A Good Forecast
Forecasting the location, timing, and nature of severe weather is a very difficult task. A comparison of the Day Two Probabalistic Outlook (figure 1) accompanying the SPC Day Two Convective Outlook issued on Saturday, July 16, 2011 and the SPC Storm Reports (figure 2) confirms the SPC did an extraordinary job in identifying the area expected to be under the greatest risk. The text of the Outlook mentioned the potential for the development of severe weather over the upper Great Lakes and mentioned that "should storms form...supercells with large hail...damaging winds and perhaps a tornado will be possible". There were no tornadoes reported across upper Lake Michigan, but the damaging wind forecast certainly verified.

Figure 1: Day Two Probabilistic Outlook issued by the SPC on Saturday, July 16, 2011. From the SPC.
Figure 2: SPC Storm Reports for Sunday, July 17, 2011. Blue dots mark the location of damaging wind gusts. From the SPC.

This storm system also dispels the common misunderstanding that all squall lines, or other modes of severe weather, are always accompanied by the passage of a low pressure system or the system's associated surface boundaries. The storms that swept across Lake Michigan formed in two distinct areas characterized by the presence of subtle surface boundaries and their interaction with an extremely unstable environment. Both areas of development were supported by lake breeze boundaries, particularly those storms that formed near Marquette, Michigan. Once the storms formed, their movement was influenced by the steering currents through a deep layer of the atmosphere and dynamics within the individual storms -- not the passage of a surface boundary such as a cold front. At 8:00 am (1200Z) on Monday morning, approximately 8 hours after the passage of the system, the low pressure system and cold front were still north of the area (link).

Widespread Strong Surface Winds
Moving at an average speed of approximately 30 knots, the system crossed Lake Michigan in slightly more than ninety minutes (radar loop). Tragically, its passage brought two weather-related fatalities, the first in the race's 103 year history. Damage to sails and rigging and participation in search efforts, prompted the withdrawal of approximately 10% of the fleet. The majority of the damaged boats and the two fatalities occurred in the area between Point Betsie and Beaver Island.

The system produced damaging surface winds across a wide area encompassing northern Lake Michigan (figure 1). The winds from a strong thunderstorm may dramatically affect a relatively small area and contribute to considerable variations in maximum wind observations across a region. The relatively sparse distribution of automated weather stations produces significant gaps in data and may contribute to an incorrect assessment of the system's fury.

Figure 1: Observed wind gusts from SPC storm reports and automated weather stations. (larger image)

A collection of high wind observations taken from 11:50 pm (0350Z) to 12:35 am (0435Z) surrounded the region near the Fox Islands. The strongest gust of the group -- 64 knots-- was reported at Charlevoix, MI at approximately 12:35 am (0435Z), and was also the highest gust produced by the system. The remaining observations in the area were 35 knots or higher, including the 45 knot gust at Buoy 45002. Most of the unfortunate Chicago-Mac participants who found themselves in this area reported wind speeds consistent with these observations in the range of 40 to 60 knots.

A Persistent Downdraft
Based upon an interpretation of the radar data and the 64 knot observation at Charlevoix, MI, it is likely that the strongest winds occurred in a relatively narrow corridor extending from the vicinity of the Fox Islands to Charlevoix, and were produced by a persistent downburst from an approaching thunderstorm. (A series of radar images documenting the storm's movement towards the coast may be viewed here.) The base reflectivity (figure 2) and base velocity imagery (figure 3) at 12:33 am (0433Z) show the leading edge of the thunderstorm and the pocket of strong winds reaching Charlevoix just two minutes before the maximum wind gust was observed. The base velocity image (figure 2) shows additional areas of strong winds associated with the leading edge of storm, but none were as strong or as widespread as the area impacting Charlevoix.

Figure 2: Base reflectivity 12:33 am (0433Z). (larger image)
Figure 3: Base velocity 12:33 am (0433Z). (larger image)

Downbursts may result when momentum from strong winds aloft are transferred to the surface by a thunderstorm downdraft. However, a review of the observed soundings at 8:00 pm (000Z) from Green Bay, WI (figure 4) and Gaylord, MI (figure 5) indicates that the winds up to nearly 500 mb (18,000 feet) were relatively light, in the range of 20 to 25 knots. Assuming that the atmosphere over the Fox Islands was similar to that over Green Bay and Gaylord, the transfer of momentum from aloft was not solely responsible for the strong winds near Charlevoix.

Figure 5: Observed sounding from KGRB at 8:00 pm (00Z). (larger image)
Figure 6: Observed sounding from KAPX at 8:00 pm (00Z). (larger image)

The observed soundings at Green Bay and Gaylord (figures 5 and 6, respectively) are similar to the conditions favoring the formation of a wet microburst, a strong downburst accompanied by significant precipitation. This environment is characterized by a layer of relatively dry air aloft overlaying a nearly saturated layer closer to the surface (annotated images). Both soundings indicate that the relative humidity between 8,000 and 14,000 was low, particularly at Green Bay where values averaged approximately 25%. This layer of dry air adds to the negative buoyancy associated with the downdraft by increasing the potential for raindrops and melting ice to evaporate, thereby enhancing cooling within the downdraft. The presence of a prolonged downburst accompanied by heavy precipitation is supported by many Chicago-Mac racers who reported very strong winds and low visibility. Along with the strong wind observation, the weather station at Charlevoix reported very heavy rain and visibility of 1.5 miles at 12:35 am (0435Z).

The data from the base reflectivity and base velocity radar imagery suggests that the waters west of Charlevoix, MI were buffeted by a combination of strong winds from the bowing segment to the west-northwest and a strong downdraft from a supercell thunderstorm to the north. Wind speeds of 45 to perhaps 70 knots are supported by radar data and observations. Although the weather station at Charlevoix observed 64 knots, it is possible that friction over land diminished the wind speeds associated with the land-falling storm and that this peak observation was somewhat less than what would have been encountered by the racers offshore.

In addition to very strong winds, many of Chicago-Mac racers reported wave heights as high as 6 to 8 feet, a significant difference from the observations at Buoy 45002. Of course, the observations at Buoy 45002 represent the conditions at only one specific point and the racers were spread out over a large area. For those competitors who were much closer to the Michigan shore, it is possible that the additional fetch supported the development of much larger waves. It is also possible that the combination of the winds from the bowing segment and the supercell downdraft combined in the area to produce significantly larger waves.

Veteran Chicago-Mac racers have encountered similar conditions during previous races and understand that coping with severe weather is part of the challenge. Skippers must prepare their boats, train their crew, maintain a watchful eye for approaching storms and "the dearest friend (and most menacing foe) of all sailors -- the wind."

About The Author

Mark Thornton has been sailing for more than 30 years and currently owns Osprey, a C&C 35. His interest in weather forecasting grew from his experiences cruising and racing on the Great Lakes. Mark is a 2006 graduate of the Penn State University Certificate of Achievement in Weather Forecasting, a two-year program that develops skills in general, tropical, and severe weather forecasting.

He is the president of LakeErieWX Marine Weather, a company dedicated to providing marine weather education and forecasting resources for recreational boaters ( Mark publishes a marine weather blog and teaches marine forecasting seminars to recreational boaters during the off-season. He has served as the Race Meteorologist for the Bayview Mackinac Race since 2014. Mark is also employed as a Teaching Assistant in the Certificate of Achievement in Weather Forecasting Program at Penn State University.

Click here to read about my marine weather forecasting seminars, webinars and presentations. Click here for the current schedule of educational programs. My complete resume can be viewed here.