3D FLASH Yosino Full 15

Two cells observed at the CG site (Fig. 4) produced more than 10 flashes. As shown in Fig. 5a, the flash at the CG site was detected at about 2345 UTC at a range of about 100 km, and produced 12 flashes within a period of 30 min. The most active cell (cell 3, Fig. 5b) produced 17 flashes during the first 10 min of its lifetime and then decreased to about two flashes per minute. One other cell produced about six flashes over a 30-min period. The most active cell on 11 November (Fig. 5c) was located in central Texas, producing about six flashes in the first 15 min of its lifetime and then decreased to about one flash per minute. This cell was located at the interface between a strong low-level circulation system and an adjacent weak, but strong, upper level flow. The thunderstorm spawned numerous wave-like features extending from the radar height, with a strong updraft at the base of the growing thunderstorm. The upper level flow also enabled upward transport of the storm, resulting in a cloud base of roughly ~10 km MSL (Fig. 5c).

Four cells that produced flashes at TX sites occurred during periods of increased upper level outflow. As shown in Fig. 5d, the most active cell (cell 1) was located at the SW edge of the radar, produced about 10 flashes over the first 10 min of its lifetime, and then decreased to about one flash per minute. The most active cell on 28 January (cell 2) produced two flashes in 15 min. The cell on 9 April (cell 4) was located near the radar center and produced 10 flashes over 10 min. The most active cell on 28 January (cell 5) produced 13 flashes over the first 10 min of its lifetime. A separate cell (cell 6) located near the radar center produced about two flashes per minute over the first 5 min of its lifetime. Figure 5 shows that the most active cells all had small, vertical extent, and were associated with storms with upwardly directed upper level outflow. These features correspond to the base of the storm (i.e., the cloud tops), which shows high vertical cloud extent, indicating low wind speeds, consistent with the weak updrafts near the thunderstorm’s base. As illustrated in Fig. 5c, the cell produced numerous wave-like features extending from the radar height, with a strong updraft at the base of the growing thunderstorm.

Here, total flash rates are computed from the RHI lightning strike and flash data. Over the period of study (May 2012 to Oct 2013), there were a total of 44 thundersnow events reported in the US (Meteorological Society of Canada 2013). Of those, 43 had data available for time of occurrence, occurrence height, and flash rates. Among the remaining event, no lightning activity was reported in seven events, and in three cases the record was incomplete. Two events, which were also reported in the previous studies by Duncan (2014) and Leighton et al. (2015), are disregarded due to the large uncertainties in flash rates. The remaining data set consists of 10 CN events and 9 USG events. For both the CN and USG cases, flash rates are computed separately for IC and CG parts of a thunderstorm, and then averaged over all events. Because we know only the occurrence times and occurrence heights of thunderstorms, only the flash occurrence height can be determined. The height of a flash can be estimated either from the main polarity or the polarity reversal. In this study, we compute flash rates by averaging over the height bins, assuming a constant flash occurrence fraction between the bins. Most of the flash events occur at similar zenith angles around the radar (~40°), which facilitates their further averaging. 827ec27edc