1. Introduction
The existence of hook echoes and rear-flank downdrafts (RFDs) in supercell thunderstorms has been well documented over the past several decades (Markowski 2002b; Stout and Huff 1953; van Tassel 1955; Fujita 1958, 1973, 1975; Browning and Donaldson 1963; Browning 1964, 1965; Lemon 1977; Burgess et al. 1977; Brandes 1978; Barnes 1978a, b). Initial investigations attempted to directly link hook echoes with tornado occurrences, but recent research has confirmed that supercell thunderstorms possessing hook echoes often fail to produce tornadoes (Markowski et al. 2002, hereafter MSR02). Using direct surface measurements from the Verification of the Origins of Rotation in Tornadoes Experiment (VORTEX; Rasmussen et al. 1994) and subsequent endeavors, general conclusions have been made relating RFD thermodynamics (equivalent potential temperature and virtual potential temperature) to tornadogenesis success and failure (MSR02; Markowski 2002a, hereafter M02; Grzych et al. 2007, hereafter GLF07). RFD outflow parcels that possess small (large) equivalent potential temperature and virtual potential temperature deficits as compared to those found in the storm inflow environment are thought to be associated with tornadogenesis success (failure). These observational findings were all focused on time periods very near tornadogenesis (±5 min). These conclusions were further reinforced through a numerical simulation conducted by Markowski et al. (2003). The present study attempts to examine the temporal and spatial variability of the thermodynamic and kinematic properties within RFDs while a tornado is ongoing.
Four “mobile mesonet” instrumented vehicles, modeled after those designed by Straka et al. (1996), were used by the Texas Tech University (TTU) Atmospheric Science Group and the TTU Wind Science and Engineering Research Center to collect data during the Wheeled Investigation of Rear-Flank Downdraft Lifecycles (WIRL) project in May and June of 2004 and 2005. Project WIRL’s efforts yielded two RFD outflow samples suitable for analysis. The first dataset was collected from a supercell located west of Lehigh, Iowa, on 12 June 2004 (case 1). The second dataset was acquired from the 9 June 2005 supercell, which passed south of Hill City, Kansas (case 2). These cases offer differing perspectives of the thermodynamic and kinematic variability within tornadic RFDs.
2. Methodology
a. Data collection
Mobile mesonet (MM) data were collected utilizing the instrumentation presented in Table 1. Instrumentation and hardware were mounted to an aluminum rack following the design of Straka et al. (1996) to limit the effects of vehicle motion on wind and pressure measurements. Consistent with the original design, a PVC housing was mounted to the rack mast to protect both temperature and relative humidity sensors from water, debris, and direct contact with sunlight. A sheet of aluminum flashing was added surrounding the PVC housing to shield the interior sensors from radio frequency (RF) noise. An electric fan was inserted into the bottom of the PVC housing to allow for a continuous aspirated airflow through the entire enclosure. Because of erroneous measurements made by the flux gate compass as a result of RF noise, the compass was removed from the instrumentation suite. Data were collected at 0.5 Hz and were continuously processed and stored by a Campbell Scientific CR23X datalogger. The datalogger, pressure sensor, and GPS receiver were housed in a hail-resistant enclosure mounted to the mast just above the vehicle’s roof (Fig. 1). Data were simultaneously transferred via serial interface from the datalogger to an on-board laptop computer running LabVIEW software. This real-time display allowed team members to better identify the RFD gust front and other noteworthy features. The entire system, including radio communications, was powered by the vehicle’s battery.
Two primary data collection routines were executed through the duration of Project WIRL. The life cycle routine was executed for slow-moving supercells or storms following a densely gridded road network. The goal of this routine was to sample various regions of the RFD outflow for as long as logistically permitted. Three individual MM probes attempted to position themselves with a 1- to 2-km spacing in the western, southern, and eastern portions of the RFD outflow. The fourth MM probe was dedicated to continuous sampling of the inflow environment so direct thermodynamic comparisons could be made between the various RFD sectors and the inflow “base state” (Fig. 2). The inflow probe was charged with helping to identify potential low-level mesocyclogenesis, storm cycling, and storm transition (e.g., into a high-precipitation supercell).
In most cases, storm motion and/or road conditions did not allow for the life cycle routine to be executed successfully. A second data collection method, the snapshot routine, was developed for situations that involved these prohibitive logistical challenges. The probes positioned themselves along a single north–south road and allowed the RFD to pass directly overhead (Fig. 3). A probe spacing of 1–2 km was desired, with one probe positioned in the ambient inflow environment, if possible. Once an individual snapshot sample was obtained, the probes attempted to reposition themselves to repeat the procedure as roads allowed.
In all cases during Project WIRL, the RFD was initially identified visually through the manifestation of a clear slot (Beebe 1959). Because of logistical challenges, the team was unable to conduct successful sampling prior to the development of this feature, suggesting that the processes driving the RFD were already ongoing when the team made its initial measurements.
b. Analysis methodologies and techniques
To ensure data consistency among the four MM probes, parameter value comparisons were made while the team was traveling in a caravan and in quiescent conditions. Instrument offsets for each probe were computed to remove any individual instrument biases relative to the overall team mean calculated using data from all the probes. The final presented data were constructed using 10-s, fully segmented averages.
