1. Introduction
Current trends in radar-based research into the structure and dynamics of severe storms have been toward increasing spatial and temporal resolution of the data. One way to achieve finer resolution is to move the radar closer to the storm. Use of mobile radars to observe convective storms and tornadoes has increased significantly in recent years, beginning with narrow beam W-band systems (e.g., Bluestein et al. 1995; Bluestein and Pazmany 2000), X-band systems (e.g., Wurman et al. 1997; Wurman and Gill 2000; Bluestein et al. 2007a), and C-band radars (Biggerstaff et al. 2005). More recently, several mobile platforms have added dual-polarization capabilities, including the University of Massachusetts “X-Pol” (Bluestein et al. 2007b), the polarimetric Doppler-on-Wheels (DOW; Wurman 2001), and the National Severe Storms Laboratory/University of Oklahoma polarimetric X-band Experimental Radar for the Examination of Storms (XERES; see Melnikov et al. 2009). Mobile radars allow researchers to collect data in close proximity to storms with increased spatial resolution. Mobile and stationary research radars achieve higher sampling rates, as often the data are collected in a sector focused on the storm, cutting down on the “wasted” time that the dish rotates and collects clear-air data.
Another emerging interest in the radar community is the use of phased array antenna (e.g., Zrnić et al. 2007; Bluestein et al. 2008a,b) and slotted waveguide array antenna (Wurman and Randall 2001) technology for rapid scanning of severe storms. In contrast to the mechanically scanning parabolic dish antennae of the Weather Surveillance Radar-1988 Doppler (WSR-88D) radars, the antennae of phased array radars are used to electronically steer the beam, allowing for more agile scanning. Recently, Heinselman et al. (2008) investigated the evolution of severe storms with the National Weather Radar Test bed (NWRT) Multifunction Phased Array Radar (MPAR) in Norman, Oklahoma. They reported that the MPAR can perform a standard volume coverage pattern 12 strategy (VCP-12; see Brown et al. 2005a) in 58 s compared to the conventional WSR-88D scan time of 258 s. However, the MPAR currently only scans a 90° sector, whereas the WSR-88D scans a full 360°, so the improved scan time is almost entirely due to a smaller coverage volume. The increased temporal resolution capability of the existing MPAR comes with trade-offs: angular resolution (1.5°–2.1°) is worse than the WSR-88D (0.9°), especially for viewing angles off broadside [figures in Heinselman et al. (2008) display MPAR data that have been oversampled in azimuth and range and WSR-88D data that have degraded spatial resolution]. MPAR will eventually be polarimetric; however, several issues and technical challenges need to be resolved before such systems come to fruition (e.g., Zhang et al. 2008, 2009).
The main focus of this study is the examination of changes in bulk microphysical characteristics (inferred from polarimetric radar data) in severe storms at time scales on the order of 1 min. In doing so, the following questions are addressed: What types of microphysical changes are present? Do the polarimetric variables change more or less rapidly than the conventional radar variables? What features will be revealed? What will dual-polarization MPAR offer? Steady-state polarimetric signatures consistent in numerous supercell storms have been documented (Kumjian and Ryzhkov 2008, hereafter KR08). However, to better understand these features it is necessary to analyze their evolution, including their genesis, mature state, and demise.
To address these questions, a scanning strategy was developed for the polarimetric prototype WSR-88D research radar operated by the National Severe Storms Laboratory in Norman, Oklahoma (KOUN). The scanning strategy achieves both rapid volumetric updates of the storm as well as fine spatial resolution. To improve the update time, the radar scans only a limited sector of 80°. The data are oversampled to achieve 0.5° azimuthal spacing, known as “super resolution” (Brown et al. 2002, 2005b). The details of this new scanning strategy (dubbed the “supersector”) are found in Table 1. Note that the volume update time is comparable to that of the MPAR (71–73 s), and improved spatial resolution and polarimetric diversity allow for detailed observations of the bulk microphysical properties within the storm.
One goal of high-resolution observations of convective storms is to better understand the process of tornadogenesis. Of particular interest is why some supercell storms exhibit a periodic regeneration of low-level vorticity maxima. In some cases, such behavior results in cyclic tornado development. The intriguing cyclic nature of supercell storms has prompted numerous studies using both observational data (e.g., Darkow and Roos 1970; Lemon and Doswell 1979; Burgess et al. 1982; Dowell and Bluestein 2002a,b; Beck et al. 2006; French et al. 2008) and numerical models (Adlerman et al. 1999; Adlerman and Droegemeier 2002, 2005). However, none of these studies has thoroughly explored the microphysical characteristics of cyclic supercells. This paper provides the first such investigation using data collected with the supersector of a cyclic nontornadic supercell storm in western Oklahoma on 1 June 2008.
The next section provides a brief review of results of previous research into the cyclic nature of supercell storms, including recent conceptual models that will serve as the basis for the current study. In section 3, a primer on the interpretation of polarimetric observations and an overview of the data and event are described. Section 4 presents the data and explores the evolution of the supercell polarimetric signatures throughout the cyclic occlusion process. Section 5 summarizes the main conclusions.
2. Background
Numerous papers have identified a cyclic process of mesocyclone (and potentially tornado) development in some supercell storms. Earlier observational studies based on visual observations (e.g., Darkow and Roos 1970; Darkow 1971), ground-based fixed Doppler radar platforms (e.g., Brandes 1977; Lemon and Doswell 1979; Burgess et al. 1982; Johnson et al. 1987), airborne radar (e.g., Dowell and Bluestein 2002a,b), and mobile radar platforms (e.g., Beck et al. 2006; French et al. 2008) facilitated our understanding of the development of a series of distinct vertical vorticity maxima at low levels as well as provided conceptual models for such cyclic development. These observational studies have been supplemented by numerical simulations of cyclic mesocyclogenesis (Adlerman et al. 1999; Adlerman and Droegemeier 2002, 2005). The body of research indicates that discrete vorticity maxima develop along the rear-flank gust front, likely a result of the tilting of environmental horizontal vorticity into the vertical followed by downstream stretching (e.g., Dowell and Bluestein 2002b). With time these low-level circulations move to the left (with respect to storm motion) as they become completely cut off from the inflow air, or are occluded. As the old circulation occludes, a new cyclonic vorticity maximum develops along the surging rear-flank downdraft (RFD) gust front. A complete cycle is presented in a conceptual model by Dowell and Bluestein (2002b; reproduced here as Fig. 1), which is a slightly modified version of the original conceptual model by Burgess et al. (1982). The numerical studies of Adlerman et al. (1999) and Adlerman and Droegemeier (2002, 2005) suggest that a strong correlation exists between cyclic tornadogenesis and the overall cyclic nature of mesocyclone development. Beck et al. (2006) also highlight this result, suggesting that nontornadic cyclic mesocyclogenesis may be analogous to cyclic tornado development.
Previous studies largely have focused on storm kinematics and dynamics at low levels, as observed through dual-Doppler synthesis. Beck et al. (2006) was the first to develop a conceptual model of the evolution of the radar reflectivity factor ZH field as it relates to cyclic mesocyclone development. They found that the appearance and evolution of the ZH “hook echo” is closely connected with the cycling behavior of the mesocyclone. Their conceptual model (Fig. 2) can be summarized in four main stages: (i) downstream of the hook echo, a deformation zone develops owing to the flow associated with storm inflow and rear-flank downdraft outflow as well as two mesocyclones, the first one being mature and associated with the hook echo and the second developing along the rear-flank gust front; (ii) as the hydrometeors associated with the hook echo encounter the deformation zone, particles are forced north and south, parallel to the axis of dilatation; (iii) as the old mesocyclone shifts rearward (with respect to the main updraft), advection of precipitation by the flow around the new mesocyclone once again results in a more cyclonic appearance of the hook echo; and (iv) the old mesocyclone dissipates in the rear of the storm while the second mesocyclone is mature and wrapped in hook echo precipitation. A third mesocyclone forms along the rear-flank gust front, beginning the cycle over again.
French et al. (2008) use high resolution single-Doppler radar data to track circulations at multiple levels in a cyclic supercell. Their results provide support for the previous conceptual models of cyclic mesocyclogenesis and associated hook echo evolution outlined above. However, they suggest that the rearward movement of circulations may not occur if the circulations develop in a nontraditional manner (i.e., not along the rear-flank gust front). In general, French et al. (2008) found that low- and midlevel circulations tend to be connected and evolve similarly.
The present study builds on the conceptual models and previous studies by adding polarimetry to the analysis of cyclic supercells. Because the entire storm was observed in each volume scan, links between low- and midlevel features can be further explored, following French et al. (2008). Additionally, the unique dataset (S-band, dual-polarization) is less affected by attenuation and differential attenuation, which plague radar observations at shorter wavelengths.
3. Event and data overview
a. Data and methodology
Because of this study’s extensive use of polarimetric radar data for its informative content regarding microphysics in the supercell, the physical interpretation of each polarimetric variable is summarized here. The reader is referred to the literature for more details and applications of polarimetric data (Herzegh and Jameson 1992; Doviak and Zrnić 1993; Zrnić and Ryzhkov 1999; Straka et al. 2000; Bringi and Chandrasekar 2001; Ryzhkov et al. 2005a).
