We opted not to do any coordinated scans between the radars. Instead we did scan setups separately so as to gain more experience and understand better the range of permissible operating parameters at each research radar. The main issues are being sure that the radar processors continue functioning when the number of samples is changed, and that the antennas can indeed scan at rates as high as 12 deg/sec (96 samples) without significant degradation of the polarimetric measurements. CHILL will use its so-called quick-scan mode at scan rates exceeding 8 deg/sec.

The SDSMT T-28 crew and airplane arrived this afternoon, and some equipment installation was completed at the Ops Center.

The NMIMT LMA now has 6 operating stations.

There were some communications problems that arose and were worked through at the morning meeting. Better definition of standing by was decided so that the mobile units would be prepared. Basically, we would call a time for standing by which means that at that time the mobiles should be ready to be dispatched. However, updates will be done before this time in order to address the possibility of dispathing the mobiles some before the actual standby time.
[back to calendar]

The severest weather was farther south with 1-1/2 inch hail south of Limon and late evening flooding and high winds (60 mph) in west central Kansas.

Addendum to May 25:

The STEPS domain is defined roughly as the region within 150 km of both research radars. This area is also covered reasonably well with the LMA centered near the center of the triple-Doppler (CHILL-SPOL-KGLD) network.

The first electrical balloon sounding will normally be done into the anvil, taking about 30-45 min to complete. Subsequent soundings will be in the inflow, taking about 15 min to complete. Ideally, the inflow sounding would correspond with a T-28 penetration; however, this would be very difficult so that if the T-28 strives to make a penetration within +/- 15 min of the launch time of the balloon this would be satisfactory.
[back to calendar]

It passed south of CHILL in a less than ideal geometric position. The T-28 made one successful pass through one of the reflectivity cores before it had an equipment failure, likely as a result of being struck by lightning. The mobiles and ballooning were dispatched eastward toward Colby KS to get ahead of the storm. By around 1645 MDT several outflow boundaries had collided and triggered new convection throughout most of the STEPS domain. Since there were no obvious regions to focus on, we went into a WSR88D-like scan mode of full 360 deg scans by both research radars. Excellent polarimetric data was taken where CHILL did the hybrid (simultaneous H and V transmissions) mode alternately with the normal H then V transmission.

These 360 deg scans continued until 18:30 MDT when we secured operations because most of the convective cells were well to the east of the STEPS domain.
[back to calendar]

We turned our attention to developing convection to the NE of S-pol from 0130 to 0245 UTC, 1 June. This complex eventually produced a storm near McCook that had characteristics of an LP supercell (high Z, low Zdr, and small Ldr), indicative of dry growth hail. This storm, although at the fringes of the radar coverage, nonetheless represents a great case (radar and LMA only). The CG lightning switched from negative to nearly all positive (CG's) around 0230 UTC. After this event, S-pol and CHILL did coordinated scanning on another strong storm to the south (0245 UTC to 0330 UTC). Again the entrainment of mid level air from the front side was obvious in the CHILL Doppler data. This entrainment forced an early demise to this event, and again drove a strong surface microburst.

Debrief comments/discussion:

It was apparent that control of the Hailstone intercept van should be placed under direction of the Ops Director and CHILL Radar Scientist for maximum efficiency for intercepting hail and intense rain cores. This will be done from now on.

It was also clear that we need to think of distributed observations at times, say using one research radar and the mobile mesonet to work developing convection (or even a boundary, or collision of boundaries, etc), while the other research radar focusses on the storm at hand, say being worked by the T-28 and EFM platforms. The Goodland WSR-88D can provide dual-Doppler with either CHILL or S-pol when the other research radar is operating in a single Doppler mode (provided the right geometry is available). This should be kept in mind for future operations. The mesonet is a pre-storm platform (although not exclusively) while the other platforms focus on mature convection.
[back to calendar]

Dual-Doppler scanning on dry line feature was concentrated between 1820 to 2245 UTC. Then sector scans were set up after 2045 UTC to observe developing convection. 2310 UTC-decision made to target developing echo with T-28 and EFM/mobile mesonet. These facilities were well positioned and obtained quality data. T-28 made 4 penetrations between 0000 UTC (4 June) to 0100. Data suggested numerous penetrations of developing clouds. Main precip core was penetrated and particles were limited to graupel no larger than 5 mm (via telemetered data). NSSL5 had two ascents, one into core and one into rear flank. This will be a very interesting storm to analyze from the point of view of storm electrification. Why were there no CG's (as depicted by NLDN data)? How large was the IC flash rate? This should be a good case for relating storm kinematics, microphysics and electrification in a low CG producer. Precipitation shaft was very narrow, and consistent with dry mid level air this storm had to work with. Storm was in range of 3-D LMA.