Ground-relative velocity data were derived by subtracting the GPS vehicle velocity from the measured wind velocity. To account for GPS drift, a constant vehicle heading was assumed once the vehicle’s speed dropped below 2.57 m s−1. Therefore, accurate wind data could be collected while a probe was stationary. Care was taken to ensure that the vehicles maintained a constant heading while decelerating. Wind velocity and pressure measurements were influenced when the vehicles made sharp changes in either speed or heading. Calculated wind speeds and directions were discarded from analysis when the vehicles’ acceleration exceeded 1.0 m s−2, in agreement with the threshold established by M02.
c. Limitations in comparing this study to previous research
The data collection routines and analysis techniques utilized in this study are not equivalent to those used by previous studies (M02; MSR02; GLF07), and the reader is urged to exercise caution when directly comparing these results.
1) Definition of the inflow base state
Determination of the inflow base state is arbitrary and imperfect. M02 and MSR02 utilized representative soundings and surface observations to determine a relatively broad inflow base state; however, the ability of these observations to accurately describe the storm inflow environment is uncertain (M02). For Project WIRL, the surface inflow base state was characterized utilizing measurements from a single MM platform residing east of the RFD gust front, within the storm inflow environment. It is thought that air parcels in this region are likely more representative of what the updraft may ingest (GLF07). This inflow probe was always located under anvil cirrus, and therefore shaded from the sun, suggesting that inflow temperatures were likely cooler using this method than observations located at greater distances from the storm.
2) Thermodynamic variables
3) Representativeness of observations
In an effort to analyze air parcels that were most susceptible to updraft ingestion, only those observations taken within a 4-km radius of the low-level mesocyclone and tornado were considered by M02, MSR02, and GLF07. Because the focus of this study was to document temporal and spatial variability within the RFD outflow, this strict condition was not upheld. Therefore, RFD outflow characteristics at distances greater than 4 km cannot be directly compared to previous findings.
3. Case 1–12 June 2004, Lehigh, supercell
a. Synoptic/mesoscale setting and observation strategy
Conditions at 0000 UTC on 12 June 2004 were sufficient to support supercell thunderstorms west of Des Moines, Iowa, northward to Minneapolis, Minnesota. A 20-m s−1 wind speed maximum had rotated around the base of a negatively tilted 500-hPa trough and was located over the region. At the surface, an eastward-translating low pressure center was located over the Iowa–Minnesota border (Fig. 4). A dryline extended southwestward from near the surface low through eastern Kansas and across the Texas Panhandle. A warm front extended east-southeastward across northern Iowa into northern Illinois. A modified upper-air sounding from Topeka, Kansas (Fig. 5),1 showed extreme instability with a surface-based convective available potential energy (CAPE) value in excess of 5700 J kg−1. The original sounding contained an erroneous surface dewpoint value, which was modified by averaging the four nearest surface dewpoint observations closest to the time of the launch. Rapid Update Cycle (RUC) analysis shows that this conditional instability also existed northward through central and western Iowa by 0000 UTC. Ahead of the dryline in central Iowa, small temperature–dewpoint spreads resulted in a low lifted condensation level of near 800 m while southeasterly surface winds aided in 0–3-km storm-relative helicity values near 200 m2 s−2. In this kinematic environment, southeasterly surface winds increase the storm-relative inflow, aligning it with the low-level shear vector.
Thunderstorm initiation occurred at the dryline–warm front triple point around 2100 UTC on 11 June 2004 and built southwestward along the dryline in north-central Iowa. By 0000 UTC on 12 June, a line of strong-to-severe storms existed across north-central Iowa, while another line of convection extended southwestward from Des Moines into northwest Missouri. Over the next hour, isolated tornadic supercells developed along the dryline between the two thunderstorm clusters (Fig. 6).
Moving southward from the northern thunderstorm complex, Project WIRL targeted isolated cells along the dryline. At 0000 UTC the team intercepted new convective development near Burnside, Iowa. This storm organized quickly and a low-level mesocyclone became visually evident by 0020 UTC. The decision was made to execute the life cycle data collection routine because of an optimal road network and slow storm motion (from 260° at 7 m s−1). Probes T3, T2, and T4 were sent west, south, and southeast, respectively, of the low-level circulation, while probe T5 was positioned farther east in the inflow environment. A 0032 UTC snapshot view of the deployment is shown in Fig. 7; tornado-relative probe positions are shown in Fig. 8. All probes drifted from their tornado-relative positions at various times in the sample but generally remained within close proximity of their assigned locations (Fig. 2). The storm produced one weak tornado (rated F1) at 0022 UTC that lasted 20 min and temporarily exhibited multiple vortex structure. At least one farmstead sustained damage.
