An example of convective initiation due to front/wave interaction
Philip N. Schumacher
National Weather Service
Sioux Falls, South Dakota
I. Introduction
On 30 April 2001, a line of severe thunderstorms with embedded supercells developed across northwest Iowa. Between 5 PM CDT and 830 PM CDT, over 30 reports of severe weather were reported including 3 tornadoes and tennis ball sized hail (Fig. 1). These storms developed on a day when severe weather was not considered likely. A weak front had just moved across southeast South Dakota and was stationary across southwest Minnesota and northwest Iowa late that morning. Little precipitation was associated with the front and dew points along the front were only in the middle 50s. During the afternoon, a weak upper-level wave moved across South Dakota initiating mid-level convection as the convection approached Sioux Falls and Brookings. As the wave approached the front, convergence increased along the front and convection rapidly developed across northwest Iowa.
II. Case review
By early afternoon, on 30 April, the surface front had become stationary and extended from Worthington Minnesota to Sioux City Iowa (Fig. 2). There was a modest temperature gradient across the front - temperatures across southeast South Dakota were generally near 70 degrees while east of the front temperatures were in the middle 70s. While there is some low level convergence evident with the front, the general wind field was 10 knots or less with a wind shift of approximately 90 degrees. While there was some moisture convergence along this front, it remained rather weak and broad. While a broad band of clouds and light precipitation was located to the east of the trough, no cumulus were developing as a result of the weak convergence. However, the visible satellite imagery also shows an area of mid-level convection across central South Dakota. This convection is associated with an upper level wave moving across western South Dakota (Fig. 3). Both the water vapor imagery and the potential vorticity show strong evidence for this wave. A darkening, indicative of subsidence right behind an upper level can be seen to the west of the maximum in 300-400 mb potential vorticity. Also notice the bright white in the water vapor imagery is coincident with the gradient in the potential vorticity. This is where the mid-level convection had developed and indicates there was significant lift associated with this wave. No 950 mb moisture convergence is associated with this convection since it originates above the surface.
By 3 PM, the surface trough had moved northwest of Sioux City and was located near Le Mars and Orange City Iowa (Fig. 4). Although a line of cumulus had developed along the trough, the convergence along the trough remains relatively weak with light northerly flow to the north of the trough and southwesterly flow to the southeast. The air mass contrast remained rather minimal along the trough with middle to upper 70s southeast of the trough and upper 60s to middle 70s to the northwest. On either side of the trough, dew points were in the middle 50s. The combined temperature and dew point resulted in a convective available potential energy (CAPE) of around 1000 J/kg. The visible satellite picturealso shows the development of convection extending from near Brookings to Sioux Falls. The radar imagery from 2032 UTC (332 PM CDT) (Fig. 5) shows a this line of convection. The strongest cell was located southwest of Brookings and eventually produce some pea-sized hail across northern Moody county. Also notice that light showers have begun to develop along the trough northwest of Sioux City Iowa.
The water vapor imagery at 3 PM shows that the wave has moved into central South Dakota (Fig. 6). This puts the best PV advection across eastern South Dakota. With the approach of the upper level wave the width of the 950 mb moisture convergence has contracted. This is likely the result of the approaching upper level wave. Research studies have shown that the approach of an upper level wave will cause an adjustment to the low level winds that result in increased low level convergence as the atmosphere tries to restore a balance between the lower and upper levels. This balance is usually restored through increased vertical motion. Rising motion will cool the atmosphere and sinking motion (subsidence) will warm the atmosphere. This response at low levels is enhance near frontal boundaries because the atmospheric stability is much lower. This is because with lower stability it takes more lift to achieve the same amount of cooler has when there is high stability. Therefore, near frontal boundaries, the response to the approach of an upper level wave will be a significant increase in upward vertical motion. To achieve this, the low level winds will become more convergence and the frontal zone will "collapse" or have a smaller width indicating stronger boundary. There is already evidence that strong lift is associated with this upper level wave as weak convection has broken out across eastern South Dakota where there is more stability and less moisture. Therefore, one would expect that once this wave interacts with the low level boundary, a much more vigorous response and stronger convection would ensue.
At 4 PM, there is evidence of this occurring (Fig. 7). The surface map and visible satellite picture at this time shows that cumulonimbus have begun to from across extreme northeast Nebraska and northwest Iowa - coincident with the location of the surface boundary. In addition, the surface winds at Le Mars Iowa and Sioux City Iowa have backed to the southeast and increased to 10 to 15 kts. This has markedly increased the convergence along the front and led to an intensification of the storms. The radar at this time shows that thunderstorms continue to develop across northwest Iowa (Fig. 8). The storms near Sioux Falls were beginning to weaken. However, farther north, the storms moving south of Brookings continued to produce heavy rain and small hail. The water vapor imagery shows that the upper wave has moved east of the Missouri River with the best gradient located over the low-level trough (Fig. 9). The best PV gradient has moved over the low level moisture convergence maximum. This would imply that the interaction between the low level front and upper level wave should increase resulting in stronger vertical motion and increased convection. As stated above, there is evidence of this in the wind field and also by the fact more showers are developing on radar. By 430 PM (2130 UTC), the radar shows a line of storms along the trough in northwestern Iowa (Fig. 10). Other storms were developing near Vermillion South Dakota. The circulation associated with the surface trough has become the primary focus for convective development. With relatively cold air aloft, CAPEs between 1000 and 1500 J/kg, and moderate shear (0-6 km shear around 40 kts), organized convection was able to develop.
By 5 PM CDT, the convergence had contracted even more as the upper wave continued to interact with the front (Fig. 11). Surface winds across northwest Iowa had switched to south or southeast with northwest across southeast South Dakota. Because of the release of the instability, vertical motion across the front increased with a rapid expansion of the cumulonimbi clouds across northwest Iowa. The radar shows strong thunderstorms across the area with several storms likely producing small hail by this time (Fig. 12). Within 15 minutes, the first severe cell would develop with golf ball sized hail being reported 15 miles east of Sheldon IA. The storms would remain severe across northwest Iowa for the next 3.5 hours. Two supercells were also observed - the first over southwest Minnesota which progressed from Chandler Minnesota to southeast of Heron Lake Minnesota. Up to 2 inch hail was reported with this cell as a half mile wide swath of hail fell. Damage to homes and cars was reported in Fulda MN. The second supercell developed southwest of Sioux City and moved to near Sloan Iowa where a tornado was observed on the ground for 5 to 10 minutes by law enforcement. In addition, to the tornado tennis ball sized hail was reported near Salix Iowa - 15 miles south of Sioux City.
III. Conclusion
This case is an illustration of the interaction of a weak surface boundary with a strong upper level wave. What was observed was the transformation of a broad ill-defined surface trough into an intense near frontal boundary due to the approach of the mid-level wave. Even though dew points were considered marginal for severe weather, the combination of a strong frontal circulation with moderate instability was enough to produce numerous reports of large hail and even 3 confirmed tornadoes. While the computers models did predict this wave to move across South Dakota, they were generally slower with the wave (having it reach the James River Valley around 0000 UTC (7 PM) instead of 2100 UTC (4 PM) and the trough was expected to move south of the CWA. Even though the model errors were small, observing these errors and anticipating their effect on the convective development is necessary in order to anticipate severe convection during the summer.
I have set up a two loops - the first is a loop of the visible satellite imagery from 1 PM to 5 PM 30 April 2001. The second loop is the water vapor imagery with potential vorticity and moisture convergence from 1 PM to 5 PM 30 April. Each loop is approximately 1 MB and may take several minutes to load if you have a slow connection. If you have comments about this case, you can contact me via e-mail.
