The Days of May 2004

The Days of May 2004

The month of May 2004 was one of the most active periods for severe weather and flooding over northern Illinois and northwest Indiana in recent history. Record volumes of water on the Des Plaines River in Lake County Illinois resulted in widespread flooding. Flash flooding from localized heavy rains along with the usual hail, wind and tornado events persisted through the entire month.

The National Weather Service office in Romeoville issued 396 warnings (each county warned counts as a warning) during May which include those for severe weather, flash flooding and marine (Lake Michigan). To be more specific all warnings were issued between May 7 and May 31, only a 3-1/2 week period. The previous 'record' for monthly warnings was set in July 2003 with 293.

What follows is a brief review on the month with images from selected events.

Prolonged periods of repeat weather occurrences usually reflect a static weather pattern or one which reverts back to a base state after short term interruptions. Positions of synoptic scale troughs and ridges as well as airmass characteristics are key players in determining the type and scope of weather elements repeatedly produced.

During the late spring and summer large scale flow patterns can become quite sluggish due to the weakening thermal gradients around the northern hemisphere. In May of 2004 the broad scale mid level (500mb) pattern featured a broad trough over the western plateau to near the west coast and flat ridging near the east coast.

A composite 500mb height map for the 31 days of May shows the broad WSW flow into the Chicago region. On any given day smaller scale transitory features can ripple within this flow stirring up assorted convective storms.

The active weather across the Midwest was so persistent during May that these mean charts are a reasonable view of the conditions that supported the regime. Dynamics underlying the events can be seen with the 300mb mean zonal wind field. Here the mean upper level jet extends from just south of James Bay in Canada across the northern Great Lakes. The (favorable) right rear entrance region would extend across the central Great Lakes into the upper Mississippi Valley.

At lower levels a mean 925mb height composite shows an amplified mean trough through the Plains balanced by a ridge off the southern Atlantic coast. The resultant southerly flow through the Plains is aimed at the western Great Lakes. The pattern is mirrored in the surface pressure mean chart where low pressure over the Plains supports the continuing influx of moist gulf air.

The demarcation of southern stream moist air and drier Canadian air can be seen in the mean specific humidity at 925mb. The gradient seen through northern Illinois depicts as much as anything where the zone of action was occurring.

Perhaps the most striking reflection of the whole process can be seen in the omega (vertical motion) composite mean for the month.

The maxima is located over Indiana into eastern Illinois with the Chicago area well within the center of action. Remember, this (and all the other composite charts) is a 31 day mean which includes many daily situations. The focus of these various parameters mentioned above over this area is witness to the persistence of the pattern.

Events:

The item which attracted national attention and was destructive on a wide scale in some areas was flooding. Both sustained river flooding over Lake County (IL) and portions of Cook and McHenry Counties as well as local flash flooding from individual storms occurred during this month. The river flooding, mainly on the Des Plaines River had some of its origin in southern Wisconsin but affected the river channel into northern Cook County. Monthly rainfall totals peaked at over 10 inches across McHenry and Lake County while 6 inches or more were common further south including northwest Indiana. Some flows and crests on the DesPlaines River set records during this time.

Individual weather events occurred in a variety of meteorological situations under the same broad synoptic regime. A few of the notable aspects of specific events are shown below.

May 18

Wind and some hail occurred during the predawn hours but the item of interest on this day was a splitting supercell over eastern Iroquois County which produced 1.5" hail over southern Newton County near Kentland. A sounding from ILX at 1800UTC showed a directionally uniform wind field through at least 15000 ft along with very little speed shear. The wet-bulb zero was near 10000 ft which coincided with a very weak thermal cap.

At 1813UTC a small but intense thunderstorm was located just west of the Iroquois-Benton County line. By 1848UTC this storm had split into two distinct cells, one moving with a right bias and one moving with a left bias (see image). In many split cases the left mover fades while the right mover becomes the dominating and main weather producing cell. In this particular instance the left moving cell intensified as it moved over southern Newton County near Kentland. The 1903UTC reflectivity image shows the strong hail producing cell. The old 'right mover' is near Fowler in Benton County.

One of the potential characteristics of a left moving supercell is that it can rotate anticyclonically (right movers are prone to rotate, if they do, cyclonically). The 1853UTC SRM (Storm Relative Motion) image clearly shows the rotational couplet is clockwise or anticyclonic. At this point the beam elevation is around 5500ft agl. The rotation attests to the good organization of the storm and its ability to hold particles aloft long enough to produce the large hail observed.

May 20

The DVN sounding at 0000 UTC of the 21^st provides a few clues as to the downstream potential over northeast Illinois. Light uniformly directional winds below 10000 ft increased to 40 kts at 15000 ft. Quite moist air below 750mb gave way to a drier zone above where the winds were increasing. Buoyancy in the form of CAPE was high at nearly 4500j/kg with minimal thermal capping in place. Vigorous convection could be expected in this environment given an initiation mechanism.

A line of strong thunderstorms had developed by 0120 UTC from central DeKalb County to northern Cook County. The deep moist layer (and precipitable water of 1.69") shown on the DVN sounding suggests heavy rain would be occurring along this band. Cells were also 'training' from west to east adding to the threat. A one hour rain total of 2.75 inches was reported just north of DeKalb.

Further east other processes were at work. The dry air noted on the DVN profile as well as the increased winds through that layer may have been instrumental in changing the character of the convection. Aside from the heavy rain associated with this line, strong surface winds were being reported in some spots. O'Hare International Airport (ORD) reported a wind gust to 50 kt (58 mph) about this time. The Doppler velocity image at 0120UTC shows the distinct signal of a downburst type wind over ORD.

May 23

At the end of a string of busy days another small tornado occurred during the evening of the 23^rd near the Mclean-Livingston County line. A line of thunderstorms extended from just northeast of KILX into Livingston County at 0111UTC. They had formed along a weak front separating very moist air to the east with somewhat drier air to the west. Cells 1 and 2 are noted in the image. Within 15 minutes these two cells had merged near the Mclean-Livingston County line. A tornado (F2) was spun up 1.7mi NW of Chenoa and dissipated 2.5 mi SW of Ocoya. The system rapidly decayed after this time as it drifted to the northeast. The accompanying SRM display at 0124UTC shows a distinct but not overly impressive rotational couplet at 5000 ft agl.

May 30

A vigorous synoptic scale trough for late May moved through the region late on this day. Morning convection with flooding was followed late in the day by a convective line which produced winds, some flash flooding and a band of small tornadoes.

The tornadoes (F0) occurred in Livingston County between Pontiac and Collum between 6PM and 7PM. They were associated with a bowing segment of the convective line as it accelerated eastward across Livingston County. Reflectivity imagery at 2336UTC of May 30 shows the bow. The notch impinging into the convection east of Pontiac reflects the subsidence and drying in the wake of the fast moving storm.

The coincident SRM image from the same time suggests a zone of strong shear near the front edge of the reflectivity. No classic tornadic couplet is evident. It is likely that strong shears along the forward edge of the storm overlaid by strong updrafts providing vertical stretching of the column were significant contributing factors to the 'string' of weak tornadoes observed in this case.

KML 7/04