The ratio of backscattered power at orthogonal polarizations is called the differential reflectivity ZDR, which was first proposed for use in observing rainfall by Seliga and Bringi (1976). The differential reflectivity ZDR is dependent on the shape, size, orientation, density, and water content of the hydrometeors in the radar sampling volume, but is independent of their concentration. In rain, ZDR is considered a good measure of the median drop size of a distribution because raindrop oblateness (and thus ZDR) increases with increasing diameter (Pruppacher and Beard 1970). For targets with isotropic scattering properties such as spherical or chaotically tumbling particles, ZDR is 0 dB. For a given particle shape, higher density or liquid water fraction yields higher ZDR. Thus, high ZDR is mostly associated with large, wet hydrometeors.
Jameson (1985) and Sachidananda and Zrnić (1986, 1987) introduced the use of differential propagation phase ΦDP for rainfall measurements. Because raindrops are oblate, the forward propagating horizontally polarized wave encounters more liquid water than the vertically polarized wave. Therefore, in rain the speed of the horizontally polarized wave is less than the vertically polarized wave, resulting in a monotonically increasing difference in phase between the orthogonally polarized waves.1 Because ΦDP is immune to radar miscalibration, is not affected by attenuation and beam blockage (Zrnić and Ryzhkov 1999), and is not biased by noise, it is an attractive variable for quantitative precipitation estimation and attenuation correction. The range derivative of ΦDP is the specific differential phase, KDP, which provides the phase shift per unit distance in the radial direction. Similar to ZDR, KDP increases as the oblateness and dielectric constant of hydrometeors increase and is not affected by spherical or isotropic hydrometeors. However, KDP is sensitive to the concentration of liquid drops (but is nearly zero for heavily aggregated snow or dry graupel/hail). Because of these characteristics and because KDP is nearly linearly related to rainfall rate in pure rain (Sachidananda and Zrnić 1987), KDP is a good indicator of the liquid water content within the radar sampling volume. The KDP is less sensitive to particle diameter than radar reflectivity factor ZH (ZH ∼ D6 whereas KDP ∼ D4.24); thus, KDP is more sensitive to the smaller drop sizes than ZH and ZDR, which are more dominated by larger particles in the sampling volume. In general, regions of high KDP are correlated with regions of high ZH below the melting layer because of each variable’s dependence on concentration. The main difference is that ZH does not distinguish between liquid and frozen particles, whereas KDP does. Other differences in the appearance of regions of enhanced ZH and enhanced KDP can be attributed to the following reasons: ZH is more heavily weighted by contributions from larger particles, KDP is difficult to estimate in the presence of substantial δ, and KDP estimates are prone to errors because of nonuniform beamfilling (NBF; Ryzhkov 2007). Because NBF and other statistical errors (including those due to strong gradients) are prevalent at low levels in severe convective storms such as supercells and KDP offers some degree of redundancy with respect to information offered by ZH, heavy use of low-level KDP features (e.g., Romine et al. 2008) requires careful control of data quality.
The correlation coefficient between the backscattered returns at horizontal and vertical polarizations at zero lag time (ρHV) is sensitive to the diversity of particle sizes, orientations, shapes, and dielectric constants within the sampling volume (Balakrishnan and Zrnić 1990). In the absence of propagation effects, the argument of ρHV is the backscatter differential phase, so the presence of resonance scatterers (for which ℜ ∼ 1 and δ is nonzero) significantly decreases ρHV. Thus, ρHV approaches unity in pure rain or pure dry hail at S band, but is decreased when a mixture of rain and hail is present. Like ZDR, ρHV is most affected by large, wet hydrometeors.
Though the majority of the analysis will focus on the polarimetric data collected using KOUN, supplemental data from the National Weather Service single-polarization WSR-88D radar in Twin Lakes, Oklahoma (KTLX), are also considered. The supersector scanning strategy for KOUN began at 0213 UTC and was used to collect data continuously through 0358 UTC, capturing most of the mature phase of the storm. After this time, a more conventional (360°) scanning strategy was implemented. The data are processed following the procedure in Ryzhkov et al. (2005a), which includes correcting the measured ZH and ZDR for attenuation and differential attenuation using the total differential phase along each radial. The data are calibrated and corrected for noise and plotted in the plan position indicator (PPI) format. With only about 6 s between each elevation scan, the vertical structure of the storm can be analyzed with more validity than in other cases where a full volume takes up to 5–6 min to complete.
Following French et al. (2008), circulations2 are identified and tracked in the Doppler velocity data when they meet the following requirements: (i) must have an azimuthal velocity difference in excess of 25 m s−1, (ii) must have a diameter between about 1 and 9 km, and (iii) must be present in at least two consecutive scans. The circulations are then tracked until they dissipate or are no longer evident in the data, even if the azimuthal velocity difference drops below the first criterion.
b. Environment
A synoptic-scale ridge over the southwestern United States and Mexico provided northwest flow over the southern Great Plains and was the dominant large-scale feature at 0000 UTC 1 June 2008. At the surface, a pronounced wind shift boundary was present associated with a stalled front, where enhanced convergence likely helped initiate convection that evening (Fig. 3). The Norman, Oklahoma, sounding from 0000 UTC 1 June 2008 (Fig. 4) reveals an environment characterized by moderate to high instability (surface-based CAPE ∼ 2900 J kg−1), and over 22 m s−1 magnitude difference in the 0–6-km vector winds. Below 1 km AGL, the hodograph depicts nearly unidirectional shear, followed by considerable veering with height of the shear over the next several kilometers.
c. Storm
Several convective storms developed shortly after 0100 UTC in western Oklahoma on 1 June 2008. One cell rapidly intensified, becoming dominant. Shortly after its rapid intensification (about 0200 UTC), the storm split with the left mover rapidly moving off to the northeast (Fig. 5). The main supercell remained mature over the next few hours, consistently displaying a well-defined hook echo appendage and “flying” or “spread eagle” forward-flank appearance (e.g., see Kumjian and Schenkman 2008) as it moved toward the southeast. The storm displayed a cyclic behavior throughout its lifetime in that it produced a series of mesocyclones, inferred by distinct cyclonic shear maxima in the observed Doppler velocities. At about 0400 UTC, a second supercell merged with the original storm. This merger led to a well-defined hook echo and intense shear couplet in the Doppler velocities (not shown) by about 0500 UTC, though it was short lived. The storm completely dissipated shortly after 0700 UTC.
Storm motion during its mature phase was crudely estimated by tracking radar features over several successive scans. The storm motion was to the southeast, depicted by the X on the hodograph in Fig. 4. This storm motion provides considerably large (>350 m2 s−2) 0–3-km storm-relative helicity (Davies-Jones et al. 1990), which can be an indicator of a storm’s potential to have a mesocyclone and tornado (e.g., Leftwich 1990; Davies and Johns 1993; Colquhoun and Riley 1996; Kerr and Darkow 1996; Rasmussen and Blanchard 1998; Thompson et al. 2003, 2007). However, it is important to note that the low-level environmental horizontal vorticity vector is oriented at a substantial angle to the low-level storm-relative wind vector. Esterheld and Giuliano (2008) propose that the crosswise vorticity indicated by such angles suggests unfavorable conditions for tornadic development. The storm did remain nontornadic throughout its lifetime, though strong low-level rotation prompted the local National Weather Service to issue a tornado warning at 0315 UTC.
Throughout its mature lifetime, the storm produced an impressive amount of lightning: data from the Oklahoma Lightning Mapping Array (see Rison et al. 1999; Thomas et al. 2004 for details) indicate more than 500 sources per square kilometer per minute. Although no tornadoes developed, it did produce numerous severe wind gusts and hailstones, some reported to be nearly 10 cm (4 in.) in diameter [National Climatic Data Center (NCDC) 2008]. The local National Weather Service issued a severe thunderstorm warning at 0415 UTC for grapefruit-sized hail. Figure 5 depicts the evolution of the storm reflectivity pattern throughout its lifetime from KTLX every 30 min.
While a mature supercell, the storm passed near two sites of the Oklahoma Mesonet (Brock et al. 1995): Weatherford and Hinton, Oklahoma. Time series observations from these sites are displayed in Fig. 6. The Weatherford site sampled the rear-flank downdraft, while the Hinton site sampled the outflow within heavy precipitation associated with the forward flank of the storm. Markowski et al. (2002) provides evidence that nontornadic supercell RFDs are characterized by larger virtual potential temperature θυ deficits than tornadic supercells. This finding is corroborated by the modeling study of Markowski et al. (2003) and by a follow-up observational study by Grzych et al. (2007). The RFD outflow sampled at the Weatherford site produces a 3.5-K decrease in θυ after an initial increase of about 2 K; the FFD outflow sampled at the Hinton site has a large θυ deficit (7.5 K). Though no precipitation was observed at Weatherford, Hinton received 27.7 mm of rain in a brief period, at times with a rainfall rate as large as 75 mm h−1.
