STORM TRACK: May 31, 1987 (Volume 10 Issue 4)

Last November, I gave a slide/video talk show on severe weather at the David W. Taylor Naval ship Research and Development Center (Bethesda, MD). With their primary interest in coastal and ocean storms, I focused more on waterspouts and the comparatively rare "tornadic waterspout" (Morton, 1966). This is a tornado that originates over land and then passes over water. Like land based tornadoes, it differs from a waterspout in intensity, duration, and translational speed (1/Golden, 1974) -all the much higher values. In addition, although upper air dynamics are usually weak (2/Golden, 1973), there often is some synoptic forcing, although far less than for land tornadoes. Still, based on their size and destructive potential, I think it is misleading to refer to them as "tornadic waterspouts". This denigrates the danger and unique characteristics of such events and makes a mischievous linkage with their more anemic cousins. A better term may be "waternado" which emphasizes both the singularity and danger of this particular vortex. Hereafter, the term waternado is used.

There is ample evidence for this type of storm. Some of the more notable accounts: 1) Venice, Italy on September 4, 1970. Such a storm struck and lifted a 25 ton passenger boat, turned it around several times in the air, and sent it to the bottom of the harbor in 30 seconds killing 18 people (3/Golden, 1971), (2) At Kialua, Kona, Hawaii, on January 28, 1971, one came ashore and traced a destructive path some 2-1/2 miles inland (4/Zipser), and (3) Several waternadoes were reported in the Miami area during an unusually stormy summer in 1968, one lifted a five ton houseboat from the water, and the other sank six cabin cruisers in a marina.

Although rarely photographed, I did have two good slides of waternadoes for the slide show, both from the southeast Florida coast: (1) Howie Bluestein's great photo over Key Biscayne; and (2) Jim Leonard's memorable June 12, 1985 vortex, about 8-10 miles off-shore from West Palm Beach (illustrated in the March 31, 1986 issue of STORM TRACK). After only a few inquiries of the several leading researchers in the filed, I found there has been no single written study on waternadoes. What I was able to discover from other related studies is the following:

1) On radar, "Rotating hook echoes could be directly associated with large waterspouts" and about 10% of waterspouts have hook echoes (1/Golden, 1974). Thus deductively, perhaps as many as one in ten are tornadic types.

2) Maximum rotational winds for "waterspouts" have been recorded "as high as ...190 knots" or about 220 mph (5/Golden, 1977).

3) Thirty percent of "waterspouts" have associated thunder and lightning (2/Golden, 1973).

4) Unlike waterspouts, if a rear flank downdraft is present, the waternado is usually unaffected (1/Golden, 1974).

5) Cloud tops are more like land tornadoes for the waternado variety with radar echoes between 30,000 and 50,000 feet (3/Golden, 1971). The waternado viewed by Jim Leonard had a cell top of 53,000 feet versus the normal 12,000 to 20,000 ft cell top for most waterspout cases.

6) Waternadoes may occur near and along the US coastline, probably within 15 miles, and from New Jersey to the southeast Florida coast.

7) Waternadoes have prevailing weak synoptic dynamics, as with waterspouts (2/Golden, 1973), but with some impetus from a cold front, some vertical shear, or a strong approaching upper wave. The Hawaiian tornado studied by Zipser resulted from interacting trade winds southwest of an elongated Northwest cold front, and a deep approaching 500 mb low.

8) In some cases, convection that produced waternadoes may be enhanced by alignment with the Gulf Stream, a land breeze with somewhat drier air, and comparatively warmer night-time temperatures from across the Gulf Stream flowing into a storm line parallel with the coast (i.e. Jim Leonard's midnight storm.)

VISUAL/AUDIBLE CLUES

9) Cone shaped or tapered vortex column for waternadoes. Waterspouts are typically cylindrical in appearance.

10) Waternadoes have higher translation speeds and longer durations than waterspouts.

11) Thunder and lightning are often present around waternadoes.

However, all of the proceeding information is largely deductive and based on aggregated waterspout data and analyses, with an apparent bias toward common (not exceptional) environments for standard waterspouts. Taking a fresh look at the existing data, here are just a few questions which should be addressed:

1) How can waternadoes be identified and measured? How many really occur? Is it as high as 1:10?

2) What synoptic patterns favor waternado formation? If a strong thermal cap is unlikely, what compensates for this? Does the "cap" extend several miles over the water? Are there any dry intrusions from land at low or mid-levels?

3) What different patterns, if any, conduce to form tornadoes over water as compared to those moving from land to water?

4) What is the diurnal, climatic, and geographic frequency of waternadoes? Do any occur over the deep ocean waters, or are they only a hazard to coastal harbors and shipping?

5) How can the casual observer tell the differences between the waterspout and waternado? Do storms which produce waternadoes have inflow bands, rain-free bases, wall or tail clouds?

6) How many mysterious small boat losses can be attributed to such storms, especially at night? What. special precautions should boaters take to secure boats, deck gear, etc., if one of these waternadoes approaches?

These are only some of the many questions regarding waternadoes. It's food for thought and ample material for several graduate theses.

1/Golden, 1974; Journal of Applied Meteorology, Vol. 13, No. 6.

2/Golden, 1973; Weatherwise, June, 1973.

3/Golden, 1971; Monthly Weather Review, Vol. 99, No. 2.

4/Zipser, Randal; graduate study

5/Golden, 1977; Journal of Applied Meteorology, Vol 16, No. 3.