Funnel clouds, landspouts and gustnadoes

A funnel cloud is a rotating, visible extension of cloud hanging from a convective cloud whose circulation has not reached the ground. Once the circulation does reach the ground, the National Weather Service treats it as a tornado, whether or not the visible cloud goes all the way down.

The three terms in the title describe three different things, and only one of them is a tornado by definition. The tornadoes on radar guide covers the signatures of the classic supercell tornado. This page covers the vortices at the edges of that picture, several of which leave little or nothing in radar data.

What a funnel cloud is

NWS Instruction 10-1605, the directive that governs the Storm Data record, defines a funnel cloud as “a rotating visible extension of cloud pendant to a cumulus/cumulonimbus with circulation not reaching the ground or water.” The NWS online glossary adds the test that separates it from a tornado: a condensation funnel is a tornado, not a funnel cloud, “if either a) it is in contact with the ground or b) a debris cloud or dust whirl is visible beneath it.”

The American Meteorological Society’s Glossary of Meteorology, as reproduced in the NWS Radar and Applications Course glossary, uses the phrase more loosely: a condensation cloud associated with a vortex “that may or may not be in contact with the surface.” In that usage the funnel is simply the visible cloud, and a tornado can contain one. Spotter reports follow the NWS convention.

Two cautions from the Storm Prediction Center’s tornado FAQ matter here. Rotation is required: “If it is a low-hanging cloud with no rotation, it is not a funnel cloud.” And the cloud is not the vortex: “In a tornado, a damaging circulation is at the surface—whether or not the cloud is.” The same FAQ notes that tornadoes can occur with no visible funnel at all, most likely because the pressure drop and lift were too weak to condense one or the air below cloud base was too dry, and that what looks like “only” a funnel cloud may be doing damage that cannot be seen from a distance.

Cold-air funnels

The NWS glossary defines a cold-air funnel as “a funnel cloud or (rarely) a small, relatively weak tornado that can develop from a small shower or thunderstorm when the air aloft is unusually cold.” NWS Louisville describes the setting: they are most common in fall and spring, when sunshine heats the lowest levels enough to build showers while temperatures around 15,000 to 20,000 feet are quite cold. The office calls them “usually harmless,” but notes that on rare occasions they touch down and produce EF0 damage.

Storm Data treats a cold-air funnel that meets the objective tornado criteria in NWSI 10-1605 as a Tornado event.

Landspouts

“Landspout” is slang, and the NWS says so in every definition. The glossary describes it as a tornado “that does not arise from organized storm-scale rotation and therefore is not associated with a wall cloud (visually) or a mesocyclone (on radar).” The AMS glossary attributes the name to Howard Bluestein’s 1985 conference paper: it looks like a weak Florida Keys waterspout over land. The SPC FAQ is blunt about the consequence of the nickname: landspouts “are tornadoes by definition; and they are capable of doing significant damage and killing people.”

The mechanism is different from the supercell case described in the mesocyclone guide. Wakimoto and Wilson (1989, Monthly Weather Review) studied 27 visual vortices in Colorado during the 1987 CINDE field project. The circulations formed as shear instabilities along radar-detected convergence lines, usually before any precipitation echo existed, and strengthened to tornadic intensity when a rapidly growing storm’s updraft moved over them. The authors hypothesized that vortex stretching did the intensifying, and estimated damage capability as high as F2.

Northeastern Colorado is the classic setting. Szoke et al. (1984, also in Monthly Weather Review) analyzed the two F2 tornadoes that crossed metropolitan Denver on 3 June 1981 and described a zone of surface convergence and cyclonic vorticity over and north of the city, apparently forced by the ridge of higher terrain extending east from south of Denver. In the 1981–1982 mesonet data they examined, that zone was associated with a disproportionate number of severe events, especially tornadoes. It is now usually called the Denver Convergence-Vorticity Zone. Non-mesocyclone tornadoes near Denver can be strong as well: Caruso and Davies (below), citing Wakimoto and Wilson, note that two of four such tornadoes on 15 June 1988 caused F2 and F3 damage.

The same process works well beyond Colorado. Caruso and Davies (2005, NWA Electronic Journal of Operational Meteorology), writing from NWS Wichita, documented an F2 non-mesocyclone tornado near Wellington, Kansas, on 27 August 2004 that lasted nearly 20 minutes, and argued that “landspout,” read by many as “weak tornado,” can be misleading.

Gustnadoes

A gustnado is not a tornado. NWSI 10-1605 defines it as “a small, usually weak whirlwind, which forms as an eddy in thunderstorm outflows. They do not connect with any cloud-base rotation and are not tornadoes.” Storm Data logs them as Thunderstorm Wind events. The AMS glossary gives the same verdict for a physical reason: gustnadoes are “shallow and short-lived vortices that are usually weak.”

The confusion is visual. A gustnado raises a spinning dust or debris cloud at ground level along the gust front, which is also what a tornado can look like from a distance. The SPC FAQ draws the line at the cloud base: in gustnadoes, “there is spinning motion at ground level but not at cloud base.” It also warns that they can still break windows and tree limbs.

What radar can and cannot see

The condensation funnel itself is invisible to the WSR-88D. It is cloud, not precipitation, and it is far smaller than the radar beam at most ranges. What the radar can sometimes see is the parent circulation. A funnel under a supercell may sit beneath a mesocyclone that is obvious in velocity data. A funnel under a small shower may sit beneath nothing detectable.

Cold-air funnels are, in NWS Louisville’s words, “difficult to detect on radar since they are very weak.” The office illustrates the point with an 18 May 2006 case in Ohio: “no rotational signature in this shower despite the EF0 cold air tornado.”

Landspouts have no mesocyclone to find. Caruso and Davies state that because these tornadoes are not preceded by midlevel or low-level mesocyclones, WSR-88D velocity data “is usually of limited use in the traditional mode of tornado warning by radar.” What radar does offer is the boundary. Within roughly 60–80 km of the radar, they note, reflectivity and velocity can identify and track the convergence lines where pre-existing vorticity lives. Wakimoto and Wilson reached a similar conclusion: Doppler radar’s ability to detect motion in clear air, and attention to the lowest elevation angles, were what mattered.

Gustnadoes are too shallow and small to resolve. The gust front that spawns them often is visible: NWS radar training material notes that outflow boundaries “can be seen as fine lines on WSR-88D, with weak reflectivity returns.” A fine line tells you where the gust front is. It does not tell you whether a whirl is spinning along it.

WeatherStory Radar decodes raw Level II reflectivity and velocity at every gate, which is where boundaries and low-level circulations appear when they appear at all.