4. Analysis and discussion
All notable supercell polarimetric signatures display a repetitive pattern that is seemingly related to the occlusion cycle. However, these cycles are not always completed fully; in some cases, cycles appear to be interrupted by a series of rapid occlusions similar to the case discussed in Beck et al. (2006). Nonetheless, this section summarizes the evolution of the Doppler and polarimetric signatures throughout a complete occlusion cycle as well as discusses the microphysical interpretation of these events. The polarimetric signatures discussed herein are the ZH hook echo, low-level hail signature, ZDR arc, KDP and ZDR columns, the updraft signature, and midlevel ZDR and ρHV rings. An example composite image from a volume scan is provided in Fig. 7, which displays the relative locations of some of the notable polarimetric features from different elevations overlaid for comparative purposes. Many of the features are found in and around the updraft and mesocyclone and are addressed separately in sections 4a–h.
a. Mesocyclone evolution
In general, the low-level kinematic features of the storm undergo an evolution closely following the patterns described in previous work (e.g., Burgess et al. 1982; Dowell and Bluestein 2002a,b). Specifically, new mesocyclones formed near the front of the ZH hook echo and moved rearward (in a storm-relative sense) during their occlusion. Figure 8 presents example data from such a cycle. At 0241:47 UTC, three circulations are present: 1) an occluded, dissipating mesocyclone at the rear of the storm (“a”), 2) an occluding mesocyclone in the hook echo (“b”), and 3) a developing circulation ahead of the hook echo in the inflow region (“c”). From 0245:27 to 0249:06 UTC, a small region of cyclonic shear was located along the edge of the FFD echo, called circulation “d.” Shortly after its dissipation, a descending reflectivity core (DRC; Rasmussen et al. 2006; Kennedy et al. 2007; Byko et al. 2009) reaches the surface at its former location. It is possible that the intensification of this low-level circulation resulted in a downward-directed vertical perturbation pressure gradient force, allowing a downdraft or local weakening of updraft to transport hydrometeors from the overhang to the surface, similar to a type-III DRC of Byko et al. (2009). Another DRC is observed by 0255:43 UTC, evident in Fig. 8 as an expansion of the area where Doppler velocities are plotted. This DRC is associated with an enhancement of inbound velocities, possibly transported from aloft, that enhance cyclonic shear associated with circulation e. By 0301:47 UTC, the second circulation has dissipated and the third circulation has occluded and moved rearward in a storm-relative sense.
b. Hook echo evolution
The evolution of the hook echo signature generally follows the model presented in Beck et al. (2006). The transition between stages 1 and 2 of the Beck et al. model is readily evident in the data from this case, even with the comparatively coarse resolution and large distance from the radar. Though the rearward shift of the old mesocyclone and development of new mesocyclone along the rear-flank gust front happened fairly rapidly, the third stage of the Beck et al. model was not as clearly observed. Recall that in that stage, the new mesocyclone begins advecting precipitation cyclonically as the old mesocyclone moves to the storm’s rear. The hook echo was observed to elongate after the old mesocyclone occluded and moved rearward, but the development of significant cyclonic curvature was prolonged. This could be due to a weaker low-level circulation in the 1 June case, or to decreased spatial resolution compared to the Beck et al. study. Also, there are other important factors governing hook echo evolution not accounted for in the Beck et al. model, such as hydrometeor fall speeds, vertical advection, and changes aloft.
An example of the measured polarimetric variables at low levels during the transition from stages 2–3 of the Beck et al. model is shown in Fig. 9. The locations of the old mesocyclone (solid circle) and developing mesocyclone (dashed circle) were determined from cyclonic shear in the Doppler velocity field. The characteristic “kink” in the hook echo is present, evidence of the old circulation occluding and moving toward the rear of the storm. Note that the radial streak of very low ρHV values extending from the rear of the storm is due to NBF effects, which combined with possible undercorrected differential attenuation, causes the similar feature in ZDR.
In the field of ZDR, the hook echo is characterized by large spatial variability and rapid changes in time. Most consistently there was a narrow strip of higher ZDR values along the inside (inflow) edge of the hook, sometimes connecting with the ZDR arc of the forward flank echo. During the RFD surge and occlusion of the old mesocyclone, the values of ZDR sometimes increased on the inside edge of the hook and decreased on the back side. The top of the hook echo, where the appendage connects to the main body of the storm, is marked by consistently higher values of ZDR, likely a result of size sorting due to its proximity to the updraft (i.e., it is the base of the ZDR column).
In general, ρHV in the hook is quite high, oftentimes higher than in any other part of the storm at low levels. This is in stark contrast to the RFD gust front (not shown), observed as a thin line in ZH characterized by very low (<0.5) ρHV, indicative of nonmeteorological scatterers such as insects or light debris. It is difficult to ascertain any significant evolution of ρHV in the hook echo, though on several occasions during the occlusion the values increased slightly. If only raindrops were present, this would be the result of a change in the drop size distribution, including (but not limited to) an increase in the number of small drops. The KDP in the hook echo tends to be fairly low, though the estimates are noisy because of the thin shape of the hook echo and its relatively low rainfall rates.
c. Low-level hail signature evolution
Numerous studies have discussed a S-band polarimetric signature observed at low levels indicative of large hail and manifested as high values of ZH collocated with near-zero value of ZDR (e.g., Wakimoto and Bringi 1988; Ryzhkov et al. 2005b; Heinselman and Ryzhkov 2006; KR08). This signature is routinely observed in supercell storms to the (storm relative) left or left rear of the mesocyclone. The low-level hail signature is present through the vast majority (>90%) of volume scans in the 1 June case. This seems to be characteristic of nontornadic supercells, which tend to display a more persistent hail signature than tornadic supercells (KR08). The occasional observation of an expansion of the hail signature may be related to changes in updraft strength, either allowing more of the lofted precipitation to fall in the case of updraft weakening, or the increase of precipitation production (including larger hail) with a strengthening updraft. Such large expansions of the hail signature often coincide with an increase in ZH in the core and in the RFD, also indicative of changing updraft strength. However, overall the evolution of the hail signature was not a reliable indicator of the occlusion-cycling process in this case.
d. ZDR arc evolution
The ZDR arc is an arc-shaped region of very high (>4 dB) ZDR along the inflow edge of the forward-flank downdraft (FFD) echo in supercells found at low levels. KR08 and Kumjian and Ryzhkov (2009, hereafter KR09) attribute the signature to size sorting due to strong veering storm-relative winds in supercell environments. The ZDR arc underwent surprisingly rapid changes in the 1 June 2008 supercell, though a repetitive pattern exists. An example of the cycle is displayed in Figs. 10 –11. The arc ZDR values increase substantially between 0252 and 0254 UTC, extending back to the inflow notch by 0255 UTC, whereupon ZDR values decrease in the panels from 0256 and 0258 UTC. The ZDR values increase again from 0259 to 0302 UTC as the arc extends back toward the inflow notch, and reach a maximum at about 0305 UTC.
This pattern repeated itself numerous times throughout the lifetime of the storm and seems to be tied to the occlusion cycle. Leading up to the occlusion of the low-level mesocyclone, the arc extends back toward the inflow and increases in magnitude and areal extent. The arc tends to reach its maximum strength and extent before the weakening of the updraft. As the circulation occludes, the arc contracts or dissipates; in other words, the region of very high ZDR rapidly shrinks. At this time, a new region of very high ZDR may form at the edge of the FFD echo. This new region becomes the new ZDR arc. The contraction of the old arc and sudden appearance of the new one is consistent with a weakening updraft. As the updraft and low-level inflow weakens, the size-sorting mechanism is inhibited, thus the magnitude of ZDR is not as large. Also, precipitation from the overhang that was previously lofted now falls into the inflow, emerging as a weak ZH echo just ahead of the edge of the FFD, characterized by very high ZDR. A new cycle commences with this new ZDR arc and repeats. The period of time over which this cycle proceeds depends on the cycling frequency of the storm, which is highly variable. The ZDR arc cycle ranged from about 7 to 20 min, and on occasion was interrupted (i.e., another new ZDR arc formed at the edge before the old arc finished its contraction and dissipation).
