earth's spin makes hurricanes turn one way north of the equator and the other way south of it. Everyone has also heard it applied to the bathroom sink — and has probably watched a sink drain the "wrong" way and concluded the whole idea is a myth.
Both cannot be right, and yet the physics does not have a size setting. The same planet turns beneath the storm and beneath the sink, at the same rate, with the same law acting on both. So what actually differs?
Reasoning it through
REASONING #Start by asking how fast the planet's rotation can bend anything.
The sideways acceleration a moving parcel picks up in the rotating frame is the Coriolis parameter times its speed. That parameter, written f, is twice the Earth's rotation rate times the sine of the latitude — about one ten-thousandth per second at forty-five degrees. It is not a big number, and crucially it is a rate, not a force you can turn up. It sets a clock: the time for the planet's spin to appreciably curve a track is roughly one over f, which is a few hours.
Now put the two flows next to that clock. Water leaves a sink in perhaps ten seconds. In ten seconds the Coriolis term can turn the flow by about f times t — a thousandth of a radian, a twentieth of a degree. Air spiralling into a hurricane is in the storm's circulation for days. Same law, same planet; one flow simply lives long enough for a slow effect to accumulate and the other does not.
There is a tidier way to say it that turns the comparison into a single number. Compare the parcel's own turning — how sharply it must curve just to go round the flow, roughly its speed divided by the size of the flow — against the planet's turning, f. The ratio is called the Rossby number: speed divided by f times length. When it is much larger than one, the flow's own dynamics dominate and rotation is a rounding error. When it approaches one, rotation is an equal partner.
Run the arithmetic. In a sink, water moves at maybe a tenth of a metre per second across a basin a third of a metre wide: the ratio comes out in the thousands. In a hurricane, winds of tens of metres per second sweep across hundreds of kilometres: the ratio comes out around one. That is a factor of a few thousand between them, and notice what produced it. The speeds are not so different — a factor of a few hundred. The length is different by a factor of a million.
So the answer to the question is that size is the knob. Rotation does not care how hard the wind blows nearly as much as it cares how far it has to travel, because distance is what buys time.
Does that mean the sink has no Coriolis signal at all? No — and this is the part usually got wrong in both directions. In 1962 Ascher Shapiro at MIT, and in 1965 Lloyd Trefethen in Sydney, drained large shallow tanks that had been covered and left to settle for around a day, so that the residual swirl from filling had died away. Drained slowly, the vortex went reliably counterclockwise in the northern experiment and clockwise in the southern one. The effect is there. It is simply about a thousandth of the strength of everything else in a real basin — the jet from the tap, the shape of the bowl, the swirl you left when you pulled your hand out — so in practice it is swamped.
The analogy
THE ANALOGY #Think of a very light crosswind, the same everywhere. Toss a ball across a room and you would never detect it. Hit a golf ball two hundred metres and the same breeze moves it several metres off line. Nothing about the wind changed between the two shots; only how long the ball was exposed to it.
a crosswind pushes always in one fixed direction, so it produces a drift, whereas the Coriolis deflection is always sideways relative to the motion itself — so instead of nudging a flow off course it curls the flow into a rotating system that can then sustain itself, which is a hurricane and not a bent trajectory.
Clarifying the model
THE MODEL #Three refinements, because the scale argument alone can be pushed too far.
First, f is not constant over the globe. It is proportional to the sine of the latitude, so it falls to zero at the equator. That is why tropical cyclones do not form within roughly five degrees of the equator: there is no rotation available to organise the convection, however warm the sea. The exception proves the rule — Typhoon Vamei formed near one and a half degrees north in 2001 and is treated as a genuine oddity.
Second, "rotation matters" is not the same as "rotation causes". The Earth's spin does not spin up a hurricane; the storm's energy comes from warm ocean water and latent heat released as moist air rises. Rotation organises what convection has already built, and fixes the sense of the spiral.
Third, the sink is not a small hurricane at all. A draining vortex conserves the angular momentum it happened to start with, amplifying it enormously as the water converges — which is exactly why any stray initial swirl beats the planetary contribution so decisively. The two systems share a picture, not a mechanism.
A picture of it
THE PICTURE #How to readEach bar is the Rossby number for one kind of flow, written as a power of ten, so a bar at 3 means the flow's own turning beats the planet's by a thousandfold. The flat line at zero is where the two are equal. Read left to right as increasing size: bars far above the line are flows too small and too brief for rotation to touch, bars at or below it are flows large enough that rotation is a full partner in the dynamics. The sink sits three and a half orders of magnitude above the line — not because the physics is absent there, but because everything else in the basin is a thousand times louder.
What became clearer
WHAT CLEARED #There is no threshold at which Coriolis switches on. There is a slow, fixed rate of deflection that only ever accumulates into something visible if the flow persists long enough, and persistence is bought almost entirely with size. Sink and storm are governed by identical equations; what separates them is a factor of a million in width, which is enough to move one term of those equations from negligible to decisive.
Where to go next
ONWARD #- Geostrophic balance — what happens once rotation becomes the dominant term rather than a competing one.
- Why the same Rossby argument makes ocean currents lean west, and why western boundary currents like the Gulf Stream are so narrow.
- The inertial oscillation — what a parcel does when rotation is the only thing acting on it.
Key terms
TERMS #| Term | What it means |
|---|---|
| Coriolis parameter (f) | twice the Earth's rotation rate times the sine of latitude; sets the rate at which the rotating frame deflects a moving parcel. |
| Rossby number | speed divided by f times length; the ratio of a flow's own turning to the planet's, and the single number that decides whether rotation matters. |
| Inertial period | roughly the time scale one over f, a few hours at mid latitudes, over which planetary deflection becomes appreciable. |
| Angular momentum conservation | the amplification of any pre-existing swirl as fluid converges inward, which dominates small drains. |
Every term the collection defines is gathered in the glossary.