b. Thermodynamic analysis
As the team approached the RFD boundary, the surface inflow conditions were characterized by a temperature of 28°C and a dewpoint of 22°C. Probe T2 entered the RFD first at 0024 UTC, followed by T3 and T4. Heavy rain and 2.54-cm hail began to wrap cyclonically around the low-level mesocyclone only a few minutes after tornadogenesis.2 The general downward trend in θe (Fig. 9) through the sample period can be explained by a decrease in potential temperature; however, the variability contained within the record is linked to fluctuations in mixing ratio. Deficits in θe of 5–10 K were measured by all the probes within the RFD outflow through the first 10 min of the sample.3 Through this time period, a decrease in both mixing ratio and potential temperature of 5 g kg−1 and 6 K, respectively, accounted for the observed deficit. Deficits in θe of 10–17 K were then measured during the last 10 min of the sample period. Perturbations in θυ also trended downward with time (Fig. 10), with deficits less than 3 K through the first 10 min of the sample and then 3–5 K through the last 10 min. Deficits in θe (θυ) of 3 K (0.5 K) measured by all the probes within the first 2 min of the RFD outflow sample were consistent with the composite tornado-favorable environments of M02, MSR02, and GLF07.
Multiple small-scale features were superimposed on the general thermodynamic trends. Prior to 0035 UTC, measured θe was relatively uniform across the team array. Deficits in θυ were smaller for T4 (located southeast of the tornado), followed by T2 (located south of the tornado), and then T3 (located west of the tornado). After 0035 UTC, probes T3 and T4 measured parcels of highly variable θe yet relatively stable θυ. Near 0037 UTC, T3 (T4) measured θe deficits of 4 K (14 K) while separated by approximately 4 km. During this time, T2 was transitioning back into the inflow environment. At 0042 UTC, θe decreased to a deficit of 16 K at the location of T4, and by the end of the sample, the two probes remaining within the RFD outflow sampled deficits of approximately 20 K.
c. Kinematic analysis
Sampled wind speeds within this RFD outflow were quite weak, though they trended stronger with time as precipitation became more dominant. Measurements of at least 8 m s−1 were observed by all the probes (Fig. 11) within the first few minutes of entering the RFD. At 0026 UTC, T2 observed a small-scale increase in wind speed from 5 m s−1 to a peak value near 14 m s−1 coincident with the positive θe anomaly previously discussed. Wind direction also backed during this period, from near 300° to 250°, followed by a veering back to near 300° as the wind speed subsided. A similar pattern was repeated at 0029 UTC, though the wind speed maximum was only 9 m s−1. It is speculated that these anomalies may have been associated with smaller embedded surges within the RFD outflow. Unfortunately, this feature would have translated through the area southeast of the tornado, while T4 was east of the RFD boundary in the inflow environment. Therefore, the potential evolution of this feature could not be documented.
The inflow vehicle, T5, experienced a gradual backing in wind direction from 0025 to 0031 UTC, accompanied by a sharp increase in wind speed from near 3 m s−1 to a peak of 12 m s−1. It is believed that this change in wind speed and direction was because of the probe moving eastward away from a short residence in the RFD outflow convergence zone. Between 0038 and 0041 UTC, T4 sampled an increase in wind speed from 8 to 18 m s−1 coincident with a 12 K decrease in θe, suggesting that in this region of the RFD, midlevel entrainment of dry air enhanced negative buoyancy and, potentially aided by precipitation drag, resulted in a downward acceleration of air parcels. The record of probe T3, located approximately 2 km away, shows no kinematic indication of this feature.
4. Case 2–9 June 2005, Hill City, supercell
a. Synoptic/mesoscale setting and observation strategy
The environment on 9 June 2005 supported the development of tornadic supercells over western Kansas near a southwestward-moving outflow boundary. An 1800 UTC sounding from Dodge City, Kansas (Fig. 12), showed conditional instability with over 3100 J kg−1 of surface-based CAPE. Regional objective analysis showed that this conditional instability existed over most of western Kansas. Surface winds were backed north and east of an outflow boundary, oriented roughly west-northwest to east-southeast by 2100 UTC, when compared to those winds to the south and west of this boundary (Fig. 13). This backing helped to increase the 0–3-km storm-relative helicity values to greater than 300 m2 s−2 along and east of the boundary (as discussed in case 1). Multiple enhancements in the low-level cumulus field are noted in close proximity to the position of the outflow boundary (Fig. 14). A dryline in western Kansas mixed eastward and provided an initiating mechanism for multiple supercells beginning around 2000 UTC.
