DPWX/Supercell thunderstorm near Brush, Colorado: 29 July 2018


Author: Patrick C. Kennedy

Several strong to severe thunderstorms over northeastern Colorado appear in the CSU-CHILL S-band reflectivity data collected in a 2.5 degree elevation PPI sector scan on 29 July 2018. The largest storm, located ~80 km east of CHILL, produced significant wind and hail damage near the town of Brush. Selected radar data and storm damage images have been assembled from the Brush area segment of the storm's track.


Multiple severe thunderstorms moved across northeastern Colorado during the afternoon and evening hours of Sunday, 29 July 2018. The CSU-CHILL radar collected primarily low elevation angle PPI scan data during much of this outbreak. Due to a failure in the horizontal polarization transmitter, S-band data was collected using only the vertical transmit polarization state. The CHILL's X-band channel (maximum range of 90 km) was also operated; it collected full, simultaneous operating mode dual polarization data.

The data presented here came from the 2239 - 2252 UTC time period when a damaging supercell thunderstorm affected the area around Brush, Colorado. According to the Storm Prediction Center's Storm Reports records, tornado-related damage was reported within a few miles of Brush between 2248 and 2253 UTC. Personnel from the Denver / Boulder NWS Forecast office identified three separate tornado damage paths in their post-event surveys NWS survey

0.7 degree elevation angle reflectivity and radial velocity loops

During the image loop time period (2238:51 - 2251:45 UTC), the CSU-CHILL radar was conducting continuous, PPI sector scans between a left azimuth limit of ~015 degrees and a right limit of ~130 degrees. To minimize the scan repetition time, only two elevation angles (0.7 and 1.0 degrees) were used. This scan regime updated the PPI sweeps at a time interval of just over one minute (1:01 to 1:05). The CSU-CHILL image frames in the following loop were interactively edited using NCAR's SOLO software. The primary editing activities involved the removal of range aliased echoes and the local unfolding of regions where radial velocity aliasing occurred. In each image frame, reflectivity (Zv in dBZ) is shown in the top panel and radial velocity (mps) is on the bottom. Brush Municipal Airport is located at ~102 degrees azimuth and 92 km range from CSU-CHILL.

During the early portion of the loop, the reflectivity data shows the inflow notch and hook echo configuration that is typical of the lower height levels in a supercell thunderstorm. With time, the inflow notch becomes less well defined as the hook appendage expands and surges towards the south-southeast. The outbound (positive) radial velocities in the southern echo appendage increase to speeds greater than 30 mps during final loop frames. Localized patches of intense convergence and cyclonic rotation are present in some of the image frames. The cross-beam pulse dimension is ~1.5 km at the storm's range, significantly reducing the capability to resolve tornado scale motions. A larger mesoscale cyclonic rotation pattern broadly centered on the hook echo can be seen.

Click play to begin animation

Delay: ms


Dual Doppler horizontal wind field at 2242 UTC

By coincidence, the CHILL scans fell into fairly good time synchronization with the operational volume scans that were being done by the NWS Denver / Front Range WSR-88D radar (KFTG) near 2242 UTC. (A CHILL volume started at 2241:56 UTC and a KFTG volume began 11 seconds later at 2242:07 UTC). The edited polar coordinate data from these two volume scans were used as inputs the the NCAR CEDRIC software to synthesize the horizontal wind field. Due to the CHILL's restricted, two elevation angle scanning, the wind field synthesis could only be done at a limited number of Cartesian grid points. The most complete wind field recovery occurred at an analysis height of 3.4 km MSL (~2.1 km above ground level near Brush).

The following dual-Doppler analysis at 3.4 km MSL shows the vector representation of the synthesized horizontal wind field plotted on top of the KFTG reflectivity field. The mesocyclone circulation is well defined; at this time and height, the highest wind speeds (~43 mps) were occurring in the north-northwesterly flow along the axis of the hook appendage. The hook appendage / strong wind flow was affecting the two color-coded points were storm damage was confirmed. These locations were the Brush Airport and the home of a Community Collaborative Rain, Hail and Snow Network (CoCoRaHS) observer. Additional data from this CoCoRaHS observer is provided in the next section.

CoCoRaHS observer data

The CoCoRaHS observer located a few km to the northeast of the Brush Airport described the 29 July 2018 storm as being "incredible". Maximum hailstone diameters were 1.75 inches (golfball sized), with the averages size being only slightly smaller (1.25 inches). The observer provided the following pictures showing a broken section of the rain gauge funnel due to a hailstone impact. The combined effects of the high winds and large hail also dislodged the CoCoRaHS rain gauge from its mounting post.

Additional documentation of the destruction produced by the wind-driven hail are shown in the following two pictures that were also provided by the CoCoRaHS observer.


As noted earlier, the CHILL S-band data was collected in vertical-only transmit polarization mode. (The >90km range to the storm was beyond the X-band channel's observation limit). In V-only mode, Linear Depolarization ratio (LDR) is the only available dual polarization measurement. In this transmit mode, the cross-pol signal component appears in the H polarization receiver channel. LDR is calculated by comparing the strength of this cross-polar return to the level of the co-polar (V in this case) return from the same sample volume. For meteorological targets, the cross-pol return is typically only a small fraction (~.001 - .01) as large as the co-polar return. When expressed on a dB scale, this magnitude relationship yields LDR values on the order of -30 to -20 dB. The final plot shows the -18 and -15 dB LDR contours as white lines plotted on top of the CHILL 0.7 degree elevation angle PPI scan data at 2243 UTC. These contour lines identify the enhanced LDR levels that affected the immediate vicinity of the CoCoRaHS observer's location. The wobbling motions and non-spherical shapes that are typical of large hailstones are physical characteristics that produce increased LDR values (Herzegh and Jameson, BAMS 1992, p. 1365-1374). (If melting was able to generate a liquid water coating on the surface of the hailstones as they descended into warmer air temperatures, additional LDR enhancement would be likely.) Finally, the presence of irregularly-shaped surface debris lifted by high winds could also contribute to the locally elevated LDR levels found near the CoCoRaHS observer's location.


Noah Newman of the CoCoRaHS main office coordinated the acquisition of the storm damage photographs from the Brush area observer. Chad Gimmestad and Robert Kleyla of the Denver / Boulder NWS office provided additional details on the results of the post-storm damage survey.