Opti National Championship Regatta Waterspout
July 28, 2012

Early on a Saturday morning in late July 2012, nearly four-hundred junior sailors were assembled at the Sandusky Sailing Club (SSC) in Sandusky, Ohio to participate in the Optimist Dinghy National Championship Regatta. Saturday was the third day of the Open Championship which was held on the protected waters of Sandusky Bay, east of downtown Sandusky and west of the Cedar Point causeway.

Optimist dinghies, affectionately called “Optis” by their owners, are competitively sailed by kids who are 15 years old or younger. The boats are small – less than 8 feet long and 4 feet wide – and carry a single sail measuring approximately 35 square feet. The sailors are typically organized into fleets based upon their age and sailing experience. But don’t allow the size of the boats or the age of the sailors deceive you -- Opti regattas are extremely competitive events (click here for image).

The LakeErieWX Forecast Summary for July 28, 2012 (click here) indicated that winds on the race course would range from eight to fourteen knots from the northeast, and wave heights would remain below one foot. The Summary also included a slight risk of showers and thunderstorms until early afternoon associated with a low pressure system located over northeast Ohio. This system had a proven track record of producing thunderstorms, with several clusters occurring over Lake Erie just before dawn (click here for image). In response to the combination of very warm Lake Erie water temperatures and cold air several thousand feet above the surface, the Summary also raised the potential for waterspouts.

The Skipper’s Meeting
As with most regattas, the day’s activities opened with a skipper’s meeting at 9:00 am, which provided an opportunity for race officials to make important announcements, answer questions and provide a short weather briefing. Several minutes into the meeting and just before the weather briefing was to begin (between 9:10 am and 9:15 am), “there’s a waterspout” was heard above the din and the competitors, parents and coaches rushed from the tent for a closer look. The waterspout was located approximately four miles east of the race course and was moving south.

Waterspout east of Cedar Point Causeway. Photograph take on July 28, 2012 by Nate Parsons. (larger image)
Waterspout Conceptual Model

Although I have sailed on Lake Erie for nearly twenty years and have “chased” waterspouts on several occasions, this was my first sighting. It was an extraordinary experience, and although I lost track of the time, I estimate the waterspout persisted for thirty to forty-five seconds. At no time was it a threat to the Opti sailors, and much to my surprise, they quickly returned to the skipper’s meeting after it dissipated.

Waterspouts: Two Varieties
Although both are intense columnar vortices and are considered “tornadoes over water”, waterspouts come in two varieties – tornadic and fair weather. The fundamental difference being the type of storm they are associated with and the manner in which they form. Tornadic waterspouts are produced by supercell thunderstorms and form when the relatively long-lived mid-level rotation within the storm extends down to the surface of the water. The self-sustaining nature of supercell thunderstorms may promote long-lived and cyclic waterspouts. Fortunately, only 5% of all thunderstorms are supercells, therefore the tornadic waterspouts occur very infrequently.

Far more common are the awkwardly-named fair weather waterspouts. Rather than descending from the cloud to the surface, fair weather waterspouts form upwards from the surface of the water in association with the updraft of a rapidly growing cumuliform cloud. Although far less powerful than their tornadic cousins, winds associated with fair weather waterspouts are capable of reaching sixty knots or more and have a long history of capsizing small boats, shredding sails and damaging equipment. They should always be avoided.

Fair Weather Waterspout Environments And Formation
Fair weather waterspouts harness the energy associated with the temperature difference between very warm water and relatively cold air several thousand feet above the surface. When analyzing the potential for waterspout development, meteorologists compare the lake temperature to the air temperature at approximately five-thousand feet to determine if the difference will be 13°C (23.4°F) or greater. Satisfying this condition determines that the atmosphere is supportive of rapidly developing thunderstorm updrafts.

Figure 1. Lake Erie: Average air temperature (red) and water temperature (blue).

Large temperature differences between the surface of the lake and five thousand feet are easier to achieve during late summer. As the chart of average monthly water and air temperatures (Figure 1) shows, Lake Erie typically reaches its maximum temperature in August, approximately one month behind the peak air temperature over the Lake. As summer gives way to fall, transient weather systems promote the invasion of much cooler air from the north which sets the stage for a dramatic increase in waterspouts from July to October (click here for graph).

However, temperature difference isn’t the only ingredient needed to produce an environment that is favorable for waterspouts. Since the vortex forms from the surface upward to the cloud base, a pre-existing source of low-level rotation (vorticity) must be available. This rotation is predominantly counter-clockwise and can be found along surface boundaries, such as a cold, warm or stationary fronts. Across the Great Lakes, waterspouts frequently tap the rotation associated with land breeze boundaries (click here for graphic) which typically form during early evening and persist until early or mid-morning.

Meteorologists use the term convergence to describe the opposing flow of wind along a surface boundary (click here for diagram). Convergence along a boundary tends to promote the upward motion of air (click here for diagram) which, if the atmospheric dynamics are supportive, leads to the development of a cloud and perhaps a thunderstorm. The combination of low level convergence and upward motion feeding the cloud act to stretch the updraft vertically (click here for animation from Penn State University), causing the pre-existing low-level rotation to increase and voila -- a waterspout forms.

