Tale of The Sheriff and A Seiche
Regular readers of my Database and Tech blog may remember last year's post on standard deviation that was built in part around an example involving seiche events on the great lakes. Yesterday a notable event took place in Munising. I completely missed seeing it live, but Alger County Sheriff Robert Hughes has kindly shared some photos.
(Click any image in this post to embiggen it)
The cause of all this watery ruckus was a severe storm system moving across the western U.P. and Lake Superior region. The following images show the morning forecast (left) and a close-to-noon radar image (right). The storms were strongest in the west and weakened as they moved toward the east.
A seiche is when water in a lake or similar basin sloshes back and forth. You get more than one slosh, and there's time if you hear about an event to get out and see it. Tuesday's event generated sizable differences in water level throughout much of the day, as you can see in the following plot of observed water levels from the National Oceanic and Atmospheric Administration's Marquette observation station. (The observation station is located on Coast Guard Station grounds).
Peak recorded lake level on this plot is 603.205 feet above mean sea level at 2:36 PM on the 22nd. The lowest recorded level is 601.906 feet some three hours later at 5:54 PM. That's a spread of 1.299 feet as recorded at the Coast Guard station in Marquette, Michigan. You can see how there is pretty good up and down activity from about 10:00 AM through 6:00 PM on the 22nd, with reduced activity continuing through the night and tapering off in the early morning hours.
Seiches can be triggered by sustained winds that push water in one direction, allowing it to slosh back when the winds die down. Seiches can also be triggered by changes in air pressure brought on by storm fronts. The next two plots show wind speeds including direction and gusting, and then barometric pressure. These plots cover the same period as the prior plot showing water level.
The wind speed plot seems to correlate best with the water level plot. It's worth noting that the low point on the pressure plot corresponds with the beginning of the heaviest period of seiche activity. It seems pretty clear that winds were a cause of Tuesday's event. I lack the knowledge to say whether the air pressure changes were also a factor.
Before you go, here are two more images from a dock just a few dozen feet away from the boat launch as you move toward the Shipwreck Tours office. The first image (on the left) shows one of the high water levels. The second image is from today with no seiche activity involved.
Want to learn more about seiches? MSU professor Randall Schaetzl maintains a nice page on the topic. The Lake Erie Seiche Disaster of 1844 is an interesting read. And dangerous seiches occur in more modern times too.
My sincere thanks to Alger County Sheriff Robert Hughes for his kindness in providing the seiche photos for this post.