This chapter covers the storm as we experienced it here in N DE. Snow started to fall early in the morning (about 5 am EST) Sunday February 16, 2003 and ended here about 11:30 am Monday February 17, 2003. Total accumulation for Bear, DE was 24 inches. There was a brief period of light snow Tuesday morning February 18 caused by the upper level low moving through the mid-Atlantic.
A. Early Sunday Morning – February 16, 2003 - 0100 (local) to 0700 (local)
On Sunday morning the snow began to fall, and it didn’t begin gradually. By the time I got up, around 6:30, it was already snowing heavily. I estimate it began here in N DE about 5:00 am, and it started much sooner than even the most recent forecasts indicated.
Figure 18 is a photograph taken from my house at 0730 EST this morning. There is already a good 3-4” of snow accumulation, and the snowfall rate has severely limited visibility. This photograph and its viewpoint will be used frequently as a reference throughout the storm.
Let’s take a look at the analysis charts for 12Z (0700 EST) on Sunday Feb 16 to get some idea of where all this snow is coming from, and perhaps an answer to the question “Why did snow start here about 6 hours before even the most recent forecast”?
The maps shown in Figure 19 represent observations at 12Z (0700) on Sunday February 16. The dominant features on the surface (Figures 19d and 19e) are the low in AL/MS and the strong high N of Lake Ontario. Note (Figure 19e) that the strong flow around the high has pushed cold air at the surface into what would normally be a warm front to the E of the low into a cold front pushing in to GA. Also note, on the 500 mb (Figure 19a) and 700 mb (Figure 19b) maps, that while the warm moist Gulf air is not advancing to the NE on the surface, it is moving over the cold air mass in the upper air flow. This is the primary cause of the current snowfall - the warm moist air being forced to rise over the cold surface air, resulting in moderate to heavy snow. The surface temperature in N DE at this time was 12F, extraordinarily cold for a snow event.
Upper level support for the surface low is indicated by the closed lows at 850 mb and 700 mb, and the short wave trough at 500 mb. This upper level flow pattern is causing strong easterly winds at 850 mb and strong (relatively warm and moist) SW flow at 700 mb and 500 mb. The closed contours on the 700 mb chart are 70% and 90% relative humidity; note that the winds at 700 mb are transporting that moisture over the cold air that is dominating the NE at the surface. So, although the extremely cold air we are feeling at the surface (12 deg F) originated in Canada, the snowflakes falling on us are coming from the Gulf of Mexico!
Figure 20 is the 12Z temperature sounding from Dulles Airport, which is the most representative sounding location for N DE. There are some interesting features to note on the sounding. First, there is very cold air near the surface with a surface temperature of -9.3C (15F). Temperature decreases to -14.1C (6.6F) at 950 mb (about 2300 ft); above this height, the warm, humid air is quite evident. Temperature increases from 950 mb (2300 ft) up to 635 mb (12,500 ft), and the air is essentially saturated in this layer (the dew point is near or equal to the temperature). Potential temperature (see Sections II.C and II.G) is increasing with height at a rapid rate (this is called a frontal inversion), and with strong SW winds aloft, this indicates significant vertical velocity, adiabatic cooling, and precipitation. And the temperature is well below freezing at all levels of the atmosphere, which means precipitation is all snow.
Figure 21 is a chart showing potential temperature and wind flow at 700 mb for 12Z Sunday Feb 16. With a strong potential temperature gradient over the mid-Atlantic, and 700 mb winds having a component across the isotherms (from higher to lower potential temperature), isentropic analysis (Section III.G) suggests precipitation over this region. Note on Figure 21 that where the 700 mb wind has a component across the potential temperature lines there is snow falling, whereas where the wind is nearly parallel to the potential temperature lines, there is no snow (from mid-PA and mid-NJ northward).
This isentropic situation with the warm, moist air over-running the cold surface air was to remain in place for about 24 hours, and is an excellent way to view the dynamics of this part of the storm. Here in N DE, we had about 20 inches by Sunday night during the time this situation prevailed. Even though the surface low did not intensify as much as the models predicted, this “isentropic lift” potion of the storm was enough to make it an historic event.
Figure 22 is the 600 mile Doppler radar image from the Weather Channel. It is portraying the situation at 0645 Sunday February 16. There is heavy snow occurring throughout the state of Delaware.
B. Late Sunday Morning – February 16, 2003 - 0700 (local) to 1300 (local)
The 6-hour period from 0700 EST (12Z) until 1300 EST (18Z) Sunday was the period with by far the heaviest snowfall rate, with about 9 inches falling in this 6-hour period. Figure 23 is a photograph taken at 10:30 EST, when the snowfall rate was about 2” - 3” per hour. Snow accumulation is about 8 inches at this time.
The maps on Figure 24 represent surface conditions at 18Z (1300) Sunday. Since upper air observations are only taken twice per day (at 00Z and 12Z), there are no upper air charts for the “off” hours of 06Z (0100 EST) and 18Z (1300 EST). On the surface, the high in Canada is still pouring cold air into the NE, and the first hint of a coastal low can be seen off the coast of SC. The station report for Philadelphia (Figure 24a) indicates heavy snow with the temperature remaining at 12 F.