Or in an example of pure coincidence, fair weather waterspouts may also form when the updraft associated with an pre-existing cumulus congestus cloud (click here for image) – fueled by the temperature difference discussed above – encounters a surface boundary, lifts and stretches the pre-existing low-level rotation up to its base.

Contrary to popular belief that a column of water is being lifted up to the cloud, the visibility of the waterspout is due to the condensation of water vapor that occurs as the air cool as it is lifted skyward.

A fledgling fair waterspout is relatively fragile and easily disrupted by too much wind, variations in wind direction or an encounter with cold air. Based on a study conducted by Wade Szilagyi, a meteorologist with Environment Canada, waterspouts rarely form when winds at approximately 5,000 feet are greater than 35 knots. It is also beneficial if the direction of the wind in the lowest few thousand feet is reasonably uniform, or in technical terms, when directional wind shear is weak. Relatively fast winds or strong wind shear inhibit the strength and continuity of the storm’s updraft and suppress the formation of the waterspout’s small-scale vortex. In addition, because rain-cooled air would quickly undercut the developing updraft, waterspouts typically form before the storm produces any precipitation.

Examining the Opti Regatta Waterspout
The storm that produced the Opti Regatta waterspout first appeared on Doppler radar east of Kelley’s Island at approximately 8:36 am (Figure 2). Observations at the Marblehead Light, several miles southwest of Kelley’s Island, indicate that surface winds in the vicinity of the storm were northwest (330°) at 12.1 knots gusting to 14.0. During the next 15 minutes, the storm strengthened significantly as it moved south-southeast at nearly 20 knots (click here for image). 

Figure 2. Doppler radar image from 8:36 am on July 28, 2012 from the National Weather Service. (Click here for a larger version.)

Typically by the end of July, Lake Erie reaches its seasonal high surface temperature. The analysis of surface water temperatures (Figure 3) issued by NOAA’s Great Lakes Coastal Forecasting System indicates the mean temperature for Lake Erie on July 28, 2012 was 76.6° F (24.8° C). Of particular interest however, is the narrow pool of warmer water (80.0° F / 26.7° C) along the Lake’s southwest corner including the area east of the Cedar Point peninsula.

Figure 3. Lake Erie Water Temperature on July 28, 2012. From the Great Lakes Coastal Forecasting System. (Click here for a larger version.)

The track of the storm suggested it remained over very warm water as it travelled south. An analysis of air temperature at approximately 5,000 feet (850 mb) indicated that temperatures over southwestern Lake Erie were approximately 55.4°F (13°C) at 8:00 am, resulting in a 24.6°F (13.7°C) difference between the surface of the water and 5,000 feet. This temperature difference slightly exceeds the 13°C threshold used to identify when the atmosphere is supportive of waterspout development.

Figure 4: Analysis of air temperature at 5,000 feet (850mb) at 8:00 am. (larger image)
Figure 5: Analysis of wind speeds at 5,000 feet (850mb) at 8:00 am. (larger image)

Winds in the lowest 5,000 feet were also supportive of waterspout development. At 8:00 am, wind speeds at 5,000 feet (850 mb) were approximately 16 knots, well below the previously discussed threshold of 35 knots. In addition, the wind direction from the surface to 5,000 feet as observed by the Doppler radar station at Cleveland (click here for image) and the ballon observations from Detroit, MI (click here for image) was remarkably consistent, satisfying the threshold that directional wind shear not be too strong.

Identifying the nature of the pre-existing low-level rotation is the most challenging aspect of an analysis of a waterspout. I suspect that a land breeze boundary or an outflow boundary associated with the earlier thunderstorms may have provided the low-level rotation. However, the sparse nature of the observational network simply doesn’t capture the data necessary to reach a definitive conclusion. In some instances, a line of storms on radar or satellite imagery suggests the existence of some type of boundary, but this was not the case on July 28. However, the analysis of surface vorticity (rotation) (click here for image) published by the Storm Prediction Center at 9:00 am on July 28 indicated an elongated area of surface vorticity (in blue) over the southern portion of Lake Erie extending west to the vicinity of the Opti Regatta. And, of course, the waterspout itself is proof that a source of low-level rotation was present.

Figure 6a: KCLE
reflectivity 8:41 am
(larger image)
Figure 6b: KCLE
reflectivity 8:46 am
(larger image)
Figure 6c: KCLE
reflectivity 8:50 am
(larger image)
Figure 6d: KCLE
reflectivity 8:55 am
(larger image)

Figure 7a: KCLE
reflectivity 9:00 am
(larger image)
Figure 7b: KCLE
reflectivity 9:05 am
(larger image)
Figure 7c: KCLE
reflectivity 9:10 am
(larger image)
Figure 7d: KCLE
reflectivity 9:14 am
(larger image)

The series of radar images in Figures 6 and 7 shows the development of the waterspout-producing storm from 8:41 am to 9:14 am. During this same period, the storm underwent significant vertical development, growing from approximately 15,000 feet to nearly 31,000 in the 25-minute period ending at 9:10 am (click here for radar imagery). This dramatic vertical development is consistent with the conceptual model of a waterspout in which a rapidly strengthening updraft transitions to a waterspout. The waterspout, with a lifespan of 30 to 45 seconds, occurred between 9:10 am and 9:14 am as the storm reached its maximum height.  And just 10 to 15 minutes after producing the waterspout, the storm reached the southern shore of Lake Erie and dissipated. But it was a great show while it lasted.