An isolated storm initiated on the dryline south of the outflow boundary organized quickly (Fig. 15) and was intercepted 20 km southwest of Hill City, around 2100 UTC. Tornadogenesis (Fig. 16) occurred at 2122 UTC, and storm updraft interaction with the outflow boundary is believed to have played a role in the development of the tornado, as discussed by Markowski et al. (1998) and Rasmussen et al. (2000). Because there was only one paved road available in the area, the decision was made to execute a north–south snapshot deployment in advance of the RFD outflow as it propagated east-northeastward (Fig. 17). Probe T2 was unable to remain in the inflow environment, negating a direct RFD-inflow comparison through the sample duration. The position of probe T2, however, allowed for a southward extension of the linear array, which provided the opportunity to capture a larger spatial representation of the RFD outflow. Tornado-relative probe positions are shown in Fig. 18. Between 2138 and 2141 UTC, probes were positioned into a linear array on U.S. Highway 283. Over the next 10 min, small southward adjustments were made, but the interprobe spacing was largely preserved.
Just west of U.S. Highway 283 (Fig. 17), the circulation underwent significant structural change. Video evidence from positions north and south of the tornadic circulation indicated a horizontally expanding dust/debris field, surrounding a small condensation funnel. At that time, the complete circulation diameter was several times wider than that shown in Fig. 16. This tornado was officially rated F2 on the Fujita scale by the National Weather Service.
b. Thermodynamic analysis
Between 2100 and 2105 UTC, when all the probes were located south of the outflow boundary, inflow conditions at the surface averaged across all four probes were characterized by a temperature of 32°C, a dewpoint of 18°C, and a mean θe (θυ) value of 355.3 (313.8) K. After crossing to the north side of the outflow boundary, but prior to intersecting the RFD outflow, mean surface inflow conditions (averaged across all four probes during the time they were stationary at position P in Fig. 17) were characterized by a temperature of 28°C and a dewpoint of 18°C, yielding a mean θe (θυ) value of 352.8 (311.7) K. The probes were also under anvil shadow during this time period. All vehicles encountered the RFD boundary between 2127:30 and 2128:15 UTC starting with T5 and followed by T4, T3, and T2. Because T2 could not remain within the inflow environment during data collection, base-state values were determined from the mean inflow sample taken to the north of the outflow boundary. As it is impossible to determine the true base state (e.g., high θe air to the south of the outflow boundary may have been ingested in the low to midlevels of the storm), the reader is urged to use caution when interpreting the absolute magnitude of thermodynamic perturbations relative to the uncertain base state. This distinction, however, has no influence on the magnitude of the spatial variability in the thermodynamic variables presented within the RFD outflow. All the probes encountered precipitation during this RFD outflow sample, but T4 and T5 experienced the heaviest precipitation and were the only two teams to observe hail (greater than 1.9 cm in diameter).4
This RFD outflow sample contained large spatial and temporal thermodynamic variability, particularly with respect to θe. Prior to the tornado crossing U.S. Highway 283 at 2136 UTC, several regions of very warm θe air were sampled by all the probes. Probe T2 (the southernmost probe) sampled small θe deficits (less than 2 K) while T3, T4, and T5 recorded multiple instances of large positive perturbations (2–5 K; Fig. 19). As mentioned earlier, the magnitude of these perturbations is somewhat uncertain because of the definition of the base state. These large increases in θe were largely generated by mixing ratio increases of 1.5–2 g kg−1. Between 2131 and 2133 UTC, large positive θe perturbations were sampled through the middle of the array, demonstrating the transient features that comprised this RFD outflow. By 2135 UTC, prolonged positive perturbations were sampled by the northern two probes, T5 and T4, while T3 still measured a small θe deficit of 1 K compared to 4 K by T2. Deficits in θυ prior to the tornado passage across U.S. Highway 283 increased with time yet remained small (Fig. 20). Two small but sharp increases in θυ were measured by T3 just after 2131 UTC, coincident with two large positive θe perturbations. This identical trend in both θe and θυ was also sampled by T5 at 2135 UTC. By 2136 UTC deficits in θυ were 2–3 K for all probes. These measurements are consistent with those findings of M02, MSR02, and GLF07.
The tornado was located 1.8 km northwest of T5 at 2133 UTC5 and crossed U.S. Highway 283 at 2136 UTC 2.7, 3.9, 4.9, and 5.5 km north of T5, T4, T3, and T2, respectively (Fig. 18). The spatial thermodynamic character within the RFD outflow to the southwest of the tornado contained marked horizontal variability. Deficits in θe of 4–6 K were measured by all the probes between 2136 and 2137 UTC, except T5, which sampled zero deficit around 2137 UTC before a negative perturbation of 2 K at 2138 UTC. Between 2138 and 2140 UTC, probes T3 and T4 saw a gradual increase in θe, resulting in perturbations at or above zero for T3 and T4 while T5 and T2 measured deficits of 2 K and 4 K, respectively. Virtual potential temperature leveled off or even increased after 2137 UTC for probes T2, T3, and T4 before a resumed general decrease after 2140 UTC. Probe T5 showed a 60-s plateau around 2138 UTC, after which a decrease continued until 2142 UTC. Probes located farther north (and closer to the tornado) measured colder θυ values, suggesting that surface parcels closest to the tornado were becoming increasingly less buoyant, particularly near the location of T5. At 2142 UTC, as the probes were abandoning the deployment, θυ deficits at the location of probes T2 and T5 were 1.5 and 6 K while separated by only 1 km. At this time the probes were roughly 7.5 km south-southwest of the circulation. Tornado dissipation was estimated to have occurred at 2148 UTC according to the National Weather Service, though the low-level circulation had become completely wrapped in rain from Project WIRL’s perspective around 2140 UTC, prohibiting confirmation of dissipation time.
