Lake Effect in South Dakota

November 28th, 2007 |


(Image source: The CIMSS/SSEC MODIS Today Page)

Lake effect clouds and snows are a common wintertime occurrence downwind of the warm waters of the Great Lakes, and such weather events have been discussed frequently on this blog (See here, here, and here, for example) and elsewhere. When cold air moves over a warm water surface, the air is warmed via convection and moistened via evaporation, and that input heat and moisture can be sufficient to generate cumuliform clouds downwind of the moisture source. If Lake waters are about 13 C warmer than the temperature at 850 mb, then clouds and snow occur downwind of the Great Lakes.

Lake Oahe is a large artificial impoundment of water on the Missouri River in central South Dakota. It is sufficiently large that, given the right wind trajectory, clouds can form downwind of the lake. Even in late November, it retains some warmth, as noted in the MODIS-derived “Sea” surface temperature product here. Note the pixels that suggest water temperatures in the 40s in the region where the clouds formed today.

A loop of the GOES-12 (GOES-EAST) imagery from today clearly shows the region where the Lake Effect clouds formed. Note how the cloud streets align themselves with the wind direction as reported at PIR (the station in central South Dakota); the other stations are Chamberlain to the southeast and Mobridge to the northeast. As the temperatures warm during the day, the thermal differences that support the cloud development weaken and the cumuliform clouds dissipate as mid-level clouds move in from the west.

Another intriguing feature in the MODIS true-color imagery above is the smear of white near Fort Thompson, just north of the southeast corner of the image. For a chilly west-northwest airflow, this region is just downwind of the prominent oxbow on the Missouri River near Fort Thompson and Lower Brule, to the north of Chamberlain and to the east of Pierre. Is this a signature of Lake/River effect snow in that area?

Lake-effect snow and strong winds over the UP of Michigan

November 27th, 2007 |

GOES-12 6.5 µm water vapor channel images (Animated GIF)

Cold air in the wake of a strong cold frontal boundary and dynamics associated with an intense shortwave trough aloft contributed to an outbreak of lake-effect snow and very strong winds over the Upper Peninsula (UP) of Michigan on 27 November 2007. AWIPS images of the GOES-12 6.5 µm “water vapor” channel (above) revealed a dynamic dry signature (darker blue enhancement) associated with a potential vorticity anomaly that was moving southeastward across northern Minnesota and the UP of Michigan. The dynamic tropopause — taken to be the pressure of the 1.5 Potential Vorticity Unit surface (below) — appeared to be as low as about 675 hPa in the vicinity of the dry water vapor image feature.

GOES-12 water vapor image + NAM12 model fields

A north-to-south cross section using NAM12 model fields through the PV anomaly feature at 12 UTC (below) showed a well-defined tropopause fold over the UP of Michigan, with the dynamic tropopause actually extending downward to below the 700 hPa level.

North-to-south NAM12 cross section

Surface wind speeds exceeded hurricane force over portions of the UP during the morning hours, with a gust to 71 mph at Copper Harbor at 6:50 AM (12:50 UTC) and a gust to 74 mph at Stannard Rock Lighthouse (on the Keweenaw Peninsula) at 9:00 AM (15:00 UTC). GOES mesoscale winds overlaid on the MODIS 11.0 µm “IR window channel” image at 17:09 UTC (below) indicated cloud-tracked wind speeds as high as 54 knots (62 mph) over the UP.

MODIS IR image + GOES mesoscale winds

Looking at a comparison of a few MODIS images and products at 17:09 UTC (below), we can see that the visible channel showed the widespread lake-effect snow (LES) bands covering much of Lake Superior, extending southward across the UP and even over extreme northern Wisconsin and northern Lake Michigan; the IR window channel brightness temperatures along many of the LES bands were colder than -20ºC (light blue enhancement), suggesting that cloud particle glaciation may have begun; however, the LES bands exhibited a generally “bright” appearance on the near-IR “Snow/Ice” channel, leading one to suspect that the bands might still composed primarily of supercooled water droplets; finally, the Cloud Phase product indicated that the majority of the LES cloud features were likely of the “Mixed Phase” category (darker gray enhancement), so although LES band glaciation may not have been complete, many of the bands probably contained a good amount of ice crystals (which would be necessary for snow to fall at the surface).