The source of large drops in the ZDR arc is melting graupel and hail. Aloft, there is sometimes a secondary band of graupel over the FFD region. Overlaying the polarimetric features from low, middle, and upper levels, a spatial offset is evident between the midlevel graupel and the low-level ZDR arc (refer to Fig. 7). This slope to the precipitation echo has been noted for many years (Browning 1964) and is indicative of precipitation being advected downwind (in a storm-relative sense) into the FFD. Based on the interpretation of KR08 and KR09, the largest drops are not advected as far by the storm-modified low-level winds and are therefore found at the edge of the FFD echo, manifested in the polarimetric observations by very high ZDR along a gradient in ZH.
e. KDP and ZDR column evolution
Vertically extensive “columns” of enhanced KDP and ZDR have been noted in many studies (e.g., Illingworth et al. 1987; Caylor and Illingworth 1987; Tuttle et al. 1989; Meischner et al. 1991; Conway and Zrnić 1993; Brandes et al. 1995; Hubbert et al. 1998; Kennedy et al. 2001; Loney et al. 2002; Scharfenberg 2003; Ryzhkov et al. 2005b; KR08; Romine et al. 2008). Though the columns are present in all types of convective storms, the spatial offset between ZDR and KDP columns in supercells is a seemingly distinct characteristic of such storms (Ryzhkov et al. 2005b; KR08; Romine et al. 2008). Generally, the KDP column is offset to the west (or northwest) of the ZDR column and is located on the western periphery of the updraft. The ZDR column is associated with updraft through size sorting and lofting of wet hydrometeors above the freezing level, whereas the KDP column is associated with the precipitation fallout of hydrometeors lofted by the updraft and a downdraft beneath it at low levels (in the case of supercells, the rear-flank downdraft). These signatures are present throughout the lifetime of the storm.
From the observed sounding, the environmental freezing level is at 4.3 km AGL. Assuming a surface-based parcel, the updraft-perturbed freezing level is at roughly 5.2 km AGL. The ZDR and KDP columns are prominently observed at 2.5° elevation PPIs (the radar beam height is about 6.0 km AGL), but they are not present in the 3.6° elevation PPI (beam height of about 8.3 km). So, the columns extend at least 2.0 km above the environmental freezing level. Unfortunately, the vertical resolution of the volume scans is insufficient to make any meaningful analysis of the evolution of the column heights. This is due partially to the storm’s substantial distance from the radar. Thus, the questions of whether there is any relation between the updraft strength and the height of the ZDR column and how the vertical structure of the ZDR column evolves in time remain unanswered. Future work may involve scanning strategies such as vertical cross sections (i.e., “genuine” RHIs) that focus on the development and evolution of the ZDR column. Such enhanced scanning strategies can be easily implemented with a polarimetric MPAR.
However, the horizontal cross sections of the columns at different levels have temporal and spatial resolution adequate enough to allow an analysis of their relation and interaction throughout the occlusion cycle. Throughout the evolution of the 1 June storm, the spatial offset is persistent. Figure 12 displays the evolution of horizontal cross sections of the ZDR and KDP columns from 0257 to 0315 UTC. At 0259 UTC, the KDP column is expanding, overtaking the ZDR column, which begins to decrease in areal size noticeably from 0300 to 0303 UTC. By 0304–0305 UTC, the KDP column dominates at midlevels, though there is some indication of regeneration of a ZDR column beginning on the southwestern flank of the storm. This corresponds to the stage II of the Beck cycle at low levels. Though the ZDR column has a small horizontal extent, the values of ZDR are extremely high.
After 0305 UTC, the updraft reintensifies, as suggested by ZH beginning to decrease coincident with an expanding ZDR column (though its magnitude decreases). As the updraft intensifies, the ZDR column (and ring) expands in areal extent as the KDP column becomes smaller. Simultaneously, a region of graupel on the southern periphery of the storm develops and expands, with the graupel evidently originating from flanking line turrets southwest of the updraft. Some of this graupel is carried by the flow around the periphery of the storm and is ingested into the updraft. This graupel ingestion and advection by the mesocyclone contributes to the ZDR column wrapping around the KDP column in a ringlike shape. Sections 4g and 4h further explore the links between these features.
Evidently, once the graupel source is cut off the ZDR column region once again begins to shrink as the KDP region begins to expand, resulting in more significant overlap between the two columns, beginning the cycle over again. Note that such an overlap indicates precipitation fallout at the edge of the updraft, since regions of high KDP (and ZH) generally imply positive ZDR. In contrast, stronger updrafts are marked with high ZDR, low KDP, and relatively low ZH. The deterioration of the ZDR column in regions of overlap with the KDP column is likely because the precipitation fallout (high ZH, KDP) implies weaker updraft that is no longer able to loft liquid drops and sort particles, as well as the presence of larger hail that would increase ZH, decrease ZDR, and have a minimal effect on KDP. Again, further dissipation of the ZDR column occurs coincident with a substantial expansion of the KDP column until regeneration of the ZDR column begins, often on the southern or southwestern flank of the storm. Typically, the complete cycle took 10–20 min in the 1 June case.
f. Updraft signature evolution
The updraft signature noted by Ryzhkov et al. (2005b) and KR08 is a deep minimum of ρHV at mid- and upper levels of the storm located within the updraft, typically extending several km above the level of the ZDR and ρHV rings. This signature is likely analogous to the “LDR cap” (e.g., Hubbert et al. 1998) observed atop the summit of the ZDR column using dual-polarization radars operating using alternating transmission of horizontally and vertically polarized waves. The signature (and collocated negative ZDR values) may be due to large, dry oblate hail (>5 cm) in the updraft (e.g., Balakrishnan and Zrnić 1990; Aydin and Zhao 1990).
Observed throughout the lifetime of the 1 June storm (Fig. 13), the values of ρHV reached as low as 0.7–0.8, and the signature was often accompanied by a polarimetric three-body scatter signature (TBSS; Zrnić 1987). The updraft signature is evident beyond the level where ZDR and KDP become zero, indicating a general lack of supercooled raindrops, though not ruling out the possibility of supercooled liquid water cloud droplets. In this sense, the updraft signature conveniently serves as a proxy for the updraft in the absence of dual-Doppler data and other oft-employed indicators (e.g., ZDR columns and bounded weak echo regions). KR08 suggested that the height of the minimum in ρHV could serve as an indicator of updraft strength. Unfortunately, the height of the ρHV minima cannot be determined accurately using this dataset owing to limited vertical resolution. Similar to the height of the ZDR column, the trend of the height of the ρHV minimum should be investigated in a study that utilizes scanning strategies with fine vertical and temporal resolution.
g. Midlevel ZDR and ρHV ring evolution
Midlevel ZDR and ρHV rings were first documented in KR08. Located near the updraft-perturbed freezing level, these striking circular or semicircular signatures are clear indicators of hydrometeor trajectories affected by the midlevel vorticity associated with the mesocyclone. The rings are characterized by enhanced positive ZDR and reduced ρHV surrounded by and encircling mainly near-zero values of ZDR and higher ρHV. The appearance of the signatures has been attributed to wet ice hydrometeors being advected cyclonically by the mesocyclone (KR08).
The midlevel rings display a cyclic behavior throughout the lifetime of the supercell case in this study. A portion of such a ZDR ring cycle is provided in Fig. 14. Using a dual-Doppler analysis, Payne et al. (2010) recently noted that a full ring occurred during the mature phase of an occlusion cycle, while a significant tornado was occurring, and that the ring dissipated once the mesocyclone was occluded. Though only single-Doppler data are available for the present case, the evolution of the midlevel rings in relation to the inferred occlusion cycles are consistent with the findings of Payne et al. (2010) and is described in detail next.
The following description of the midlevel ZDR ring evolution begins at the mature stage. Note that the ZDR and ρHV rings follow similar evolution because of their common origin, so for brevity only the ZDR ring is described. Payne et al. (2010) found that the cyclonic vorticity maximum was located at the center of the ring. Thus, the full ring is likely present during the divided mesocyclone stage of supercell evolution (Lemon and Doswell 1979), where the right-front (storm relative) half of the ring is associated with updraft and the left half of the ring is associated with heavy precipitation. This is evident upon examining the observed midlevel KDP; the KDP column is directly tied to the RFD (Ryzhkov et al. 2005b), and the collocation of large values of KDP, ZH, and positive ZDR are typical of regions with high concentrations of hydrometeors and large liquid water contents (e.g., Scharfenberg 2003; Scharfenberg et al. 2005). This is distinct from the right half of the ring, where the region of positive ZDR is generally collocated with a relative minimum in ZH and near-zero KDP. This implies a deficit of small drops, indicative of size sorting within the updraft.
The ring then gradually loses its shape, becoming less structured in appearance as it erodes (i.e., regions of positive ZDR disappear) from the west and northwest, and on some occasions the south. This is evident in Fig. 14 between 0248 and 0250 UTC, which coincides with the occlusion of the low-level mesocyclone and its subsequent rearward movement. At this time, the eastern portion of the ring (the ZDR column) may decrease in magnitude and/or areal extent. During the dissipation of the midlevel ring, a TBSS appears on the north or northwest side of the mesocyclone, serving as an obvious indicator of large hail aloft. Often more prominent is the polarimetric TBSS (e.g., Hubbert and Bringi 2000): a region of extremely high ZDR (7–10 dB) and very low ρHV (<0.5) extending outward radially from the rear of the hail core. These trends suggest a weakening of the updraft as the mesocyclone occludes. The loss of positive ZDR regions on the western side of the ring indicates that ice phase hydrometeors such as hail or graupel now dominate the resolution volume, as discussed previously. A decrease in areal extent of the ZDR column implies a smaller region of updraft where vertical velocities are sufficient to vigorously sort out and loft the smaller particles, or that the supply of these particles is somehow reduced.