LMA data collection extended from 2257 UTC 3 June to 0110 UTC 4 June. First lightning at roughly 2257 UTC. Flash rates were 1 flash/10 sec at 2305 UTC, peaking at 1 flash/5 sec at 2358 UTC. Based on qualitative inspection of the LMA record by NMIMT, this storm was deemed to be a "significant" lightning producer.

Addendum: There were 6 passes by the T-28 at 2 different altitudes, roughly 6 and 6.6 km MSL. Temperatures were about -8 and -14 C. Although we did not know it in real time, precipitation particle sizes ranged up to 1 cm. There was graupel near new and old updraft cores, and rimed aggregates and smaller rimed single crystals in other portions. There were not many pristine small ice crystals in the regions sampled. The electric field structure was reasonably simple, and there was evidence of lightning at a frequency of about one or more flashes per minute. Peak updrafts were about 10 m/s.
[back to calendar]

At 7:35 pm, radar emphasis shifted to another cell near Wray, Colorado, which was norht of and much stronger than the earlier targeted convection and which was exhibiting nearly 100 per cent positive CG lightning. This cell was propagating SE towards the eastern dual Doppler lobe, but decayed about 9:00 pm, before entering that lobe. SPol took sector scans of this cell from about 7:35 till its demise. No mobile mesonet observations were obtained.
[back to calendar]

Addendum: As part of the morning debriefing, STEPS investigators summarized observations to date, noting especially that several of the observed storms, albeit weak and high based, have exhibited very interesting electric field and lightning characteristics, with reverse polarity (negative within the anvil and positive below) in at least two of the cases. This observation matches some recent observations from the Oklahoma region, but is counter to the established conceptual model of electric field distribution in most convective storms.

Observational highlights of the event included:

T-28 made multiple penetrations in developing and mature convection from roughly 2145 to 2245 UTC. Observed liquid water contents peaked at 1.5 g/m3, at -8 C. Heavy icing was picked up by the aircraft. Copious large particles were detected on most penetrations. T-28 studied two convective elements in the line while coordinated dual-Doppler scanning was conducted in the western lobe.

NSSL5 launched a total of five EFM's, with good data obtained on four of these flights. Balloons were launched in the vicinity of the T-28 passes. EFM's were obtained in the leading anvil, core (two flights) and post core environments. This should provide a very nice cross section of the E-field structure. Interferometer reported the horizontally propagating flashes near the time of the balloon ascents.

Mobile mesonet appeared to obtain a great dataset. Originally they sampled the well defined gust front of this system along an east-west line along route 36, just west of Idalia. Then they shifted their east most probes to a position east of Idalia, and north-south along 385 to study the gust front and convective passage in more detail. Numerous gustnadoes were observed by the probes.

Hailstone van got a core penetration near Joes CO where we received a report of 3/4 hail at 2143 UTC from a NWS spotter.

Dual-Doppler coverage was continued in east lobe after NSSL, mesonet and T-28 left the event. Doppler observed very strong rear inflow jet, as well as a well defined gust front. Peak outflows were 50 knots. This MCS was one of three MCS strung out along a N-S line. The southern most MCS was in the OK panhandle, and was the most intense of the three. Our MCS was the second in this N-S line, while the third MCS was situated in western Nebraska. This latter system was more of a parallel-stratiform system. MCSs #1 and #2 were more of the asymmetric MCS type with the stratiform region to the north of the bowed convective line.

RHI and PPI coverage was continued from 0030 (12 June) to approximately 0130 UTC (12 June) to document large stratiform region that was producing large horizontally propagating flashes at a rate of 1 per 5 minutes. Possibly a sprite producing case.

MGLASS soundings were obtained at Limon and Goodland at 18 and 21 UTC. Also, a post convective sounding was taken at Flagler at approximately 2245 UTC. A pre-storm sounding was also obtained at Goodland at approximately this same time. These soundings should have provided excellent documentation of this system.
[back to calendar]

The southern end of the bowed boundary produced the storm that ended up being our target for the T-28, NSSL5 and the intercept team. All units were all well positioned but the storm died extremely rapidly, as it was undercut by a major outflow boundary. This latter feature was oriented east-west and was generated by convection in extreme SE Colorado. The cell that did form SE of CHILL near 2300 UTC produced some IC lightning, but very few, if any CG's. This storm was reported having an inverted charge structure, positive underlying negative. Polarimetric radar indicated a brief period of large drops with this precipitation shaft. The T-28 made two passes in the rapidly decaying convection. One EFM was launched into the decaying storm but the quality of the data on this flight was unclear at the time of this writing. The mobile mesonet worked the dryline prior to convective development. All facilities performed well in this case.
[back to calendar]

The T28 completed about 6 passes through the storm between 6:40 pm and 7:40 pm as it propagated from the CHILL site to northeast of Goodland. One electric balloon was launched from the Goodland WSFO into the leading anvil of the storm at about 7:10 pm, and an additional balloon was launched behind the storm at 7:50 pm. Launches into the core of the storm were curtailed due to 70-75 MPH surface winds. The mobile mesonets began transects of the storm starting about 4 miles west of Goodland. Other than the strong winds, the mobile mesonets did not document any hail. Other reports suggested occasional observations of pea-size hail. The storm system exhibited a large amount of IC lightning, but only a modest number of CG strokes. The CG strokes were predominantly of negative polarity.