c. Kinematic analysis
Several kinematic features were measured in this RFD outflow sample that could be tied to the previously discussed thermodynamic fluctuations. Wind directions for T3 and T4 backed locally after 2131 UTC (Fig. 21). Considering the concomitant increase in measured θe (Fig. 19) and a local maximum in measured θυ (Fig. 20), it is unknown what finescale feature may have been sampled or where it may have originated. Once the feature passed around 2133 UTC, wind speeds decreased immediately, especially at the position of T3. Probes T2, T3, and T4 all shared similar wind speed trends as the tornadic circulation approached, highlighted by a significant increase in wind speed as the circulation crossed U.S. Highway 283 north of T5. The T5 recorded an instantaneous (0.5 Hz) peak of 42 m s−1 at 2137 UTC, while peak wind gusts of 37 and 32 m s−1 were recorded by T4 and T3, respectively. Probe T2 did not experience wind speeds greater than 15 m s−1 during this period; this probe was only 1 km south of T3 and 2.5 km south of T5. Following the passage of the circulation to the north of the MM array, a secondary wind maximum was observed between 2138 and 2141 UTC by all the probes. This secondary embedded surge coincides with the increase in θe previously discussed at this time. Though high-resolution radar data were not available to capture this feature, this kinematic surge may be similar to that sampled by Finley and Lee (2004) and Lee (2004) and to the double gust front structure discussed by Wurman et al. (2007) through dual-Doppler analysis.
5. Concluding remarks
Project WIRL collected high-resolution surface data within RFD outflows during the late spring of 2004 and 2005. Data were collected, using a fleet of four mobile mesonet vehicles, on 12 June 2004 near Lehigh, Iowa (case 1), and 9 June 2005 near Hill City, Kansas (case 2). These data were then analyzed to examine thermodynamic and kinematic variability within the RFD outflows of each supercell. Different data collection techniques were employed for each case.
Case 1 provided an RFD outflow sample in which measured θe and θυ perturbations were small (−3 K and −1 K, respectively) within 5 minutes of tornadogenesis, consistent with the findings of M02, MSR02, and GLF07. These parameters trended colder with time over the 20-min duration of the tornado. The magnitude of the cooling was correlated to the duration of precipitation at each probe’s location. Deficits in θe (θυ) of 8–10 (2.5–3.5) K were measured by all probes within the RFD outflow during the life of the tornado (Table 3). A maximum θe deficit of 17 K was measured 2.5 km west-northwest of the tornado just before team visibility of the tornado was obscured by precipitation. This RFD sample contained relatively weak surface winds, especially prior to the onset of precipitation.
The supercell in case 2 appeared to have interacted with a surface baroclinic (outflow) boundary and was highlighted by multiple large tornadoes. Data acquisition occurred during the first tornadic cycle. Analyses indicated only small thermodynamic differences in comparison with the inflow environment. Deficits in θe and θυ were generally less than 4 K until very late in the sample, at which time heavy precipitation was experienced by all the probes located in the RFD outflow (Table 4). In fact, θe excesses of 3–5 K were measured by multiple probes early in the RFD sample. The location for determining the base-state observations north of the outflow boundary likely had an impact on these perturbation values. Higher θe air that originated south of the outflow boundary may have been advected above the outflow boundary, ingested by the storm and mixed down through the RFD. Deficits in θυ generally remained less than 3 K for all teams through the first 10 min of the sample and 5 K for the second 10 min of the sample. Heavy precipitation occurred at each probe location through the entire 20-min sampling period, with hail occurring at the locations of the northern two probes. Several transient features were sampled in this RFD outflow. It is unfortunate that measurements of surface parcel thermodynamics could not be made immediately surrounding the circulation, as the transient nature of this sample indicated the RFD outflow parcel composition was highly variable both spatially and temporally.
In both cases presented, there are observed relationships between localized kinematic maxima and maxima of thermodynamic perturbations of either sign. Surface observations alone cannot determine the cause for these fluctuations. However, these fluctuations reinforce the suggestion that parcels of highly variable thermodynamic composition reside within the RFD and are unevenly transported toward the surface.
Higher-spatial-resolution data from within the RFD in close proximity to the principal updraft are needed to better quantify the thermodynamics affecting a tornadic circulation. Because manned measurement systems are limited to only those portions of the RFD where samples can be made safely, surface thermodynamics from key regions northeast and north of a low-level mesocyclone cannot always be easily obtained. Future observational work will address this challenge with several easily deployable “StickNet” (Weiss and Schroeder 2008) towers arrayed to capture the full horizontal variability and temporal evolution of surface thermodynamics and kinematics in these critical areas. Emphasis will also be placed on sampling the initial RFD intersection with the surface, usually to the west or southwest of the parent low-level mesocyclone, such that the full surface thermodynamic evolution of the RFD can be documented.