MODIS images and products (Animated GIF)

Finally, let’s take a closer look at the MODIS true color imagery over the surrounding region — some very interesting features are evident once we zoom in and take advantage of the 250-meter resolution of MODIS data using the SSEC MODIS Today website. A close-up view over northern Minnesota (below) reveals a long, narrow gravity wave feature along the leading edge of a thin cloud deck; also note that ice formation in Upper Red Lake (just northeast of the cloud edge and gravity wave structure) is well underway — not surprising, given that the air temperatures in that area were below 0º F (-18º C).

MODIS 250m resolution true color image

Farther to the southeast, a close-up view centered over central Wisconsin (below) shows the mesoscale “banded” nature of the snow on the ground over that area (especially along the southern periphery of the snow cover). The snow depth in these “streaks” was probably only about 1 inch or less (judging from the cooperative observer snow depth reports that were only a Trace across that particular region), but the snow streaks really stood out against the surrounding areas of bare ground. Such mesoscale “snow steaks” are not uncommon to see on satellite imagery following “light” snowfall events — these satellite signatures help to underscore the difficulty in forecasting snowfall accumulation amounts over any given location.

MODIS 250m resolution true color image

Wave structure on top of a stratus cloud deck

November 20th, 2007 |

GOES-12 visible images (Animated GIF)

GOES-12 visible channel images (above) revealed
an interesting wave structure along the top of an extensive stratus cloud deck
that covered much of Iowa, southern Wisconsin, and northern Illinois on 20
November 2007
. An AWIPS 4-panel image showing the MODIS and GOES-12
visible and IR window channels (below) demonstrated
the better wave detection capabilities of the higher spatial resolution MODIS
data. The GOES-12 and MODIS IR brightness temperatures in the region of the
wave signatures were generally in the +1ºC to +5ºC range, with the GOES
Sounder Cloud Top Height
indicating tops around 4700 feet in that
area (tan enhancement); the MODIS
Cloud Phase product
confirmed that the cloud in that region was
likely composed of supercooled water droplets (blue enhancement).

MODIS and GOES visible + IR images


GOES-12 visible image + NAM 850mb isotherms

Much of the wave structure on satellite imagery seemed to be located along
a southwest-to-northeast oriented baroclinic zone (indicated
by a tighter packing of the 850 mb isotherms)
, with the individual banding
elements oriented generally perpendicular to the axis of the baroclinic
zone (above); however, radar echoes that developed
a few hours later were generally aligned closer with the axis of the baroclinic
zone (below). A northwest-to-southeast cross section
of NAM12 model output (along
line D-D’
orthogonal to the baroclinic zone axis)
revealed elevated
pockets of frontogenesis and omega (within
the 600-850 mb layer)
which may have played a role in the formation of
the regions of banding seen on both satellite imagery and radar reflectivity.

Radar reflectivity + NAM 850mb isotherms


MODIS 250m resolution true color image

A 250-m resolution MODIS true color image (above) from
the SSEC
site shows the cloud top waves in great detail over
northeastern Iowa. Note that some of the wave structure and orientation (just
south of the Iowa/Minnesota border) was similar to that seen on the radar images.

Snow cover in New England

November 19th, 2007 |

MODIS true color + false color images (Animated GIF)

A comparison of the MODIS “true color” RGB image (Red=channel 01, Green=channel 04, Blue=channel 03) and the corresponding “false color” RGB image (Red=channel 02, Green=channel 07, Blue=channel 07) from 19 November 2007 (above) shows snow cover over parts of New York, Vermont, New Hampshire, and Maine (extending northward into portions of southern Quebec in Canada). Both snow cover and clouds appear white on the true color image, but deep snow cover appears as darker shades of red (with clouds composed of ice crystals appearing as a lighter shades of red) on the false color image — this makes it relatively easy to discriminate snow cover from supercooled water droplet clouds (which appear as shades of white on the false color image). In fact, a few small patches of supercooled water droplet cloud can be seen over the region of deeper snow clover (along and just north of the US/Canada border). Snow depth data from the NOAA National Operational Hydrologic Remote Sensing Center (NOHRSC) indicated a number of sites reporting 5-10 inches (13-25 cm) of snow on the ground that morning.