MGLASS soundings were taken at 18, 21 and 00 UTC near Straton CO and near Colby KS. An additional sounding was launched from Colby at 1:30-2:00 UTC as the 00 UTC sounding was partially lost due to power problems.

Radar observations ended at 8:20 pm, as the storm system propagated east of the eastern lobe.

Additional severe storms were triggerred as this system propagated across the dry line near Colby KS, with hail up to quarter size and two brief tornadoes reported in the vicinity of Hoxie KS. STEPS observations were not taken on these storms.
[back to calendar]

The T28 was called out at 5pm along with the storm intercept and the mobile ballooning units, which were vectored to near Idalia for soundings and ground verification. The T28 did transects of the storm system between 5:50 and 6:50 pm, experiencing significant turbulence, icing, hail, and frequent lightning. An initial tandem electric balloon was launched into the updraft, but failed as the instrument package separated from the balloon shortly after launch. Two successful balloon launches were obtained on the western side of the storm system. The lightning mapping network showed a large positive charge layer within the storm at heights between 6 and 9 km. Radars continued coordinated scans of the system through about 9pm, as it moved northeast of the Goodland area, within the eastern dual-Doppler lobe.
[back to calendar]

Coordinated observations were also taken with two DOW radars, which were located about 10 m N of Joes CO and 10 m S of Cope CO.
[back to calendar]

One particular highlight was the fact that the storm became a right mover near Wheeler around 2330 UTC. This behavior was accompanied by the onset of positive CG's and the occurrence of a tornado. The first tornado observation was noted around 2330 UTC, az/range of 57 degrees, 75 km from CHILL (reported by NSSL5). Prior to this time the storm produced only one positive CG, and was primarily an intracloud lightning producer. The storm was reported as having an inverted polarity structure prior to the turning and onset of positive CG's.

Near 2350 UTC, the storm featured a well-identified BWER with tops to 15 km. Ldr values as large as -15 db in the hail shaft. Reflectivites of 60 dBZ were maintained to 10 km AGL for a good deal of the time the storm was in the east lobe.

First EFM was launched near Wheeler into the mesocyclone. NSSL5 was then directed south and east and ended up setting up for further launches near Brewster KS. Second EFM launch was 3 mi. west of Brewster, again into mesocyclone. Third EFM was also into mesocyclone near 0045 UTC at Edson, KS. The fourth EFM was sent up into a strong core that was just north of I-70, with was north of the main core that the previous EFM's had been launched into. This storm was targeted since it was producing positive CG's and had a mesocyclone. If NNSL5 would have stuck with the southern storm that they had been working, they would have been directly in the path of the mesocyclone associated with the northern storm. This location was undesirable for obvious reasons, and hence the northern storm was selected. The Hailstone van set up near Brewster too, and got a great core penetration near 0030 UTC 30 June. There were numerous reports of hail exceeding golf ball size in this area. The manual hail catcher was used, but the software on the SDSMT hail box was not operative after several reboots.

T-28 penetrations were conducted between roughly 2240 and 2340 UTC, 29 June. A total of 6 passes were made. Pass 4 and 6 were most interesting, with a peak updraft of 33 m/s on one of these passes. Heavy rain, moderate hail and continuous lightning were observed on these passes. Unfortunately the engine on the T-28 seized just before landing and the aircraft is down for the rest of STEPS. The aircraft was able to land at the Goodland airport, thanks to the tremendous skill of pilot Tom Root. We are all thankful that Tom made it back to Goodland safely. The T-28 provided 14 flights for STEPS and performed very well.

The mobile mesonet got great observations including hail intercepts and a tornado siting.

Radar observations were ended around 0245 UTC 30 June, with CHILL examining the hailstorm as it moved off to the SE. To date this event marks the top case collected in STEPS.
[back to calendar]

At 2342Z, NSSL5 dispatched to north of Wray for a possible anvil launch, but the EFM was never launched. Secured operations around 0015Z, July 2.
[back to calendar]

Around 2050Z, the ballooning unit came west to intercept the western squall-line, but it didn't hold together once the outflow had outrun the line itself. The line and all other echoes were covered with multi-elevation angle surveillance scans with RHIs every 5-10 deg along the line.