Acknowledgments
Funding for the 2004–05 WIRL field campaign was provided by the National Science Foundation Grant IGERT-0221688 and the Texas Tech University (TTU) department of Geosciences. Additional support for data analysis was provided by the Department of Commerce National Institute of Standards and Technology/TTU Cooperative Agreement Award 70NANB8H0059. We thank all those individuals from TTU who volunteered countless hours in the field to make data collection possible. Particular thanks are extended to Ian Giammanco for his time spent with mobile mesonet and communication preparations. We also thank two anonymous reviewers for their suggestions and input.
REFERENCES
Barnes, S. L., 1978a: Oklahoma thunderstorms on 29–30 April 1970. Part I: Morphology of a tornadic storm. Mon. Wea. Rev., 106 , 673–684.
Barnes, S. L., 1978b: Oklahoma thunderstorms on 29–30 April 1970. Part II: Radar-observed merger of twin hook echoes. Mon. Wea. Rev., 106 , 685–696.
Beebe, R. G., 1959: Notes on the Scottsbluff, Nebraska tornado, 27 June 1955. Bull. Amer. Meteor. Soc., 40 , 109–116.
Brandes, E. A., 1978: Mesocyclone evolution and tornadogenesis: Some observations. Mon. Wea. Rev., 106 , 995–1011.
Browning, K. A., 1964: Airflow and precipitation trajectories within severe local storms which travel to the right of the winds. J. Atmos. Sci., 21 , 634–639.
Browning, K. A., 1965: The evolution of tornadic storms. J. Atmos. Sci., 22 , 664–668.
Browning, K. A., and R. J. Donaldson, 1963: Airflow and structure of a tornadic storm. J. Atmos. Sci., 20 , 533–545.
Burgess, D. W., R. A. Brown, L. R. Lemon, and C. R. Safford, 1977: Evolution of a tornadic thunderstorm. Preprints, 10th Conf. on Severe Local Storms, Omaha, NE, Amer. Meteor. Soc., 84–89.
Finley, C. A., and B. D. Lee, 2004: High resolution mobile mesonet observations of RFD surges in the June 9 Basset, Nebraska supercell during Project ANSWERS 2003. Preprints, 22nd Conf. on Severe Local Storms, Hyannis, MA, Amer. Meteor. Soc., P11.3.
Fujita, T. T., 1958: Mesoanalysis of the Illinois tornadoes of 9 April 1953. J. Meteor., 15 , 288–296.
Fujita, T. T., 1973: Proposed mechanism of tornado formation from rotating thunderstorms. Preprints, Eighth Conf. on Severe Local Storms, Denver, CO, Amer. Meteor. Soc., 191–196.
Fujita, T. T., 1975: New evidence from the April 3–4, 1974 tornadoes. Preprints, Ninth Conf. on Severe Local Storms, Norman, OK, Amer. Meteor. Soc., 248–255.
Grzych, M. L., B. D. Lee, and C. A. Finley, 2007: Thermodynamic analysis of supercell rear-flank downdrafts from project ANSWERS. Mon. Wea. Rev., 135 , 240–246.
Koch, S. E., M. desJardins, and P. J. Kocin, 1983: An interactive Barnes objective map analysis scheme for use with satellite and conventional data. J. Climate Appl. Meteor., 22 , 1487–1503.
Lee, B. D., 2004: Thermodynamic and kinematic analysis of multiple RFD surges for the 24 June 2003 Manchester, South Dakota cyclic tornadic supercell during Project ANSWERS 2003. Preprints, 22nd Conf. on Severe Local Storms, Hyannis, MA, Amer. Meteor. Soc., P11.2.
Lemon, L. R., 1977: Severe thunderstorm evolution: Its use in a new technique for radar warnings. Preprints, 10th Conf. on Severe Local Storms, Omaha, NE, Amer. Meteor. Soc., 77–83.
Markowski, P. M., 2002a: Mobile mesonet observations on 3 May 1999. Wea. Forecasting, 17 , 430–444.
Markowski, P. M., 2002b: Hook echoes and rear-flank downdrafts: A review. Mon. Wea. Rev., 130 , 852–876.
Markowski, P. M., E. N. Rasmussen, and J. M. Straka, 1998: The occurrence of tornadoes in supercells interacting with boundaries during VORTEX-95. Wea. Forecasting, 13 , 852–859.
Markowski, P. M., J. M. Straka, and E. N. Rasmussen, 2002: Direct thermodynamic observations within the rear-flank downdraft of nontornadic and tornadic supercells. Mon. Wea. Rev., 130 , 1692–1721.
Markowski, P. M., J. M. Straka, and E. N. Rasmussen, 2003: Tornadogenesis resulting from the transport of circulation by a downdraft: Idealized numerical simulations. J. Atmos. Sci., 60 , 795–823.