C. Sunday Afternoon – February 16, 2003 - 1300 (local) to 1900 (local)
Snowfall rate decreased somewhat during this time frame; an estimated total of 4 inches fell from 1300 to 1900 Sunday (local time). Figure 25 shows conditions in Bear, DE at 4:25 pm EST; there is well over a foot of snow at this time. Examination of the weather maps in Figure 26 reveal several significant items. First, compare the 500 mb observation from 1900 local time Sunday (Figure 26a) with the 500 mb observation 12 hours earlier (Figure 19a). Noting the trough in the 500 mb height field on both figures, there is a noticeable increase in amplitude and decrease in wavelength, both indicators of intensification. Again comparing Figures 26 and 19, the closed lows at 700 mb and 850 mb and at the surface have moved very little over this 12 hour period. This slow movement of the entire system is a major reason for the heavy snow accumulation.
On the surface maps, the high over Canada is still strong (Figure 26d and 26e), and the surface observation at Philadelphia (Figure 26d - not visible in reduced size map) indicate a temperature of 13 F and moderate snow. The Weather Channel surface map shows continuing snowfall from southern Missouri through NJ, and the northern edge of the snowfall has moved from southern PA/NJ at 0700 EST to encompass virtually all of PA/NJ at 1900 EST. Also the coastal low has started to take over, and is producing heavy rain in eastern VA/NC/SC.
Figure 27 is the Feb 17 00Z (1900 EST Sunday Feb 16) temperature sounding from Dulles Airport. Note that warm air continues to over-run the cold shallow surface air. Temperature at about 720 mb is close to the freezing mark (0 deg C), but precipitation should still be all snow with this temperature profile. The warm air (925 mb up to 600 mb) is completely saturated, and the steep slope (from left to right) of the temperature sounding in this range indicates a steep potential temperature gradient as well. With 700 mb winds continuing to come from the SW, isentropic lift is still very strong (see Figure 28).
Figure 28 shows the 700 mb potential temperature and wind fields for 00Z Feb 17 (1900 Feb 16). Figure 29 is the 600 mile Doppler radar image from 10:22 pm Sunday. The strong potential temperature gradient remains over the mid-Atlantic, and 700 mb winds are now blowing more directly across the isotherms than was the case at 12Z (compare Figures 28 and 21). Note also that the region of positive lift (where the wind arrows are crossing the potential temperature isotherms) has moved farther north in the last 12 hours, as has the region of snowfall (see the Weather Channel surface maps in Figures 26 and 19).
D. Sunday Evening – February 16, 2003 - 1900 (local) to 0100 (local)
We received another 4 inches of snowfall accumulation during this 6 hour period. The photograph in Figure 30 was taken at 10:00 pm EST Sunday evening, at which time there was nearly 20 inches of snow on the ground, and it was still snowing fairly heavily. The surface map (Figure 31) indicates a developing low off the coast of NC, and wind is still from the north over N DE. At this point, with the coastal low just beginning to develop, we have already received about 20 inches of snow here in N DE, a result mainly from the over-running (isentropic lift) of moist Gulf air on top of the extremely cold arctic air mass forced into the NE by the Canadian High.
E. Early Monday Morning – February 17, 2003 - 0100 (local) to 0700 (local)
The picture in Figure 32 was taken at 0700 Monday February 17; it is not snowing at this time. Snow accumulation during this 6 hour period was only about 1 inch. One reason for this is that there was a “dry slot” over N DE during much of this 6 hour period (see the WC Surface image for 0546 2/17/03 in Figure 33e and Figure 36). There were brief periods of sleet overnight as the Canadian high is starting to move out and the coastal low is becoming the primary surface feature (see surface map in Figure 33d).
By early Monday morning the surface winds were out of the NNE and the surface temperature has risen to 21F, although there some brief periods of sleet overnight.. Examination of the 850 mb chart in Figure 33 indicates strong SE winds over DE and the 0C isotherm over S DE at this time, and the Dulles 12Z sounding (Figure 34) indicates temperatures very near the freezing mark at about 800 mb.
Figure 34 is the Feb 17 12Z (0700 EST) temperature sounding from Dulles Airport. There continues to be significant warm air intrusion from about 925 mb to 700 mb. Temperature at 800 mb is right at the freezing mark. Since this is for Dulles, and N DE is closer to the coast (and relatively warm air), it is likely that temperatures over N DE were above freezing at some level. This observation is borne out by the fact that I did hear some sleet during the night.
Figure 35 is the 700 mb potential temperature and wind fields for 12Z Feb 17 (0700). The potential temperature gradient and wind speeds have weakened over the mid-Atlantic; this is one of the reasons for the diminished snowfall during this period.
The 600 mile Doppler image in Figure 36 clearly shows the dry slot over N DE. This was not the end of the snowfall for us however; snow started falling again shortly after the time of the images in Figures 33 and 36.
E. Late Monday Morning – February 17, 2003 - 0700 (local) to 1300 (local)
We received an additional 3 inches of snow during this 6-hour period, bringing the storm total to about 24 inches in Bear, DE. The picture at left (Figure 37) was taken at 2:00 pm Monday February 17. The snow had stopped falling and there was considerable wind and drifting snow. The surface maps in Figures 38a and 38b show low pressure centers over eastern TN and off the coast of NC/SC. The coastal low is the dominant feature, and the heaviest snow has moved into New England.
The surface map (Figure 38a) has sleet and temperature of 24F at Philadelphia, while the Weather Channel map (Figure 38b) still has snow in N DE and PHL. The high pressure area has moved to east of Maine, and the snow is almost over for us in Bear, DE.
Go to top of this page
Go to next chapter