Lake Michigan Ducted Moisture Plume

January 23, 2005

Lake effect snow bands are a routine phenomena over the Great Lakes during the cold seasons. The physical configuration of Lake Michigan (long and thin) and north-south orientation dictate that single plume snow bands are often formed when the synoptic flow is from the north. This maximizes the fetch (terminating at the south shore) and allows mass convergence from the east and west shorelines to focus the plume.

Boundary layer lapse rates near adiabatic (often super adiabatic over the water) and a uniform, lacking in directional shear, wind field in the adiabatic layer are necessary conditions for this phenomena to be sustained. Since these events are usually post frontal, synoptic scale subsidence places an inversion over the mixed boundary layer. This caps vertical growth of the cloud from buoyancy considerations as well erodes the cloud turrets with dry air mixing above the inversion.

Snowfall from these plumes is often maximized in proximity to the shoreline or just inland for perhaps 10-20 nmi. Clouds dissipate further downstream due to mixing and lack of thermal buoyancy support.

In rare cases the effects of these plumes can be observed at a great distance from the originating source. On January 22, 2005 a deepening upper trough and surface cyclogenesis along the east coast of North America created a broad northerly flow over the Great Lakes parallel to the long axis of Lake Michigan. The result was an extended (lake effect) moisture zone which originated over Lake Superior, was maintained over Lake Michigan and persisted well south to beyond the Ohio River. An IR satellite image at 0515 UTC of January 23^rd reveals the extent of the cloud field, nearly 750 mi. Arguably some cloud elements reached northern Alabama and snow flurries were reported as far south as Hopkinsville, Ky and Nashville, Tenn..

A corresponding fog product image at 0500 UTC is accompanied by surface station data. Here the shoreline mass convergence, reverse lake breeze, into the mid lake (Michigan) plume can clearly be noted. Computer analysis of surface stream lines reinforces the lake convergence but also reveals this convergent tendency persists southward along the Indiana-Illinois border.

While one might attribute the extensive length of the plume over land to pure advection one would also expect turbulent mixing within the cold drier air to limit the horizontal extent of cloud existence. In this particular case however there is evidence that the thermal structure of the lower atmosphere played a critical role in maintain the cloud zone for close to 350 miles beyond the south shore of Lake Michigan.

No actual soundings are available during this time but Eta40 (mesoEta) model soundings were used to reconstruct the atmospheric profile along and in proximity to the cloud zone. The sample points are noted in this image.

Over the open water at Point-A around 0600 UTC the model sounding reflected the expected steep super adiabatic lapse rate in the boundary layer capped by an inversion at 900mb. The airmass was indicated to be very dry above the inversion. Wind profile showed a slight veering right off the surface but uniform winds in direction and speed above the immediate surface through the inversion. Whether the veering was actually present in reality is unknown.

Moving downstream at this time (0600 UTC) Points D and E sample the area where satellite images show the clouds to be present. Both points are very similar with a sharp inversion height around 825mb, having risen from 900mb over the Lake. The moist / dry discontinuity persists through the inversion. Observed cloud heights through this zone averaged around 4000ft agl which fits within the moist layer shown. On a plan view, the WSeta analysis of the 875mb level at 0600 UTC also showed, quite remarkably, a well organized zone of maximum RH extending from southern Lake Michigan southward along virtually the entire length of the Indiana-Illinois border. Compare this with clouds seen in the 0500 UTC fog product.

Looking on either side of the cloud zone, Points B and C we see a much more chaotic profile. Assuming these have some bearing on fact, the inversions, if they exist, are poorly defined. Moisture is more diffuse although the wind profile remains uniform.

In another view, a north-south cross section from the Lake south along the plume zone highlights the omega (vertical motion) forced by the water trapped under a large area of synoptic subsidence (solid lines). Subsequently three hours later at 0900 UTC we note the vertical motion has expanded to the south a little. The upward slope of the inversion from north to south, best seen in the rise of the RH color image, reflect the rise in the inversion seen in the individual soundings earlier.

Clouds became less organized after around 0700 UTC and the plume faded as an organized entity well before 1200 UTC. Eta40 model profiles at 1200 UTC from the primary points D and E show much less structure than 6 hrs earlier with the northern point (blue in the image) becoming very stable toward the surface under the influence of somewhat colder air.

Extended length plumes such as this, while rare, are not unknown. The source region in this case, Lake Michigan, was augmented by the advection of low level moisture from eastern Lake Superior. Once over land the normal mixing out of the moisture was inhibited but the bounded thermal structure of the inversion around 875mb. The steep lapse rates below the inversion coupled with an ample moisture supply within the zone and subsequent small but not trivial CAPE values noted in the soundings (dashed lines in boundary layer) were sufficient to maintain rising parcels which saturated and were then trapped. The mass convergence which extended well south of the Lake as seen in the surface stream lines were likely another mechanism to sustain slight rising motion within the cloud zone (mass continuity). Once initiated cloud parcels were thermally bound (ducted) within the inversion layer until mixing from the sides and some from above took its toll.

Thermal structure conditions either side of the cloud zone were not as conducive to maintaining a trapped cloud layer. One can speculate, however, whether the presence of added moisture from the Lake plus gentle mass convergence was sufficient to modify the thermal profile to better support the observed conditions. In other words there was nothing intrinsically different between the airmass over central Illinois (Point C) or central Indiana (Point B) verses that in the cloud zone (Points D,E) except for the mass convergence and injection of lake sourced moisture.