Rasmussen, E. N., J. M. Straka, R. P. Davies-Jones, C. A. Doswell III, F. H. Carr, M. D. Eilts, and D. R. MacGorman, 1994: Verification of the Origins of Rotation in Tornadoes Experiment: VORTEX. Bull. Amer. Meteor. Soc., 75 , 995–1006.
Rasmussen, E. N., S. Richardson, J. M. Straka, P. M. Markowski, and D. O. Blanchard, 2000: The association of significant tornadoes with a baroclinic boundary on 2 June 1995. Mon. Wea. Rev., 128 , 174–191.
Stout, G. E., and F. A. Huff, 1953: Radar records Illinois tornadogenesis. Bull. Amer. Meteor. Soc., 34 , 281–284.
Straka, J. M., E. N. Rasmussen, and S. E. Fredrickson, 1996: A mobile mesonet for finescale meteorological observations. J. Atmos. Oceanic Technol., 13 , 921–936.
van Tassel, E. L., 1955: The North Platte Valley tornado outbreak of June 27, 1955. Mon. Wea. Rev., 117 , 255–264.
Weiss, C. C., and J. L. Schroeder, 2008: StickNet—A new portable, rapidly-deployable, surface observation system. Preprints. 24th Conf. on IIPS, New Orleans, LA, Amer. Meteor. Soc., P4A.1. [Available online at http://ams.confex.com/ams/pdfpapers/134047.pdf.].
Wurman, J., Y. Richardson, C. Alexander, S. Weygandt, and P. F. Zhang, 2007: Dual-Doppler and single-Doppler analysis of a tornadic storm undergoing mergers and repeated tornadogenesis. Mon. Wea. Rev., 135 , 736–758.
TTU MM platform including wind monitor (A), static pressure port (B), GPS antenna (C), PVC housing (D) containing “fast” temperature and “slow” temperature–relative humidity probes, and hail-resistant enclosure (E) containing pressure sensor, datalogger, and GPS receiver. The background tornado occurred in rural Kent County, TX, on 12 Jun 2005. Photo taken by Ian Giammanco.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Schematic diagram of the life cycle data collection routine. The RFD boundary is indicated by a solid black line while individual MM teams are marked with an M. Location of the low-level mesocyclone or tornado is denoted with a T. The storm motion vector is provided. Vehicles were to utilize the road network to maintain their respective mesocyclone or tornado-relative positions as much as possible as the storm moved eastward. The idealized road grid spacing is approximately 1.6 km × 1.6 km.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Same as Fig. 2, but for the snapshot data collection routine.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Subjective surface analysis from 0000 UTC 11 Jun 2004. The surface low pressure center, warm front, and dryline are illustrated using standard synoptic symbols. Objectively analyzed isodrosotherms are also contoured every 5°F as thin gray lines.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Modified skew T–logp diagram from Topeka, KS, at 0000 UTC 12 Jun 2004.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
KDMX NEXRAD image on 0055 UTC 12 Jun 2004. The black arrow indicates the storm that was targeted for WIRL operations.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Locations of all four TTU MM probes at 0032 UTC 12 Jun 2004. The T denotes the location of the tornado at this time. The “P” indicates where all probes were located at 0022 UTC when the surface circulation was first documented. The dashed line indicates the tornado’s path until it was no longer visible at 0042 UTC. An inset of the state of Iowa is provided for reference.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Locations of TTU probes (a) T3, (b) T2, (c) T4, and (d) T5 relative to the tornado during the sampling period. The origin represents the location of the tornado. Data points represent relative locations at 1-min increments with the times of 0025 (I), 0030 (II), 0035 (III), 0040 (IV), and 0045 (V) UTC identified. Range rings are incremented every 2 km with the innermost ring representing 2 km.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Time series of θe perturbation (K) compared to that of the inflow environment (T5) during the 12 Jun 2004 Lehigh, tornadic RFD sample. Above the time series, down arrows denote the time that individual MM probes entered the RFD from the inflow, while up arrows denote the time that individual MM probes entered the inflow environment from the RFD. The thin solid bars at the bottom of the plot indicate the duration of precipitation for each probe, with circles indicating hail duration, where documented. The bold bar at the bottom of the figure denotes the time duration the tornado was visible by WIRL.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Same as Fig. 9, but for θυ (K).
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Plots of ground-relative wind speed (solid line, m s−1) and direction (dotted line, degrees) for (a) T3, (b) T2, (c) T4, and (d) T5 during the 12 Jun 2004 Lehigh, tornadic RFD sample. The horizontal bar along the bottom of each plot denotes the time duration the tornado was visible to WIRL. Downward-pointing arrows indicate when each probe entered the RFD, while upward-pointing arrows indicate when each probe exited the RFD, where applicable.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Skew T–logp diagram from Dodge City, KS, at 1800 UTC 9 Jun 2005.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Subjective surface analysis from 2100 UTC 9 Jun 2005. The primary surface outflow boundary is indicated by a dashed line. Objectively analyzed isodrosotherms are also contoured every 5°F as thin gray lines.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Visible satellite imagery at (a) 1910, (b) 1940, (c) 2010, (d) 2040, and (e) 2045 UTC 9 Jun 2005. The location of outflow boundaries and the dryline are illustrated, and the target storm is denoted by the letter “S.”