Operations were secured at 2356 Z.
[back to calendar]

By 2309Z, we had a strong, classic supercell in Nebraska and a weaker cell in northwestern Kansas near St. Francis area. This storm southwest of the supercell began to exhibit classic LP characteristics such as a very erect cloud, with very little precipitation out the base as reported by Morris Weisman who had accompanied the ballooning unit.

At 2315Z, the Goodland WFO issued a severe storm warning and at 2333 issued a tornado warning (based on radar). At 2342Z NSSL5 reported a well-defined wall cloud and we had a definite hook echo on the CHILL radar. Tops of the supercell storm were now to 17.5 km above CHILL. S-Pol reported about 20 mm of rain from the LP storm (it was also producing -CGs as seen on the NLDN), and the supercell storm produced about 60 mm of rain (and +CGs) in a one-hour period. The accumulated rain from the LP (visually verified) was contaminated some by the earlier, weaker cells in this same area before Morris' report.

NSSL5 launched 3 EFMs, two of which where in the reflectivity notch on the southeast side of the main reflectivity core. These two almost certainly went into the updraft region, though this is yet to be confirmed. Additionally, the 21Z environmental sounding was taken between the two storms and about 4 mi south of Benkelman. An 18Z sounding had also been taken at Benkelman.

Other NWS reports included: one inch hail (5:50-6:00pm MDT) 3 mi east of Max NE in Dundee Cnty, funnel (6:00pm) 5 mi south of Stratton in Hitchcock Cnty and golfball-sized hail 4 mi west of Stratton, and golfball-sized hail (6:40pm) 6.5 mi southeast of Stratton.

The northeastern supercell storm formed on the moist-air side of the boundary near the NE-KS border, while the LP storm formed on the drier-air side.

Sydney storm: By 0214Z, we abandoned the Benkelman storms since the last balloon was on its way down and turned our attention to a severe supercell that had passed near Sydney NE where it produced funnel clouds and some tornadoes. This storm had taken a marked right turn from its original eastward track and was projected to pass southestward near S-Pol and continue on toward Goodland if it survived. This storm had 65 dBZ and strong rotational characteristics as determined by both research radars. S-Pol was closest to the storm and had a well-defined tornadic signature. This storm also produced golfball-, baseball-, and grapefruit-sized hail in the area of Haxton-Holyoke CO. The largest sizes have not been confirmed. By 0330Z the storm was clearly weakening and beginning to turn back eastward. We secured operations at 0423Z when the storm was no longer a threat.

Although the Benkelman storms were 120-140 km from CHILL and 80-120 km from S-Pol, they were well covered in coordinated scans by both radars. S-Pol may have been scanning the Benkelman storms (especially the LP storm) and/or their predecessor(s) when they were much closer, perhaps 20-60 km. Both research radars also scanned the Sydney storm but it was aligned mostly along and north of the SPol-CHILL baseline.

The LMA channel 3 interference did clear up between the Benkelman and Sydney storms.
[back to calendar]

Apparently, the cool front (more a wind shift) line came through the domain past midnight. One severe storm developed in western Kit Carson County and moved northeastward through the western Doppler lobes and produced +CGs. A severe storm warning was issued by Goodland, with the expectation of up to golfball-sized hail and strong winds. The large-size end of the hail size spectrum has not been verified, but the storm did produce hail up to about one inch.
[back to calendar]

An intense storm with golfball-sized hail was reported in Yuma County, but was too far and in poor geometry for effective dual-Doppler work so we stayed with the cells in the eastern lobe. Once these eastern-lobe cells had begun to decay and were now about 50-50 +/-CG, we shifted to very rapidly growing +CG storms along the Nebraska-Kansas border west of Benkelman NE.

Other reports of note: wind gusts to 70 mph and 2.5 inch hail in northwest corner of Cheyenne County KS, also in northwest corner of Rawlins County KS, significant mesocyclonic signature in western most of these storms (Cheyenne Cnty one), 70 dBZ, a positive Zdr column, and a splitting storm east of Imperial Nebraska. Most storms around the perimeter of the 2D lightning mapping were +CGs at some time or another in their lifetimes.

All-in-all, it was a very active day with several storms reaching severe stages with large hail. However, most storms were fairly short-lived, perhaps 1/2-3/4 hour, but still intense in spite of the relatively low values of shear. A key was apparently a fairly deep layer of moisture. We took a couple of surveillance scans near the end and wrapped up operations at 0152Z. Soundings were taken at Goodland at 18, 21, and 24Z. The last sounding likely went through cloud.
[back to calendar]