Examination of a Training Cold Season Heavy Rain Event Over the Ohio River Valley

During the period 3-4 January 2000 heavy convective precipitation exceeding five inches fell across northern Kentucky, southern Indiana and southern Ohio along the Ohio River Valley. This event basically fit the "synoptic" type heavy rain scenario described by Maddox et al. (1979) as it was associated with a slow-moving synoptic scale cold front which was quasi-stationary for a period during the event. Heavy rain and severe thunderstorms developed along and south of the Ohio River Valley as a wave developed on the front over eastern Oklahoma and moved northeastward toward Illinois. Elevated thunderstorms were observed north of the quasi-stationary boundary, also contributing to heavy precipitation in southern Indiana and Ohio.

This presentation will focus upon the synoptic and mesoscale forcing of this event which was atypical for the winter months in this part of the United States. Unusually warm, moist air (temperatures and dew points > 60° F ) moved northward into Tennessee and Kentucky during the period, accompanied by an elevated region of instability diagnosed using max q_e CAPEs (CAPE computed by lifting the parcel in the lowest 300 mb with the highest q_e value). Precipitable water (PW) values of over one inch, which is greater than 200% of normal, were advected into Tennessee and Kentucky, setting the stage for a significant precipitation event. Surface moisture convergence fields revealed a series of moisture convergence maxima which formed upstream from the region of surface-based thunderstorms and then translated downstream. Interestingly, the elevated thunderstorms located further north were not directly correlated to the surface moisture convergence fields but were more associated with strong moisture transport diagnosed on low-level isentropic surfaces. WSR-88D radar imagery will be shown to illustrate how the training cells evolved with respect to the surface moisture convergence fields. Storm motion vectors computed from WSR-88D data will be compared to those computed using the method devised by Corfidi et al.(1996). In this method the storm motion vector is approximated as the sum of the propagation vector, estimated using a vector opposite to the low-level jet (LLJ), and the cell motion vector, estimated by the average wind in the 850-300 mb layer.

Upper-level forcing will also be examined through an analysis of the LLJ and upper-level jet (ULJ) to note how these features evolved and interacted during the course of this event. The LLJ, which was southwesterly at greater than 25 m s^-1, played a pivotal role in maintaining the inflow of moisture and unstable air into the region over a 36 hour period. The presence of a long-wave trough situated over the central Midwest was also critical since it was slow moving and associated with an elongated ULJ axis on the eastern side of the trough, anchored well to the west-northwest of the region of heavy rainfall. A Q-vector analysis will also be shown to document the synoptic scale support of upward vertical motion throughout the period.