Meteorological Review of the Thanksgiving
The first snow storm of the 2004-2005 winter season for northeast Illinois occurred during the afternoon and early evening of November 24th, Thanksgiving eve. Aside from up to 5 inches of wet snow in the western suburbs, strong northeast winds resulted in power outages especially over southern Cook and eastern Will counties of Illinois and portions of northwest Indiana.
From the meteorological and forecast perspective this event presented a number of challenges but also displayed some of the classic signals seen in synoptic scale weather systems. The following discussion will attempt to highlight a few of these items.
Background Synoptics and Concerns:
As the week began, one in a series of upper troughs which had swung through the southwest states was ejecting into the southern Plains. An enhanced water vapor image at 2315UTC (515CST) on Tuesday shows this feature as it works into west Texas. Thunderstorms were ongoing from southern Missouri southward into east Texas. The upper low was forecast to initiate a surface low over Oklahoma which would then move northeast along the Ohio River on Wednesday. A composite map of forecast tracks from the two primary forecast models (GFS and Eta) for a 24hr period beginning at 0600UTC on the 24th displays the uniformity of the forecasts.
Indeed by 1200UTC of the 24th a surface low was developing over northern Arkansas. Moist air was spreading into the Ohio Valley. The picture was a bit complicated by a weak cool front which had slipped through northern Illinois Tuesday afternoon. While significant cooling did not occur with the passage of the front it did serve to act as a source of cooler and drier air just to the north in Wisconsin and was associated with an upper level jet which came into play later.
From a forecast point of view some of the uncertainties were:
Model guidance going into Monday and Tuesday had been reasonably consistent tracking a surface low along the Ohio River during the day Wednesday. The forecast surface positions of both the Eta and GFS between 24/06UTC and 25/00Z are in fairly close agreement but are biased to the south (and too fast) of where the system actually moved.
In attempting to assess where the anticipated precipitation band would occur one would determine the processes expected to be critical and how they evolved. In this case the banded forcing to the north of the system in a zone where temperature profiles indicated maximum snow flakes would be created was a reasonable place to begin. The zone around -13C to -17C is often a good place to look. The meso-eta 24hr forecast valid Wednesday noon is representative of model output for this event. Of interest is the frontogenic forcing in the 500-600mb layer (~15000ft-18000 ft) which would suggest where the atmosphere was most productive. The layer was chosen using a time section display for ORD (Eta model).
The display shows a band of strong forcing from west central Illinois into the Chicago area. The red lines are potential temperature lines which are compressed within the band, as they should be. Winds in the 500-600mb layer are also strongly convergent both in both direction and deceleration. Yellow contours are the forecast surface isobars with the low center approaching southwest Indiana. Later model guidance using the same parameters was similar for the afternoon period of the 24th.
Another common factor in organized precipitation bands is the upper divergence associated with dual jet cores. A 12hr 300mb forecast from the meso-eta valid at 1800UTC of November 24th reveals two jet maxima. The first is pulling away over Lake Superior and is associated with the weak cool front mentioned earlier. The second is working into the Ohio Valley ahead of the short wave trough pushing into the mid Mississippi Valley. Between the two dynamic forces are creating a zone of divergence which facilitates deep upward motion in the atmosphere.
The signals appears to be lining up but the type of precipitation or timing of change from rain to snow was still an open question.
Storm Evolution and Results
The short wave moved steadily northeast as forecast. However, the track was a bit further north and it maintained a vigorous core circulation vs weakening. This can be seen in the water vapor imagery at 1900UTC of November 24th. The circulation can clearly be seen over eastern Illinois with a swirl extending back to just east of St. Louis. Thunderstorms, some severe, were beginning over southern and central Indiana. A few lightning strikes can be seen as far north as Kankakee and Bloomington.
Visible imagery at 2125UTC is even more striking. The center of the circulation appears to be just south of Champaign. The low sun angle highlights the convective towers over Indiana as well as those around the center. It is useful to monitor the cooling of cloud tops to get a perspective on whether precipitation intensity and coverage is expanding. The 1800UTC IR satellite image does indeed show cooling cloud tops across northern and central Illinois.
Rain was falling across northern Illinois during the morning and early afternoon of the 24th. It had turned to snow by late morning around Peoria and southwest into northern Missouri where up to 6 inches was reported. During the early afternoon dynamic and thermodynamic forcing (cooling at mid levels) was turning the rain to snow over much of northeast Illinois. A 3-hour loop from 1918UTC to 22UTC from the KLOT radar is when this change was ongoing. The blue in the loop reflects stronger radar returns. Two items of note. A nearly stationary band of enhanced return is positioned from LaSalle County northeastward into northern Cook County. Also, around 1900UTC there is a flurry of lightning activity. None appears concentrated or related to any specific echo feature. It does perhaps reflect the intensity of the rain/snow falling.
In a broader view a composite radar/surface plot at 2100UTC clearly shows the circulation west of Indianapolis and its relation to the frontogenic band over northern Illinois. In another aspect, a reflectivity cross section through the KLOT site at 2000UTC shows echo tops not much above 20000 ft but some pockets of enhanced elevated return.
The reality of the situation as it unfolded was that the expected strong forcing mechanisms over northern Illinois did materialize. The mid level circulation maintained a vigorous state for longer than anticipated and was north of the forecast track. This sequence resulted in a both an adjustment north of the primary precipitation zone and, by virtue of strong vertical motion, a rapid change to snow during the early afternoon and with greater intensity. The intensity of snowfall over a 3-4 hour period during the later afternoon negated any affect of warm ground temperatures. It simply snowed too hard to melt efficiently.
The analyzed snowfall map from this event reflects a couple of the other concerns [dashed lines are a smoothing attempt to account for lack of snowfall reports]. The strongest reported winds during this sequence were from Gary and Calumet Harbor. A minimum zone of snow accumulation from Lake County Indiana and southeast Cook County southwestward likely reflects the influence of the warm water from Lake Michigan. Additionally there is a marked snowfall gradient from near Quad Cities to McHenry County. A visible satellite image the next day highlights the demarcation. Such a pattern was expected given the nature of the forcing. The question always is where the band will be.