Next, regeneration of the ZDR ring occurs on the south or southwest side of the storm, likely associated with new updrafts. At this time, the polarimetric TBSS decreases in prominence or disappears altogether. The positive ZDR values may increase in magnitude and in areal extent. It is unclear whether the increase in magnitude of ZDR is due to an increase in updraft strength; in fact, in some instances the increase in ZDR was associated with more precipitation falling into the updraft (i.e., an increase of ZH in the weak echo region). Regeneration of the ring on both sides continues until a full ZDR ring redevelops. In later cycles, regeneration of the ZDR ring follows a similar pattern (Fig. 15) as low-level convergence increases, coincident with a circulation intensifying as it moves rearward and occludes.
h. The graupel belt and its evolution
The evolution of the midlevel ZDR ring is evidently closely tied to a feature called the “graupel belt” herein. The graupel belt is characterized by very high ρHV values (>0.98) and ZDR near zero, indicative of ice phase hydrometeors. The region is found on the southwest, south, and southeast flanks of the supercell (Fig. 16) and is likely analogous to the “embryo curtain” of Browning and Foote (1976) or the “embryo corridor” suggested by Ziegler et al. (1983) and Nelson (1983). Indeed, it appears as if the graupel belt provides embryos for rapid hail growth in the moisture-rich updraft. The ingestion of graupel particles into the updraft is visible as an abrupt increase in ZDR and a sharp decrease in ρHV, producing the midlevel rings. This sudden transition in the polarimetric variables suggests that the dry graupel particles begin to acquire a significant water fraction, which causes the dielectric constant to drastically increase, leading to a signature analogous to that of the melting layer. This acquisition of water could be due to melting as the graupel encounters warmer air within the updraft (KR08), and/or due to the onset of wet growth. With respect to the latter process, it is speculated that the ZDR ring could be interpreted as an indicator of the path wet-growth hailstones take as they pass through the edge of the updraft, advected by the mesocyclone. In both cases the polarimetric signatures would be very similar and probably impossible to distinguish. Interestingly, hailstone trajectories calculated in previous studies (e.g., Xu 1983; Nelson 1983; Foote 1984) show the preferred path of large hailstone growth along the right flank of the updraft, coinciding with the observed location of the ZDR ring in supercells.
The rapid scanning implemented for this case allows for the identification of the sources of the graupel in the graupel belt. Previous studies have suggested that embryos come from flanking line turrets (e.g., Heymsfield and Musil 1982; Heymsfield 1982), or from particles recycled from the upwind overhang (e.g., Browning and Foote 1976; Knupp and Cotton 1982; Conway and Zrnić 1993). In fact, both of these sources are observed in the present case. Flanking line convection is observed to produce “puffs” of graupel on the southwestern flank of the storm. Analysis of the three-dimensional structure of the storm indicates that the echoes originate from below the level of the graupel belt, and enhanced ZDR beneath the melting layer collocated with low-to-moderate ZH is indicative of size sorting by the updrafts in these flanking line cells. The graupel produced in these convective cells is advected with the flow around the right flank of the storm and into the inflow/updraft region. Note that at certain times during the storm lifetime, this graupel source is cut off and the graupel belt (and thus ZDR ring) erodes from the southwest side. This is apparently related to the behavior of the RFD gust front, along which there is convergence and the associated flanking line cells.
Later in the storm lifetime, graupel falls into the graupel belt region from above during an updraft “collapse” (Lemon and Doswell 1979) or weakening stage. As the bounded weak echo region fills in, ZH increases in the inflow region from above. Successive scans show the increase in ZH at lower elevations, similar to a DRC. Graupel originating from aloft tends to be confined more to the southeast and eastern flanks of the updraft in the 1 June supercell. It is not clear whether the difference in location is characteristic of all supercells or simply fortuitous for this event. Note that this graupel pathway could be active at times other than the weakening stage; an increase in ZH in the inflow region may simply mean an increase in this graupel pathway.
5. Summary
The evolution of supercell polarimetric signatures as related to the cyclic occlusion process was presented. The 1 June 2008 cyclic nontornadic supercell was sampled using a polarimetric WSR-88D radar with which data with enhanced spatial and temporal resolution were collected, allowing for such an investigation of rapidly evolving features. Also, the use of a longer radar wavelength reduces problems with attenuation and differential attenuation that inhibit a full view of the storm in other studies that use radars with shorter wavelengths.
The polarimetric signatures investigated in this study included the hook echo, ZDR arc, hail signature, ZDR and KDP columns, updraft signature, midlevel ZDR and ρHV rings, and the graupel belt. These supercell signatures have repetitive evolutionary patterns that correspond to the storm’s occlusion cycle. Each signature’s complete cycle was described in detail, and a microphysical interpretation of these cycles was presented. Future studies should investigate the generality of these evolutionary patterns. If found to be typical of supercell storms, these patterns could augment traditional data in beneficial ways for operational meteorologists to identify the onset of occlusion cycles, the frequency of which may be an important factor of the storm’s potential to produce tornadoes (Beck et al. 2006).
In addition to the previously documented supercell signatures, this study identifies the graupel belt, a region of graupel (indicated by very high ρHV and low ZDR above the melting layer) that wraps around the updraft on its southern and southeastern flanks, analogous to the embryo corridor identified in previous hail growth and modeling studies. The graupel belt is intimately connected with the evolution of the midlevel ZDR and ρHV rings. The polarimetric characteristics of the graupel belt and midlevel rings provide evidence for the ingestion of graupel into the updraft (as hypothesized by KR08), whereupon melting and/or wet growth occurs. Based on previous trajectory analyses and hail growth studies, these features may indicate the ongoing development of large hail. The possible sources of the graupel embryos are identified as 1) flanking line turrets to the southwest of the updraft, as well as 2) graupel falling from the anvil overhang and backsheared anvil, especially upon weakening of the updraft. Both of these source regions have been suggested in previous studies; there is clear evidence for both occurring in the 1 June case.
In general, the radar-observed features of the storm evolved rapidly, even on scales faster than what was observed in this study. Polarimetric variables in particular are especially sensitive to these rapid changes. This is partly because of the dynamic range of the variables. For example, a 3-dBZ change in ZH (which varies from roughly 0–60 dBZ) corresponds to about 5% of the range, whereas in ZDR (which varies roughly 0–4 dB at S band), a 2-dB (50%) change is much more noticeable. Another contributing factor is that polarimetric variables are more sensitive to changes in the particle size distribution than ZH, especially the concentration of larger raindrops and resonance scatterers. Because of this, the more noticeable changes in polarimetric variables have a more meaningful physical interpretation than those of ZH. These drastic changes in the conventional radar and polarimetric signatures between volume scans indicate that updates faster than 70 s are desired. The supersector scanning strategy maximizes the mechanical accelerations of the antenna dish with the WSR-88D. Thus, to decrease the scan time in the future, there are two options: reduce the width of the sector, or reduce the number of samples collected at each radial. The latter will increase the noisiness of the measurements. For isolated storms, using a 40° sector may be sufficient depending on range, and will approximately halve the update time. Nonetheless, the results of this study suggest that certain bulk microphysical properties of supercells may change on scales faster than what was sampled, which underscores the utility of rapid-scanning polarimetric radars such as the future polarimetric MPAR. This conclusion is consistent with previous work that has shown rapid changes in conventional radar and kinematic properties of storms on small time scales.
A limitation of this study was relatively poor vertical resolution, which is another opportunity for improvement with a future polarimetric MPAR, which is capable of performing genuine RHIs. Fortunately, with the current trend in radar meteorology for higher spatial and temporal resolution data as well as polarization diversity, more rapid-scan polarimetric observations of cyclic supercells are likely. These future studies should investigate the generality of the documented patterns of evolution of the polarimetric signatures as related to cyclic mesocyclogenesis. The apparent link between microphysics and supercell kinematics found in this investigation should be further explored to elucidate the role microphysics plays in the behavior and evolution of such storms.
Acknowledgments
The authors thank the NSSL/CIMMS employees who maintain and operate the KOUN radar for research-grade operations and who worked to allow such versatile control of the radar for this study. This work is partially funded by NSF Grant ATM-0532107. Additional funding was provided by NOAA/Office of Oceanic and Atmospheric Research under NOAA/University of Oklahoma Cooperative Agreement NA17RJ1227, U.S. Department of Commerce. Paul Markowski (PSU), David Dowell (NCAR), and an additional anonymous reviewer provided constructive comments and suggestions that significantly improved the manuscript. Additionally, the authors thank Michael French (OU) and Dušan Zrnić (NSSL) for reviewing earlier versions of the manuscript, as well as Alexander Schenkman and Joseph Picca (OU) for useful discussions.
REFERENCES
Adlerman, E. J., and K. K. Droegemeier, 2002: The sensitivity of numerically simulated cyclic mesocyclogenesis to variations in physical and computational parameters. Mon. Wea. Rev., 130 , 2671–2691.
