This Partners Project teamed the Indianapolis National Weather Service Office and Purdue University to determine the events that led to the formation of the Pulaski County and West Lafayette tornadoes which were not captured in realtime by the Doppler radar tornado warning criteria. Their research showed that the key to the formation of both of these tornadoes is that the inflow into an existing mesocyclone was affected by the outflow from an approaching convective system. In the case of the West Lafayette tornado, the presence of a moderately strong thunderstorm, with some color velocity signature (but below the threshold value for warning) and which was flanked to the south by a bow-echo radar feature, set the stage. The West Lafayette tornado was an F4 event that produced 3 deaths, 70 injuries, and an estimated $3.5 million in property damage along its 14-mile track, while the earlier Pulaski County was an F0 tornado.
Severe storm meteorologists have long recognized that isolated storm cells, ahead of more organized convective lines and clusters, must be monitored very carefully as potential tornado-producing storms. The approach of strong, low-level outflow winds (associated with the bow echo storm) can provide enhanced shear vorticity in a favorable region for the northern storm cell to spin up quickly to the tornado stage. The two cases fit a conceptual model that has been developed from the data set. Any mesocyclone that interacts with an approaching gust front, convective cluster, or cell should be monitored closely for tornadic development. A successful application of this conceptual model occurred on 13 May 1995 when radar data indicated that the collapsing storm had evolved into a bowing reflectivity signature pushing across a strong inflow notch feature. A tornado warning was issued, based on the recognition that the pattern fit the conceptual model, and an F1 tornado was reported a short time later. In this case, as with the ones studied in this project, a strong mesocyclone persisted for a considerable amount of time (45 minutes) before the added trigger of thunderstorm outflow appeared to trigger the tornadogenesis.