Lake Champlain Hydrodynamics — the seiche and the layers
loading the latest lake data…
All summer Lake Champlain is divided into two layers: warm water on top, cold water beneath, and a thin boundary between them that will not mix. Wind piles the warm layer against one end of the basin, the boundary tilts, and when the wind drops the whole thing rocks back — a wave tens of meters tall that takes days to cross the lake and never breaks the surface. A summer of it, measured:
A wind blowing down the lake pushes the whole water column the same way, top to bottom. That is most of what the current meters see, and underneath it the internal wave is invisible: raw, the current at 7.5 m and the current at 27.5 m move together, +0.82. Subtract the average of all ten meters — the part of the flow that is the whole column moving as one — and what is left is the wave. Above the boundary and below it, the water runs in opposite directions.
Smoothed over a day, because everything faster than that — gusts, the meters' own noise — is not the wave and hides it. The correlation quoted is on the unsmoothed series.
Periods present in the boundary's rise and fall. The peak is broad because the thing driving it — weather — is not periodic; the lake has a natural period and the wind sets it going at irregular intervals.
The surface seiche tips the lake by inches. This one moves the boundary between warm and cold water by tens of meters, and everything that lives in the water column moves up and down with it: the depth a fish finds its preferred temperature, the depth cold nutrient-rich water reaches, whether the warm surface layer touches a shallow shelf or hangs above it.
It is also the one part of the lake's physics that a shore observer cannot see at all. Without a string of thermistors hanging under a buoy, there is no way to know it is happening.
The two layers running in opposite directions is the wave. What holds them in that arrangement is the rotation of the earth. model.py leaves it out: the second line states no rotation. The argument for why that matters has so far been a length scale: the internal Rossby radius here is about 4.7 km (2.9 miles) against a basin 11–16 km (7–10 miles) wide, so the internal seiche ought to be a rotation-influenced, shore-trapped wave rather than the flat one-dimensional standing wave the model draws.
This site's own record starts in 2007. The surface seiche was measured long before that, on paper charts, and the measurement survives: Myer and Gruendling's Limnology of Lake Champlain (1979) tabulates the periods seen at six stations around the lake and names the mode each one probably belongs to. Their observations are the marks below. The curve behind them is the same quantity measured the modern way — twenty years of hourly gauge data, the lake's end-to-end tilt, with its own background noise divided out so that a mode can be seen against it at all.
The buoy sees one place in fine detail and the monitoring series below sees seven deep stations for thirty-four summers. In 1969 Gruendling measured twenty stations from a five-meter (16-foot) bay to ninety meters (295 feet) of open water, and the result is the clearest statement of the thing that governs everything else here: the depth of the station.
The state monitoring program has been lowering a sonde down the same deep stations since 1992, a few times a month — far too coarse to catch a wave that takes six days, and the only way to ask whether the layering itself is changing. It is. The surface of the lake in July and August has warmed +0.47 °C a decade while the water at the bottom has not moved at all, so the two layers are diverging in temperature: the difference across the boundary has grown +0.54 °C a decade. A more strongly layered lake mixes less, holds its warm surface longer, and favors cyanobacteria.