On 29–30 May 2012, the Deep Convective Clouds and Chemistry experiment observed a supercell thunderstorm on the southern end of a broken line of severe storms in Oklahoma. This study focuses on an approximately 70 min period during which three mobile Doppler radars operated and a balloon‐borne electric field meter, radiosonde, and particle imager flew through the storm. An overview of the relationships among flash rates, very high frequency (VHF) source densities, and Doppler‐radar‐derived storm parameters is presented. Furthermore, the evolution of the flash distribution relative to the midlevel storm's kinematics and microphysics is examined at two times during a period of rapid storm intensification. The timing of increases in VHF counts in the 8–10 km above ground level (agl) layer, which contained the largest VHF source counts, is similar to the timing of increases in updraft mass flux, in updraft volume, and in graupel volume at approximately 5–9 km agl. Although some increases in VHF source counts had little or no corresponding increase in one or more of the other storm parameters, at least one other parameter had an increase near the time of every VHF increase, a pattern which suggests a common dependence on updraft pulses, as expected from the noninductive graupel‐ice electrification mechanism. A classic bounded weak lightning region was observed initially during storm intensification, but late in the period it appeared to be due to a wake in the flow around the updraft, rather than due to a precipitation cascade around the updraft core as is usually observed.

Previous studies have found that the vertical distribution of sources of very high frequency (VHF) signals from discharges mapped by Lightning Mapping Arrays typically have a secondary maximum in a storm's overshooting top. Low rates of these sources tend to occur continually throughout the lifetime of the overshooting top (OT), rather than sources occurring in the episodic distinct flashes observed at lower altitudes. This study examines the evolution of the VHF OT signature (VHF OT, defined here as a sustained period of ≥4 VHF sources per minute per 200 m layer above a storm's level of neutral buoyancy (LNB)) relative to the evolution of radar reflectivity and IR imagery of overshooting tops in three supercell and two multicell storms. The VHF OT began before OTs were detected in IR satellite imagery of the supercell storms. No OT was observed in IR imagery for either multicell storm, but this lack may have been due to the spatial and temporal resolution of the current IR imager. The VHF OT began near the time the 18 or 30 dBZ echo top rose above the LNB and ended near the time it fell below the LNB. There were too few radar volume scans of OTs in the multicell storms for a correlation analysis. In the supercell storms, however, the maximum altitude of VHF sources typically was less than or equal to the altitude of 18 dBZ echo tops and was correlated with these echo tops (linear correlation coefficient ≥0.86) during the period of the VHF OT.