Skip Navigation

May-June 2011

ResizeResize Text: Original Large XLarge

Weather Queries

I wonder if we are not overcomplicating the tornado process? If I take a spoon and spin water in a glass, the surface of the water rises at the sides and lowers at the center, creating a funnel. Why wouldn't this easily explain why tornadoes form in supercell thunderstorms? Supercells have copious moist air, providing buoyant energy and mass. The Coriolis effect starts the air spinning. With time, the spin will increase, and eventually the center of circulation will touch the ground. Am I going too far in simplifying the tornadic process?

Steve Dryja
Muskego, Wisconsin

Yes, but you deserve more than a one-word answer. The Coriolis force is significant everywhere except near the Equator, and its effect is most evident when there is a horizontal pressure gradient, that is, a difference in pressure between two points at the same elevation above sea level. If the Earth were not rotating, air would flow in a straight line from higher to lower pressure. The Earth's rotation gives rise to the Coriolis force and, away from the Equator, it deflects air from a straight-line path toward lower pressure. In the Northern Hemisphere, the deflection is toward the right, so, after hours of action by the Coriolis force, air begins to circulate counterclockwise around lower pressure. Friction forces, especially near the Earth's surface, cause the air to spiral inward. Once counterclockwise flow around a center of low pressure is established, the Coriolis force alone does not intensify the circulation. That can only happen if the pressure gradient increases.

Vorticity is a measure of rotation in the air. Pure rotary motion is one example of vorticity, but there is a second source: wind shear. Wind shear occurs whenever the wind (direction and/or speed) changes from one point to another. The connection between shear and rotation is easy to see by considering a stream in which the flow of water increases from the bank to the middle. If you place a large stick in the stream and orient it perpendicular to the shore, it will begin to rotate as the end nearer the middle moves downstream faster than the end nearer the shore. As long as different parts of the stick experience different velocities, it will continue to rotate.

A stick in a stream illustrates horizontal shear. An example of strong horizontal shear in the atmosphere may be found along the edges of the jet stream, where the wind speed can change by 50 mph in just 10 to 20 miles of horizontal distance. Much stronger shear often occurs in the vertical direction. For example, the wind speed may change by 20 mph within a few thousand feet of altitude. This vertical shear leads to rotation around a horizontal axis, just as a stick in a stream rotates about a vertical axis.

Caption: Illustrating how horizontal vorticity arising from unidirectional wind shear can be tilted into the vertical by an updraft, thereby spawning a supercell thunderstorm with counter-rotating updrafts. A more detailed explanation is in the text. From Tornado Alley: Monster Storms of the Great Plains, by Howard B. Bluestein, 1999, p. 69.

Caption: Illustrating how horizontal vorticity arising from unidirectional wind shear can be tilted into the vertical by an updraft, thereby spawning a supercell thunderstorm with counter-rotating updrafts. A more detailed explanation is in the text. From Tornado Alley: Monster Storms of the Great Plains, by Howard B. Bluestein, 1999, p. 69.

The figure illustrates how vertical shear can promote rotation in thunderstorms. Thunderstorms that have internal rotation are called supercell thunderstorms, and they sometimes spawn tornadoes. At the center of the figure is a rough sketch of a supercell thunderstorm. Note the compass directions near the left edge. On the southwest corner, the stack of arrows depicts a vertical wind profile. The winds are relative to storm motion, and, for simplicity, they all lie in a vertical plane oriented east to west. The storm experiences inflow from the east at lower levels (see the 3-D arrow at low levels near the center of the image). At higher levels, the wind blows from the west, increasing with altitude. High winds near the top of the storm help to spread the anvil toward the east (3-D arrow at high levels). The shear in this wind profile leads to rotation about horizontal axes, one of which is shown at low levels near the east edge of the diagram. The axis of rotation is called a vortex line, and the arrows about the axis show the sense of rotation.

Every developing or mature thunderstorm has an updraft near its base. When the updraft encounters a horizontal vortex line, it carries the line upward into an inverted “U” as shown. The sides of the “U” are nearly vertical, and the sense of rotation, indicated by the arrows, is cyclonic (counterclockwise) on the left side, and anticyclonic on the right side. The strength of rotation depends on the strength of the original vertical shear, so it should not be surprising that supercell thunderstorms require strong low-level shear as a precondition for their development.

As seen at the center of the figure, precipitation forms in the updraft, most readily between the two centers of rotation, and initiates a downdraft, often helping to drive the two centers farther apart. The environmental wind vectors seldom lie in the same plane, as in the special case illustrated here. The result is that one or the other rotating updraft weakens, in most cases, the anticyclonic one.

Supercell thunderstorms produce the most violent and longest lasting tornadoes, but these are in the minority. Nonsupercell thunderstorms produce the great majority of tornadoes, which are generally smaller, shorter-lived, and less violent. All tornadoes require a source of rotation. In supercell thunderstorms, the principal source of rotation comes from strong low-level shear when vortex lines are tilted into the vertical. For most other tornadoes, the rotation (about a vertical axis) is present in the low-level air flow before the storm develops, for example, along a wind-shift line, and it need not be pronounced.