Adlerman, E. J., and K. K. Droegemeier, 2005: The dependence of numerically simulated cyclic mesocyclogenesis upon environmental vertical wind shear. Mon. Wea. Rev., 133 , 3595–3623.
Adlerman, E. J., K. K. Droegemeier, and R. P. Davies-Jones, 1999: A numerical simulation of cyclic mesocyclogenesis. J. Atmos. Sci., 56 , 2045–2069.
Aydin, K., and Y. Zhao, 1990: A computational study of polarimetric radar observables in hail. IEEE Trans. Geosci. Remote Sens., 28 , 412–422.
Balakrishnan, N., and D. S. Zrnić, 1990: Use of polarization to characterize precipitation and discriminate large hail. J. Atmos. Sci., 47 , 1525–1540.
Beck, J. R., J. L. Schroeder, and J. M. Wurman, 2006: High-resolution dual-Doppler analyses of the 29 May 2001 Kress, Texas, cyclic supercell. Mon. Wea. Rev., 134 , 3125–2148.
Biggerstaff, M. I., and Coauthors, 2005: The Shared Mobile Atmospheric Research and Teaching Radar: A collaboration to enhance research and teaching. Bull. Amer. Meteor. Soc., 86 , 1263–1274.
Bluestein, H. B., and A. L. Pazmany, 2000: Observations of tornadoes and other convective phenomena with a mobile, 3-mm wavelength Doppler radar: The spring 1999 field experiment. Bull. Amer. Meteor. Soc., 81 , 2939–2951.
Bluestein, H. B., A. L. Pazmany, J. C. Galloway, and R. E. Mcintosh, 1995: Studies of the substructure of severe convective storms using a mobile 3-mm-wavelength Doppler radar. Bull. Amer. Meteor. Soc., 76 , 2155–2169.
Bluestein, H. B., C. C. Weiss, M. M. French, E. M. Holthaus, R. L. Tanamachi, S. Frasier, and A. L. Pazmany, 2007a: The structure of tornadoes near Attica, Kansas, on 12 May 2004: High-resolution, mobile, Doppler radar observations. Mon. Wea. Rev., 135 , 475–506.
Bluestein, H. B., M. M. French, R. L. Tanamachi, S. Frasier, K. Hardwick, F. Juyent, and A. L. Pazmany, 2007b: Close-range observations of tornadoes in supercells made with a dual-polarization, X-band, mobile, Doppler radar. Mon. Wea. Rev., 135 , 1522–1543.
Bluestein, H. B., R. Tanamachi, J. Houser, I. Popstefanija, B. Seeger, R. Bluth, and J. B. Knorr, 2008a: Use of a mobile, phased-array, X-band Doppler radar to study severe convective storms and tornadoes. Extended Abstracts, Fifth European Conf. on Radar in Meteorology and Hydrology, Helsinki, Finland, Finnish Meteorological Institute, 4.1.
Bluestein, H. B., and Coauthors, 2008b: Severe-storm data collected in the Southern Plains by three mobile Doppler radars during the spring, 2007 and 2008. Extended Abstracts, 24th Conf. on Severe Local Storms, Savannah, GA, Amer. Meteor. Soc., 5.3.
Brandes, E. A., 1977: Flow in severe thunderstorms observed by dual-Doppler radar. Mon. Wea. Rev., 105 , 113–120.
Brandes, E. A., J. Vivekanandan, J. D. Tuttle, and C. J. Kessinger, 1995: A study of thunderstorm microphysics with multiparameter radar and aircraft observations. Mon. Wea. Rev., 123 , 3129–3143.
Bringi, V. N., and V. Chandrasekar, 2001: Polarimetric Doppler Weather Radar: Principles and Applications. Cambridge University Press, 636 pp.
Brock, F. V., K. C. Crawford, R. L. Elliott, G. W. Cuperus, S. J. Stadler, H. L. Johnson, and M. D. Eilts, 1995: The Oklahoma Mesonet: A technical overview. J. Atmos. Oceanic Technol., 12 , 5–19.
Brown, R. A., V. T. Wood, and D. Sirmans, 2002: Improved tornado detection using simulated and actual WSR-88D data with enhanced resolution. J. Atmos. Oceanic Technol., 19 , 1759–1771.
Brown, R. A., R. M. Steadham, B. A. Flickinger, R. R. Lee, D. Sirmans, and V. T. Wood, 2005a: New WSR-88D volume coverage pattern 12: Results of field tests. Wea. Forecasting, 20 , 385–393.
Brown, R. A., B. A. Flickinger, E. Forren, D. M. Schultz, D. Sirmans, P. L. Spencer, V. T. Wood, and C. L. Ziegler, 2005b: Improved detection of severe storms using experimental fine-resolution WSR-88D measurements. Wea. Forecasting, 20 , 3–14.
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., and G. B. Foote, 1976: Airflow and hail growth in supercell storms and some implications for hail suppression. Quart. J. Roy. Meteor. Soc., 102 , 499–533.
Burgess, D. W., V. T. Wood, and R. A. Brown, 1982: Mesocyclone evolution statistics. Preprints, 12th Conf. on Severe Local Storms, San Antonio, TX, Amer. Meteor. Soc., 422–424.
Byko, Z., P. M. Markowski, Y. Richardson, J. Wurman, and E. Adlerman, 2009: Descending reflectivity cores in supercell thunderstorms observed by mobile radars and in a high- resolution numerical simulation. Wea. Forecasting, 24 , 155–186.
Caylor, I. C., and A. J. Illingworth, 1987: Radar observations and modeling of warm rain initiation. Quart. J. Roy. Meteor. Soc., 113 , 1171–1191.
Colquhoun, J. R., and P. A. Riley, 1996: Relationships between tornado intensity and various wind and thermodynamic variables. Wea. Forecasting, 11 , 360–371.
Conway, J. W., and D. S. Zrnić, 1993: A study of production and hail growth using dual-Doppler and multiparameter radars. Mon. Wea. Rev., 121 , 2511–2528.
Darkow, G. L., 1971: Periodic tornado production by long-lived parent thunderstorms. Preprints, Seventh Conf. on Severe Local Storms, Kansas City, MO, Amer. Meteor. Soc., 214–217.
Darkow, G. L., and J. C. Roos, 1970: Multiple tornado producing thunderstorms and their apparent cyclic variations in intensity. Preprints, 14th Conf. on Radar Meteorology, Tucson, AZ, Amer. Meteor. Soc., 305–308.
Davies, J. M., and R. H. Johns, 1993: Some wind and instability parameters associated with strong and violent tornadoes. Part I: Wind shear and helicity. The Tornado: Its Structure, Dynamics, Prediction, and Hazards, Geophys. Monogr., Vol. 79, Amer. Geophys. Union, 573–582.
Davies-Jones, R., D. W. Burgess, and M. Foster, 1990: Test of helicity as a forecast parameter. Preprints, 16th Conf. on Severe Local Storms, Kananaskis Park, Alberta, Canada, Amer. Meteor. Soc., 588–592.
Doviak, R. J., and D. S. Zrnić, 1993: Doppler Radar and Weather Observations. Academic Press, 562 pp.
Dowell, D. C., and H. B. Bluestein, 2002a: The 8 June 1995 McLean, Texas, storm. Part I: Observations of cyclic tornadogenesis. Mon. Wea. Rev., 130 , 2626–2648.
Dowell, D. C., and H. B. Bluestein, 2002b: The 8 June 1995 McLean, Texas, storm. Part II: Cyclic tornado formation, maintenance, and dissipation. Mon. Wea. Rev., 130 , 2649–2670.
Esterheld, J. M., and D. J. Giuliano, 2008: Discriminating between tornadic and nontornadic supercells: A new hodograph technique. Electron. J. Severe Storms Meteor., 3 , 1–50.
Foote, G. B., 1984: A study of hail growth utilizing observed storm conditions. J. Climate Appl. Meteor., 23 , 84–101.
French, M. M., H. B. Bluestein, D. C. Dowell, L. J. Wicker, M. R. Kramar, and A. L. Pazmany, 2008: High-resolution, mobile Doppler radar observations of cyclic mesocyclogenesis in a supercell. Mon. Wea. Rev., 136 , 4997–5016.
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.
Heinselman, P. L., and A. V. Ryzhkov, 2006: Validation of polarimetric hail detection. Wea. Forecasting, 21 , 839–850.
Heinselman, P. L., D. L. Priegnitz, K. L. Manross, T. M. Smith, and R. W. Adams, 2008: Rapid sampling of severe storms by the National Weather Radar Testbed phased array radar. Wea. Forecasting, 23 , 808–824.
Herzegh, P. H., and A. R. Jameson, 1992: Observing precipitation through dual-polarization radar measurements. Bull. Amer. Meteor. Soc., 73 , 1365–1374.
Heymsfield, A. J., 1982: A comparative study of the rates of development of potential graupel and hail embryos in High Plains storms. J. Atmos. Sci., 39 , 2867–2897.
