The tornado which formed in northeast Kingsbury County was considered a non-supercell tornado and was similar to a landspout. According to a farmer who observed the tornado as it formed less than a half mile away, the circulation was first observed near the surface and appeared to build up toward the cloud base. Even though he was only a couple of hundred feet from the tornado at one point, he noted that there was very little wind where he was located, while debris was being lifted in the air by the tornado. He also noticed it was clear to his south, with the southern edge of the cumulonimbus cloud located above him. This would appear to indicate that the tornado was located within the updraft of the storm where convergence may have been maximized.
What is a landspout? A landspout is basically a tornado which forms within the updraft of a thunderstorm. According to Jon Davies, they are favored in areas where there is warm and moist air which can produce low-level instability, along with a low-level boundary. The low-level instability will cause air to accelerate upward just above the surface. To determine how much low-level instability is present, we examine a quantity called convective available potential energy (CAPE) below 3 km. The higher the CAPE is below 3 km, the more quickly air will accelerate up into the atmosphere. This has the effect of stretching air parcels and increasing the spin or rotation within the updraft. But how much spin with there be? This depends on the presence of a boundary. Basically, a boundary is a front or any "line" that separates winds of different directions. These boundaries have weak rotation associated with them. However, when a strong updraft occurs on these boundaries, they can concentrate this rotation and result in the development of a tornado or landspout. While not as important, large changes in wind speed between the surface and 3000 feet can alse increase the probability of landspouts - especially if the boundary is relatively weak. These types of tornadoes can reach an intensity of F2, although most are F0.
What makes landspouts different from supercell tornadoes, such as what hit Manchester, SD, in June 2003, is that they can occur with ordinary thunderstorms. Ordinary thunderstorms are storms which develop and then dissipate within an hour. In most cases, these storms will produce lightning, brief heavy rain and, in a few cases, small hail. These storms are generally not associated with tornadoes. However, when these storms develop where there is sufficient low-level instability, 0 to 3 km CAPE greater than 75 J/kg, and there is a strong boundary present, then these landspouts can form. That is in contrast to a supercell thunderstorm in which a strong circulation will form 10,000-30,000 feet above the surface up to 2 hours before the tornado forms. The circulation with a supercell tornado is typically observed to lower toward the surface just prior to tornado formation, and the width of these circulations can be up to 5 miles. In addition, the reflectivity field may show a "hook" on the south side of the storm as rain is wrapped around the back side of the circulation. Compare that to non-supercell, landspout type of tornadoes in which the circulation builds up from the surface, the circulation is generally less than a mile wide and the developement of the tornado can happen on the order of a few minutes. No hook is observed in reflectivity and rarely does a strong circulation even appear on radar.
So how do forecasters issue tornado warnings for non-supercell tornadoes or landspouts? It is critical for meteorologists to be aware of the storm-scale environment. Based upon Davies work, forecasters look for locations with a high amount of instability and a boundary such as a dry line, cold front or warm front. Once these locations are identified, they look for storm development within these areas. Forecasters then look for rapidly developing storms which may have stronger updrafts and may even show weak rotation at low levels on radar. Finally, forecasters need to rely on spotters who can relay reports of funnels or tornadoes. Because of the rapidity with which the tornadoes can develop, lead times on warnings tend to be much shorter.
Let's look at the environment that was present over eastern Kingsbury County (highlighted yellow in all images) around 6 pm on May 2. The graphic below shows that the axis of the largest 3 km CAPE extended into eastern Kingsbury County. Notice that an axis of high instability extends from eastern eastern Kingsbury county into southwest Minnesota. The largest instability was located over eastern Kingsbury County into western Brookings County. This would be an area where strong low level updrafts could develop.
The picture above also shows mean sea level pressure countoured in green. Notice a trough of low pressure extends across central Kingsbury County. This indicates that there may be a boundary present. To see how strong the boundary is, we can display a quantity called vorticity at the surface. In general terms, vorticity is the amount of rotation in the atmosphere due to changes in wind speed or direction along a surface. On the image below, the red and white colors are areas where voriticty is high. Notice that the axis of highest vorticity is located from western Kingsbury County to near Sioux Falls, SD. So the boundary was present to focus rotation in the presence of a strong updraft.
Finally, forecasters can examine the change in wind speed and direction with height through the lower atmosphere called shear. For supercell tornadoes, it is usually necessary to have a lot of shear to increase the rotation within the thunderstorm. For non-supercell tornadoes, this may not be quite as important but the presence of high shear could help enhance the rotation in the updraft. The image below shows that the shear below 1 km was 15 to 20 kts. Generally, values around 20 kts are considered necessary for tornadoes, however, the presence of the strong boundary and high instability can help make up for the weaker shear.
To provide you context, below is the radar image from 6:19 pm CDT. You can compare the location of the thunderstorm to the location of the surface boundary and instability in the images above.