The missing ingredient, not yet discussed, is a mechanism to concentrate the rotation into a much smaller area. This ingredient is called convergence. Stirring liquid in a cup imparts rotation mechanically. A better example of what happens in tornado formation is pulling the plug at the bottom of a bathtub. In order to drain from the tub under the influence of gravity, water must converge toward the opening. If there is any rotation in the water prior to pulling the plug, it intensifies as the water nears the drain. As the water spins faster, centrifugal force causes a funnel to form over the drain. Conservation of angular momentum, the same physical principle that explains why a figure skater spins faster as she draws her arms close to her body, is responsible for the increasing rotation as water approaches the drain.

In a thunderstorm, supercell or otherwise, rising air becomes buoyant and accelerates upward. This causes a reduction in pressure near the base of the thundercloud and beneath it. This in turn increases the pressure gradient and causes convergence near the base of the updraft, all of which draws down the radius of rotation and increases the rotary wind speed. Whether a tornado forms depends upon how much vorticity is present at low levels before the thunderstorm forms and how much convergence is induced by buoyancy forces within the thunderstorm updraft.

Even the foregoing is an oversimplification. The genesis of tornadoes is very complicated, and the subject of active research. Vortex2 is a multiyear field project to expand knowledge about the formation of tornadoes. Consult the Weatherwise issue from January/February 2010 for details, or visit

In July, I observed a small projection from beneath an isolated cloud. It took some time to conclude that the projection was indeed rotating, perhaps one revolution every 40 seconds. I have no idea how long it was rotating before I observed it, but I watched this “funnel cloud” for nearly six more minutes before it dissipated. (What a time not to have a camera!) The day was hot and humid, with a few hit-and-miss thundershowers, but the cloud with the funnel beneath it was maybe 1000 feet high, which seems far short of what is required to generate a tornado. What did I observe?

George M. Gumbert
Bowling Green, Kentucky

I periodically review questions I have held for a long time (because I couldn't find an answer to them when I first received them). I received this question in 1989. In seeing it again, I recalled an article published in Monthly Weather Review by Howard B. Bluestein, an expert on tornadoes and other severe weather: “More Observations of Small Funnel Clouds and Other Tubular Clouds” (December, 2005, pp. 3714–3720). Bluestein presents a number of his personal photos of small funnels, some connected to fairly innocuous-looking cumulus clouds and some detached from clouds altogether. I offer thanks to him for permission to reprint one of those photos here, which may resemble what George Gumbert saw.

Caption: A miniature funnel extending from the base of a cumulus cloud in eastern Colorado.

Caption: A miniature funnel extending from the base of a cumulus cloud in eastern Colorado.

The funnel is a narrow wisp just above the center of the photo. It extended downward from a high-based, towering cumulus cloud, of which only part of the base is visible. Such funnels are rare and often fleeting. If you see one without your camera at the ready, chances are it will be gone before you can retrieve your camera. The examples shown by Bluestein and the one described in the question all occurred during the warmest part of the day, when a deep layer of air is well mixed from the ground up, and cumulus clouds, from small to huge (cumulonimbus), dot the sky. Tiny funnels are matters of curiosity rather than cause for alarm, unless they are pendant from the base of a rapidly developing thunderstorm, in which case they may grow larger and extend downward.

Funnels, usually small, have also been observed at the base of stratocumulus clouds, separated towering cumulus clouds, and sometimes small cumulonimbus clouds, when a deep cold air mass moves over a much warmer surface. These cold air funnels have been sighted in the United States Midwest from late spring to early fall. They also occur along the California coast in winter, when storms move in off the Pacific with very cold air aloft. Finally, they occur over the Great Lakes and the Gulf Stream when cold air passes over warm water in late fall and winter. In the rare instances where they touch down, the damage is only slight. The reason is that the convection that gives rise to cold air funnels is usually shallow in comparison to that occurring in warm, humid air masses, and the atmospheric instability is moderate rather than large.

All these funnels have one thing in common: they are associated with convection, predominantly vertical motions driven by buoyancy forces within a cloud. Turbulence (irregular fluctuations in air motion) invariably accompanies convection and consists of wind shear on scales from meters to hundreds of meters. As noted in the preceding answer, shear is a source of vorticity. Along the edge of a convective updraft, the vertical wind speed changes quickly over small horizontal distances: horizontal shear of the vertical wind. (This explains why aircraft shake when they enter a cumulus cloud.) The vortex lines arising from this shear are horizontal, but they can be tilted into the vertical. Then, if they are stretched by being accelerated into an updraft, the pressure might drop enough to condense a visible funnel. In mountainous country, vortices may be generated as air flows over ridge tops or around prominent peaks. Convection can stretch these vortices and generate ephemeral funnel clouds.

Small funnels, whether attached to cumulus clouds or detached from them, are small-scale, short-lived, and innocuous curiosities, seldom observed. Some plausible mechanisms for their formation are discussed above, but the details remain largely undiscovered.

Weatherwise Contributing Editor THOMAS SCHLATTER is a retired meteorologist and volunteer at NOAA's Earth System Research Laboratory in Boulder, Colorado. Submit queries to the author at, or by mail in care of Weatherwise, Taylor & Francis, 325 Chestnut St., Suite 800, Philadelphia, PA 19106.

In this Issue

On this Topic

Privacy Policy

© 2018 Taylor & Francis Group · 530 Walnut Street, Suite 850, Philadelphia, PA · 19106