Heymsfield, A. J., and D. J. Musil, 1982: Case study of a hailstorm in Colorado. Part II: Particle growth processes at midlevels deduced from in situ measurements. J. Atmos. Sci., 39 , 2847–2866.
Hubbert, J., and V. N. Bringi, 2000: The effects of three-body scattering on differential reflectivity signatures. J. Atmos. Oceanic Technol., 17 , 51–61.
Hubbert, J., V. N. Bringi, and L. D. Carey, 1998: CSU-CHILL polarimetric measurements from a severe hailstorm in eastern Colorado. J. Appl. Meteor., 37 , 749–755.
Illingworth, A. J., J. W. F. Goddard, and S. M. Cherry, 1987: Polarization radar studies of precipitation development in convective storms. Quart. J. Roy. Meteor. Soc., 113 , 469–489.
Jameson, A. R., 1985: Microphysical interpretation of multiparameter radar measurements in rain. Part III: Interpretation and measurement of differential propagation phase shift between orthogonal linear polarizations. J. Atmos. Sci., 42 , 607–614.
Johnson, K. W., P. S. Ray, B. C. Johnson, and R. P. Davies-Jones, 1987: Observations related to the rotational dynamics of the 20 May 1977 tornadic storms. Mon. Wea. Rev., 115 , 2463–2478.
Kennedy, A., J. M. Straka, and E. N. Rasmussen, 2007: A statistical study of the association of DRCs with supercells and tornadoes. Wea. Forecasting, 22 , 1191–1199.
Kennedy, P. C., S. A. Rutledge, and W. A. Petersen, 2001: Polarimetric radar observations of hail formation. J. Appl. Meteor., 40 , 1347–1366.
Kerr, B. W., and G. L. Darkow, 1996: Storm-relative winds and helicity in the tornadic thunderstorm environment. Wea. Forecasting, 11 , 489–505.
Knupp, K. R., and W. R. Cotton, 1982: An intense, quasi-steady thunderstorm over mountain terrain. Part II: Doppler radar observations of the storm morphological structure. J. Atmos. Sci., 39 , 343–358.
Kumjian, M. R., and A. V. Ryzhkov, 2008: Polarimetric signatures in supercell thunderstorms. J. Appl. Meteor. Climatol., 47 , 1940–1961.
Kumjian, M. R., and A. D. Schenkman, 2008: Interpretation of the “flying eagle” radar signature in supercells. Extended Abstracts, 24th Conf. on Severe Local Storms, Savannah, GA, Amer. Meteor. Soc., P14.2. [Available online at http://ams.confex.com/ams/pdfpapers/141917.pdf].
Kumjian, M. R., and A. V. Ryzhkov, 2009: Storm-relative helicity revealed from polarimetric radar measurements. J. Atmos. Sci., 66 , 667–685.
Leftwich, P. W., 1990: On the use of helicity in operational assessment of severe local storm potential. Preprints, 16th Conf. on Severe Local Storms, Kananaskis Park, Alberta, Canada, Amer. Meteor. Soc., 269–274.
Lemon, L. R., and C. A. Doswell III, 1979: Severe thunderstorm evolution and mesocyclone structure as related to tornadogenesis. Mon. Wea. Rev., 107 , 1184–1197.
Loney, M. L., D. S. Zrnicì, J. M. Straka, and A. V. Ryzhkov, 2002: Enhanced polarimetric radar signatures above the melting level in a supercell storm. J. Appl. Meteor., 41 , 1179–1194.
Markowski, P. M., J. M. Straka, and E. N. Rasmussen, 2002: Direct surface thermodynamic observations within the rear-flank downdrafts 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.
Meischner, P. F., V. N. Bringi, D. Heimann, and H. Höller, 1991: A squall line in Southern Germany: Kinematics and precipitation formation as deduced by advanced polarimetric and Doppler radar measurements. Mon. Wea. Rev., 119 , 678–701.
Melnikov, V., D. Zrnic, A. Ryzhkov, A. Zahrai, and J. Carter, 2009: Validation of attenuation orrection at X band performed with collocated S-band polarimetric radar. Extended Abstracts, 34th Conf. on Radar Meteorology, Williamsburg, VA, Amer. Meteor. Soc., 11A.5.
NCDC, cited. 2008: Storm events database. [Available online at http://www4.ncdc.noaa.gov/cgi-win/wwcgi.dll?wwEvent~Storms].
Nelson, S. P., 1983: The influence of storm flow structure on hail growth. J. Atmos. Sci., 40 , 1965–1983.
Payne, C. D., T. J. Schuur, D. R. MacGorman, M. I. Biggerstaff, K. Kuhlman, and W. D. Rust, 2010: Polarimetric and electrical characteristics of a lightning ring in a supercell storm. Mon. Wea. Rev., 138 , 2405–2425.
Pruppacher, H. R., and K. V. Beard, 1970: A wind tunnel investigation of the internal circulation and shape of water drops falling at terminal velocity in air. Quart. J. Roy. Meteor. Soc., 96 , 247–256.
Rasmussen, E. N., and D. O. Blanchard, 1998: A baseline climatology of sounding-derived supercell and tornado forecast parameters. Wea. Forecasting, 13 , 1146–1164.
Rasmussen, E. N., J. M. Straka, M. S. Gilmore, and R. P. Davies-Jones, 2006: A preliminary survey of rear-flank descending reflectivity cores in supercell storms. Wea. Forecasting, 21 , 923–928.
Rison, W., R. J. Thomas, P. R. Krehbiel, T. Hamlin, and J. Harlin, 1999: A GPS-based three-dimensional lightning mapping array: Initial observations in New Mexico. Geophys. Res. Lett., 26 , 3573–3576.
Romine, G. S., D. W. Burgess, and R. B. Wilhelmson, 2008: A dual-polarization radar based assessment of the 8 May 2003 Oklahoma City area tornadic supercell. Mon. Wea. Rev., 136 , 2849–2870.
Ryzhkov, A. V., 2007: The impact of beam broadening on the quality of radar polarimetric data. J. Atmos. Oceanic Technol., 24 , 729–744.
Ryzhkov, A. V., T. J. Schuur, D. W. Burgess, P. L. Heinselman, S. E. Giangrande, and D. S. Zrnić, 2005a: The Joint Polarization Experiment: Polarimetric rainfall measurements and hydrometeor classification. Bull. Amer. Meteor. Soc., 86 , 809–824.
Ryzhkov, A. V., T. J. Schuur, D. W. Burgess, and D. S. Zrnić, 2005b: Polarimetric tornado detection. J. Appl. Meteor., 44 , 557–570.
Sachidananda, M., and D. S. Zrnić, 1986: Differential propagation phase shift and rainfall rate estimation. Radio Sci., 21 , 235–247.
Sachidananda, M., and D. S. Zrnić, 1987: Rain rate estimated from differential polarization measurements. J. Atmos. Oceanic Technol., 4 , 588–598.
Scharfenberg, K. A., 2003: Polarimetric radar signatures in microburst-producing thunderstorms. Preprints, 31st Int. Conf. on Radar Meteorology, Seattle, WA, Amer. Meteor. Soc., 581–584.
Scharfenberg, K. A., and Coauthors, 2005: The Joint Polarization Experiment: Polarimetric radar in forecasting and warning decision making. Wea. Forecasting, 20 , 775–788.
Seliga, T. A., and V. N. Bringi, 1976: Potential use of radar differential reflectivity measurements at orthogonal polarizations for measuring precipitation. J. Appl. Meteor., 15 , 69–76.
Straka, J. M., D. S. Zrnić, and A. V. Ryzhkov, 2000: Bulk hydrometeor classification and quantification using polarimetric radar data: Synthesis of relations. J. Appl. Meteor., 39 , 1341–1372.
Thomas, R. J., P. R. Krehbiel, W. Rison, S. Hunyady, W. Winn, T. Hamlin, and J. Harlin, 2004: Accuracy of the lightning mapping array. J. Geophys. Res., 109 , D14207. doi:10.1029/2004JD004549.
Thompson, R. L., R. Edwards, J. A. Hart, K. L. Elmore, and P. Markowski, 2003: Close proximity soundings within supercell environments obtained from the Rapid Update Cycle. Wea. Forecasting, 18 , 1243–1261.
Thompson, R. L., C. M. Mead, and R. Edwards, 2007: Effective storm-relative helicity and bulk shear in supercell thunderstorm environments. Wea. Forecasting, 22 , 102–115.
Tuttle, J. D., V. N. Bringi, H. D. Orville, and F. J. Kopp, 1989: Multiparameter radar study of a microburst: Comparison with model results. J. Atmos. Sci., 46 , 601–620.
Wakimoto, R. M., and V. Bringi, 1988: Dual-polarization observations of microbursts associated with intense convection: The 20 July storm during the MIST project. Mon. Wea. Rev., 116 , 1521–1539.
Wurman, J. M., 2001: The DOW multiple-Doppler network. Preprints, 30th Int. Conf. on Radar Meteorology, Munich, Germany, Amer. Meteor. Soc., 95–97.
