Optional Enrichment Article: Why the Wind Spins
All large flows on Earth appear to turn along their path, because of the Coriolis effect, which arises because the Earth rotates. You can find serious discussions of the Coriolis effect in any good textbook on meteorology or oceanography. The material in this optional Enrichment article provides a more intuitive version. This isn’t a complete explanation… but we don’t know of a complete explanation that avoids serious math and physics.
The rotation of the Earth is surprisingly fast. If you buckled a belt around the Earth at the equator, the belt would need to be 25,000 miles (about 40,000 km) long. The Earth spins once every 24 hours. This means that the Earth’s rotation is moving a point on the equator at just over 1000 miles per hour (1600 km/hr). In comparison, a point near the pole is essentially stationary (the pole itself stays in place as it rotates). Dr. Alley has walked around the South Pole in three steps, but he couldn’t walk around the equator in a day. If this doesn’t make sense, get a tennis ball or the head of a friendly classmate, draw an equator, and try it out!
If you stand on the equator, the wind is not blowing 1000 miles per hour. This is fortunate indeed because the “mere” 150-mile-per-hour winds occasionally reached in hurricanes can cause immense disasters. Instead, the air near the surface of the Earth is, on average, moving with the surface, dragged along by the mountains and trees and such. You may have seen something similar if you have driven down a highway at 65 miles per hour with a bug stuck to your windshield. The wind at 65 miles per hour surely is strong enough to blow a bug, but the wind very close to the windshield is very much slower than 65 miles per hour—really strong winds don’t blow close to surfaces, because of the friction of the surface, and instead, the air close to the surface does almost the same thing that the surface is doing.
Now, suppose that you watch a parcel of air that rises from the sun-warmed surface at the equator and begins moving towards the North Pole in a convection cell. Once the air has moved many miles north, the Earth under it is no longer rotating at 1000 miles per hour, but is somewhat slower, perhaps 900 miles per hour, and the farther north the air goes, the slower the surface is moving, dropping toward zero at the pole. But aloft, there aren’t trees and mountains to slow the air down from the 1000 miles per hour it had at the equator. Thus, this continues to move at 1000 miles per hour, and “gets ahead” of the Earth beneath. The Earth rotates to the east—you see the sunrise in the east as the rotation of the Earth brings the sun into view—so, the equatorial surface air is moving east at 1000 miles per hour. As this air rises and moves northward over slower-moving land, the wind will appear to turn to the right or east as it blows, getting ahead of the ground. Wind heading south from the equator will also move east ahead of the surface, making a left turn.
Similarly, wind moving from the pole to the equator in the returning limb of a convection cell will lag behind the rotating surface of the Earth, again seeming to turn to the right in the northern hemisphere (and to the left in the southern hemisphere). Thus, the wind cannot go directly to where it “wants” to go; instead, it turns and tends to go in circles. The circular airflow around low-pressure systems and hurricanes occurs because the Earth rotates.
As noted earlier, more-precise definitions are possible of this “Coriolis effect,” which explains why all large flows turn on a rotating Earth. The intuitive explanation given above will fail you if you think of a wind moving due-east or due-west because those winds also turn. Starting from the conservation of angular momentum might be better. Notice, however, that the explanation given above will get the right answer for you, in how much the wind will turn, and which way.
Notice also that Coriolis turning affects large, fast flows, not large slow flows or small ones at any speed. The geological convection deep in the Earth is large, but it is too slow to feel Coriolis much. The difference in rotation speed between opposite sides of a kitchen sink or a bathroom toilet is so tiny that Coriolis turning has no significant effect on the direction that the water swirls as it goes down. Instead, those water swirls are controlled by the design of the sink or toilet, and by any motion in the water at the time the drain was opened; get the water swirling in a sink and then pull the plug, and the water usually will keep swirling in the way you started it as it goes down. Dr. Alley has seen both clockwise and counterclockwise flows in Pennsylvania and Greenland, and in New Zealand and Antarctica.
More Enrichment: Why Cold Air Sinks but Valleys Are Warmer than Mountain Peaks
When air moves up, it expands, which requires that work is done in pushing away other air to make room for the expansion. The work requires energy, which comes from the heat energy in the air, so the rising air cools. Similarly, when air moves down, it contracts as the surrounding, higher-pressure air squeezes the sinking air parcel, and this squeezing is work that is done on the sinking parcel and warms the parcel. If this is happening near the surface of the Earth, and the air is dry, the change in temperature is about 1oC per 100 m of vertical motion. This applies everywhere, at all times. So, it was complete nonsense in the 2004 movie, The Day After Tomorrow, when huge storms brought air down from above so rapidly that the air didn’t have time to warm up.
As noted in the main text, air can cool by expanding as it rises, but also by losing energy by radiation or by conduction into colder land or water beneath. Imagine that a “chunk” or parcel of air, sitting somewhere on the side of a hill, cools a little by losing energy, perhaps by radiating energy to space as the sun goes down in the evening. Will that air parcel sink now, flowing along the hill into a valley? The answer is that it will sink if, after it has sunk and warmed, it is colder than the air that started out in the valley and had to be displaced as the sinking air arrived.
Consider an example. You measure the temperature of your parcel, add 1oC for the warming from sinking 100 m, and if your air is still colder than the air it must displace 100 m below, then your parcel will sink. (If you do this really carefully, friction comes in as well—if your parcel plus 1oC would be only a tiny bit colder than the air it must displace, motion is unlikely; you need a notable difference to overcome the friction and really move.)
Overall, a balanced, stationary atmosphere will cool upward by about 1oC per 100 m under dry conditions, and somewhat less under wet conditions, as described in the main text. Vertical motions will be triggered when cooling or warming creates air that is anomalously cold or warm relative to this stationary profile. So, cold air on a mountaintop won’t necessarily sink, unless that mountaintop air is colder than you would expect from this profile. But on an October evening in the Appalachians, when fog develops and holds heat in the valleys while the mountaintop radiates heat to space, the mountaintop air will become anomalously cold and sink to the valleys.