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
KGLD NEXRAD image from 2008 UTC 9 Jun 2005. The black arrow indicates the storm that was targeted for WIRL operations.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Photograph looking west-northwest of the tornado on 9 Jun 2005 that was 20 km southwest of Hill City. Photo taken by Eric Thoen at 2125 UTC.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Locations of all four TTU MM probes at 2134:30 UTC 9 Jun 2005. The T denotes the location of the tornado at this time. The P indicates where all the probes were located at 2122 UTC when the surface circulation was first documented. The dashed line indicates the path of the tornado, determined by the National Weather Service. An inset of the state of Kansas is provided for reference.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Locations of TTU probes T2 (open circles), T3 (closed circles), T4 (asterisks), and T5 (upside-down triangles) relative to the tornado during the sampling period. Traces are marked in 1-min increments with the times of 2125 (I), 2130 (II), 2135 (III), and 2140 (IV) UTC identified. Range rings are incremented every 2 km with the innermost ring representing 2 km.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
(a) Time series of θe perturbation (K) during the 9 Jun 2005 Hill City tornadic RFD sample. Above the time series, down arrows denote the time that individual MM probes entered the RFD from the inflow. The thin solid bars at the bottom of the plot indicate the duration of precipitation for each probe, with circles indicating hail duration, where documented. The bold bar at the bottom of the figure denotes the time duration the tornado was on the ground and visible to WIRL. (b) Time-to-space conversion of the data presented in (a) from 2130 to 2140 UTC (this time window is denoted by the solid vertical lines) with station plots overlaid. Each station plot represents a 10-s-avg plotted every 30 s. Wind barbs are ground relative. The T at the origin represents the location of the tornado, and U.S. Highway 283 is also shown.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Plots of ground-relative wind speed (solid line, m s−1) and direction (dotted line, degrees) for (a) T5, (b) T4, (c) T3, and (d) T2 during the 9 Jun 2005 Hill City tornadic RFD sample. The horizontal bar along the bottom of each plot denotes the time duration the tornado was visible to WIRL. Downward-pointing arrows indicate when each probe entered the RFD.
Citation: Monthly Weather Review 136, 7; 10.1175/2007MWR2285.1
Instrumentation and accessories used on the MM platform.
Changes in θυ with reflectivity, when omitting the liquid water vapor mixing ratio term from the equation of virtual potential temperature. In this example, potential temperature equals 305 K and mixing ratio equals 14 g kg−1.
Summary statistics for each MM probe during the Lehigh tornadic RFD sample. Time (UTC), comments on existing conditions, relative humidity (RH), equivalent potential temperature (θe), virtual potential temperature (θυ), wind speed (WS), and wind direction (WD) are all shown. T3, T2, and T4 θe and θυ measurements are perturbations from the base state (inflow) as measured by T5. The tornado was no longer visible to team members after 0042 UTC.
Summary statistics for each MM probe during the Hill City tornadic RFD sample. Time in UTC, comments on existing conditions, relative humidity (RH), equivalent potential temperature perturbation (θ′e), virtual potential temperature perturbation (θ′υ), wind speed (WS), and wind direction (WD) are all shown. The base state of θ′e is 352.8 K and of θ′υ is 311.7 K. The tornado was no longer visible to team members after 2140 UTC.
The sounding from Topeka was preferred because the dryline had mixed through the closer location of Lincoln, NE, by 0000 UTC, and midlevel lapse rates were thought to be more representative of the storm environment than at the location of Davenport, IA.
For the 12 June 2004 case, precipitation times and type were determined using video footage collected from each individual probe. All first encounters with precipitation were accurately accounted; however, T2 did not have a complete record through the duration of the sampling period. It will be assumed that precipitation ceased when this probe transitioned back into the inflow environment. Hail may also have been encountered by T2 at some point, but cannot be verified.
The plateau in θe measured by T4 from 0027 to 0032 UTC was a result of crossing back into the inflow temporarily.
For the 9 Jun 2005 case, precipitation times and type were determined using video footage collected from each individual probe. All first encounters with precipitation were accurately accounted, and all teams remained in precipitation through the duration of the sample. Hail was experienced by T4 and T5 at 2131 UTC; however, it is unknown when the hail fall ended as accurate documentation ceases after 2140 UTC while hail was still ongoing. Given the southward motion of these two teams, and the lack of measured hail by the two probes farther south, it was assumed that the teams did not encounter hail much beyond this time.
The center of the tornado path is based on National Weather Service documentation. Given the width of the circulation at this time, T5 was closer than 1.8 km to the outer periphery of the tornadic circulation; however, the exact distance cannot be accurately determined.