Wurman, J. M., and S. Gill, 2000: Finescale radar observations of the Dimmitt, Texas (2 June 1995), tornado. Mon. Wea. Rev., 128 , 2135–2164.
Wurman, J. M., and M. Randall, 2001: An inexpensive, mobile, rapid-scan radar. Preprints, 30th Int. Conf. on Radar Meteorology, Munich, Germany, Amer. Meteor. Soc., 98–100.
Wurman, J. M., J. M. Straka, E. N. Rasmussen, M. Randell, and A. Zahrai, 1997: Design and deployment of a portable, pencil-beam, pulsed, 3-cm Doppler radar. J. Atmos. Oceanic Technol., 14 , 1502–1512.
Xu, J. L., 1983: Hail growth in a three-dimensional cloud model. J. Atmos. Sci., 40 , 185–203.
Zhang, G., R. J. Doviak, and D. S. Zrnić, 2008: Phased array radar polarimetry for weather sensing: Challenges and opportunities. Proc. IEEE Int. Geoscience and Remote Sensing Symp., Boston, MA, IEEE Geoscience and Remote Sensing Society, FR3.203.1.
Zhang, G., R. J. Doviak, D. S. Zrnić, J. Crain, D. Staiman, and Y. Al-Rashid, 2009: Phased array radar polarimetry for weather sensing: A theoretical formulation for bias corrections. IEEE Trans. Geosci. Remote Sens., 47 , 3679–3689. doi:10.1109/TGRS.2009.2029332.
Ziegler, C. L., P. S. Ray, and N. C. Knight, 1983: Hail growth in an Oklahoma multicell storm. J. Atmos. Sci., 40 , 1768–1791.
Zrnić, D. S., 1987: Three-body scattering produces precipitation signature of special diagnostic value. Radio Sci., 22 , 76–86.
Zrnić, D. S., and A. V. Ryzhkov, 1999: Polarimetry for weather surveillance radars. Bull. Amer. Meteor. Soc., 80 , 389–346.
Zrnić, D. S., and Coauthors, 2007: Agile-beam phased array radar for weather observations. Bull. Amer. Meteor. Soc., 88 , 1753–1766.
Conceptual model of cyclic tornadogenesis from Dowell and Bluestein (2002b). (left) The numbered circles identify vortices, thick lines indicate wind shift lines, and shading represents the tornado tracks. (right) Shading (speckling) indicates updraft (downdraft), and dashed (solid) outlines indicate regions of production of cyclonic vertical vorticity by tilting of horizontal vorticity (stretching of vertical vorticity). Arrows indicate vortex-relative trajectories.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Conceptual model of the idealized storm-relative streamlines and the evolution of the ZH field and hook echo (solid black outline) from Beck et al. (2006). Speckling indicates regions of deformation.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Mesoscale features at 0000 UTC 1 Jun 2008. Wind data come from the National Weather Service surface stations; half barbs represent 2.5 m s−1 whereas full barbs represent 5 m s−1. The 500-hPa heights are contoured every 30 m. The dashed line indicates a surface wind shift boundary associated with a stationary front. The large star indicates the first radar echo of the storm. The gray circle markers with “W” and “H” correspond to the Weatherford and Hinton surface stations of the Oklahoma Mesonet, respectively. Abbreviated state names are in gray.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Sounding and 0–6-km hodograph (inset) at 0000 UTC 1 Jun 2008 from Norman, OK. The thick black line is the temperature profile, the dashed line is the dewpoint temperature profile. The hodograph (in m s−1) is labeled at 850, 700, and 500 hPa. The observed storm motion is indicated by a large X.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Radar reflectivity fields every half hour from 0105 to 0630 UTC 1 Jun 2008. Data are displayed as 0.5° elevation angle PPIs, observed using the National Weather Service WSR-88D radar in Twin Lakes, OK (KTLX). (top left) The locations of the Weatherford and Hinton Mesonet sites are indicated by black dots. The storm of interest in this paper is labeled with an A. The second supercell (that merges with storm A by about 0430 UTC) is labeled with a B. The left-moving split of storm A is labeled L. The location of the first echo of storm A is indicated in the top left.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Time series of two Oklahoma Mesonet sites starting at 0200 UTC 1 Jun 2008 for a 6-h period: (a) Weatherford and (b) Hinton. The black line is θυ (in K), barbs show 10-m wind direction and speed (half barb is 2.5 m s−1, full is 5 m s−1); gray filled area is precipitation accumulation (in mm).
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Overlay of low- and midlevel radar features compiled from three elevation angle scans from the volume coverage pattern beginning at 0301:41 UTC. The outlines show the 35-dBZ contour at low levels (solid line, darkest shading, 0.0° elevation), midlevels (2.2° elevation scan, dashed line, lighter shading), and upper levels (5.6° elevation scan, lightest shading). The hail signature (denoted as ZDR less than 0.5 dB, in dark blue) and ZDR arc (in purple) were determined from the 0.0° scan; the ZDR ring and column (red), the regions of graupel (light blue), ρHV ring (yellow), and KDP column (green) were determined from the 2.2° scan. Regions indicated as graupel correspond to areas of moderate ZH above the melting layer (where ZDR is near 0 dB) with very high (>0.98) values of ρHV. Approximate heights AGL of the radar beam at the storm echo centroid are about 0.9, 5.0, and 11.5 km for the 0.0°, 2.2°, and 5.6° elevation angle scans.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Doppler velocity data displayed on PPIs from the 0.5° elevation angle scan on 1 Jun 2008, at the times indicated (times in UTC). The 35-dBZ radar reflectivity contour is overlaid. The approximate center of each analyzed circulation is indicated by an arrow. A threshold of ZH ≥ 4 dBZ is used in the presentation of the velocity data. Distances are relative to KOUN, so the domain shown here is located west and north of the radar.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Example of low-level PPI scans in fields of (a) reflectivity factor ZH, (b) differential reflectivity ZDR, (c) specific differential phase KDP, and (d) cross correlation coefficient ρHV. The 35-dBZ contour of ZH is overlaid on (b)–(d). The occluding mesocyclone (second circulation) is denoted as a solid circle, whereas the newly developed mesocyclone (third circulation) is denoted as the dashed circle. Cyclonic shear identified from Doppler velocity data (see Fig. 8). Two of the circulations from Fig. 8 have been omitted for clarity. Data are from 0247:53 UTC 1 Jun 2008 at 0.5° elevation.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Example of the evolution of the ZDR arc from 0251:26 to 0258:03 UTC 1 Jun 2008, from the lowest elevation angle in each volume scan. The 35-dBZ contour of ZH is overlaid. The arc is initially centered at approximately x = −100 km and y = 40 km.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
As in Fig. 10, but for the data from 0259:16 through 0305:19 UTC.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Evolution of midlevel KDP (light gray shading for values >1° km−1) and ZDR (dark gray shading for values >2 dB) columns represented as quasi-horizontal cross sections as observed from the 1.5° elevation angle scans from 0257:08 through 0315:17 UTC 1 Jun 2008. The 35-dBZ contour of ZH is overlaid. The beam height in the vicinity of the columns was approximately 4.0 km AGL.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
PPI display of an elongated updraft signature in ρHV (centered at about x = −107 km, y = 37 km) from 0301:09 UTC 1 Jun 2008, taken at 4.5° elevation. The solid black line is the 35-dBZ contour of ZH, while the dashed line encircles the region of ZDR less than −0.5 dB. The extremely low values of ρHV behind the storm are associated with the TBSS. The beam height at the location of the updraft signature is approximately 9.7 km AGL.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Evolution of the midlevel ZDR ring from 0245:44 through 0258:26 UTC 1 Jun 2008. Data from 2.2° elevation scans. The radar beam is approximately 5.6 km AGL at the ring.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Regeneration of a midlevel ZDR ring. As in Fig. 14, but for the time period 0319:01 through 0331:03 UTC. The ring is sampled at approximately 4.6 km AGL. Arrows point to regions of strong regeneration.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Evolution of the graupel belt displayed in the field of ρHV. The graupel belt is defined as the region of very high ρHV wrapping around the southern flank of the updraft. The arrow indicates a source region of graupel. Data are from the same scans as Fig. 14 (2.2° elevation angle). The radar beam is at approximately 5.6 km AGL.
Citation: Monthly Weather Review 138, 10; 10.1175/2010MWR3322.1
Specifics of the “supersector” scanning strategy employed in this paper.
This assumes no differential phase shift upon backscatter, δ. In pure rain at S band, δ is negligible; it only becomes significant for particles of the size at which resonance scattering effects occur. For a radar transmitting at wavelength λ and a particle of diameter D and complex dielectric constant ε, resonance scattering occurs where the ratio ℜ = D|ε|1/2/λ is on the order of unity.
“Circulation” is used here instead of “mesocyclone” because only single-Doppler velocity data are available. See French et al. (2008) for a discussion of the ambiguity in determining mesocyclones from single-Doppler data.
