1. Introduction
Since the advent of multiple Doppler wind retrievals (Doviak et al. 1976; Ray et al. 1985; Miller and Anderson 1991), the meteorological community has learned a great deal about weather phenomena where direct in situ measurements are difficult to acquire. Beginning in the late 1970s, techniques were developed using Doppler-derived winds to retrieve dynamic and thermodynamic quantities, such as perturbation pressure and buoyancy (Gal-Chen 1978; Hane and Scott 1978; Hane et al. 1981; Pasken and Lin 1982; Brandes 1984; Roux 1985; Hane and Ray 1985). One such weather phenomenon where these techniques proved particularly useful is the supercell thunderstorm. Although many processes within the spectrum of supercells are still unknown, Doppler wind and (thermodynamic) dynamic retrievals have provided insightful information that has confirmed some hypotheses’ where in situ data was lacking. For example, Cai and Wakimoto (2001) used Doppler-derived quantities to confirm how the pressure field influences storm movement in supercells.
While having access to multiple radars to perform dual- or triple-Doppler wind retrievals is preferred, that is often not that case. For most events, only single-Doppler radar is available for data analysis. Peace et al. (1969) were one of the first to suggest the use of a single radar to derive the horizontal motion fields. Their method was modified and used by Kraus (1974) to study convective storms in New England. The method has since been refined, and is now referred to as the synthetic dual-Doppler (SDD) technique and is described in detail by Bluestein and Hazen (1989) and Klimowski and Marwitz (1992). It was additionally used to study a mesoscale convective system (Bluestein et al. 1994) and a mesoscale winter vortex (Laird et al. 2001). The SDD technique has several limiting factors, including vector of propagation near the radar site [i.e., the feature in question must travel fairly quickly at a close distance (20–40 km) to the radar site]. More importantly, the SDD technique also invokes a quasi-steady-state assumption, where the flow fields cannot change significantly between radar volume scans. Because of these limitations, the SDD technique has rarely been used to retrieve airflow within supercell thunderstorms, which can undergo rapid strengthening (or weakening) of low-level velocity in the span of only a few radar volume scans, especially during tornadogenesis. However, the mesoscale structure of a mature supercell thunderstorm can often maintain a quasi–steady state for several radar volume scans (Lemon and Doswell 1979; Weisman and Klemp 1984).
The Super Tuesday tornado outbreak of 5–6 February 2008 provided two cases where a SDD analysis might be successfully applied to a mature supercell. One case occurred within range of the Doppler radar near Nashville, Tennessee (KOHX; case A), and is the primary focus of this study. This case occurred after midnight (near 0700 UTC), when the Super Tuesday outbreak transitioned to a low convective available potential energy (CAPE), high shear event. The second case (case B) occurred near Memphis, Tennessee; however, it did not readily meet the quasi-steady-state assumption and is therefore only used for comparisons of the SDD method against case A. Dynamic and thermodynamic retrievals were performed on both cases. Additionally, a vertical vorticity budget analysis was performed on case A, following Roberts and Wilson (1995).
The primary purpose of this paper is to examine the kinematic and dynamic structure of case A, a supercell thunderstorm occurring in a nocturnal, high shear, low CAPE environment (common characteristics of a cold season outbreak) in the Southeast, using an analysis method not readily used on supercell thunderstorms. The dynamic retrievals are utilized to determine the dominant updraft forcing mechanisms. These results are compared with previous work, including true dual-Doppler analyses of supercell storms that occur in the Great Plains during the afternoon and early evening hours. This comparison should not only demonstrate the relative success (or failure) of the SDD technique, but also show storm-scale differences between the typical spring season supercell storm of the high plains (often high CAPE and high shear) and one that developed during the nocturnal hours of this cold season outbreak (low CAPE, high shear).
2. Event and case overview
a. Super Tuesday tornado outbreak of 5–6 February 2008
The Super Tuesday tornado outbreak produced 87 tornadoes, caused 57 fatalities and hundreds of injuries within a 15-h period between 2100 UTC 5 February and 1200 UTC 6 February 2008. The number of fatalities alone was the most in a single outbreak since 1985 (Hayes 2009).1 A map of the tornado tracks and their enhanced Fujita ratings is shown in Fig. 1. Favorable conditions for a significant outbreak began appearing in forecast models days before the event. By 3 February, southerly flow induced by strengthening low pressure east of the Rockies, and high pressure moving off the East Coast produced strong warm air and moisture advection throughout the southeastern United States. By Tuesday, 5 February, dewpoints were nearing 21°C as far north as Tennessee and Kentucky. This unstable air mass, along with sufficient environmental wind shear aided by the development of a strong low-level jet, provided an environment ripe for supercell development.
Map of the southeastern United States showing tornado tracks and ratings for all tornadoes during the Super Tuesday tornado outbreak of 5–6 Feb 2008 (Hayes 2009). Tornado tracks associated with the two storms being studied have been highlighted.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Map of the southeastern United States showing tornado tracks and ratings for all tornadoes during the Super Tuesday tornado outbreak of 5–6 Feb 2008 (Hayes 2009). Tornado tracks associated with the two storms being studied have been highlighted.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Map of the southeastern United States showing tornado tracks and ratings for all tornadoes during the Super Tuesday tornado outbreak of 5–6 Feb 2008 (Hayes 2009). Tornado tracks associated with the two storms being studied have been highlighted.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
During the morning of 5 February, a squall line developed across Texas ahead of a surface cold front. At 1630 UTC, the Storm Prediction Center (SPC) outlined large portions of the Mississippi and Ohio River Valleys for a moderate and high risk of severe weather. By 2100 UTC, a broad area of surface based CAPE (SBCAPE) greater than 1000 J kg−1 covered the southeastern United States, along with 0–1-km storm-relative helicity (SRH) approaching 300 m2 s−2. The SPC mesoanalysis SRH assumes a right-moving supercell using the Bunkers et al. (2000) technique. Supercells readily developed in this environment, and the first tornado was reported at 2132 UTC, near the axis of greatest SRH in southeastern Arkansas (Fig. 2).
SPC mesoanalysis graphics for (top) SBCAPE (solid contours, J kg−1) and surface-based convective inhibition (SBCIN) (dashed contours and shading, J kg−1) and (bottom) 0–1-km SRH (m2 s−2) at 2100 UTC 5 Feb 2008 centered on the southeast United States. The wind barbs represent (top) surface wind (kt; 1 kt = 0.5144 m s−1) and (bottom) predicted storm motion (kt). The black dot in southeastern Arkansas indicates the approximate location of the first reported tornado in the outbreak.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
SPC mesoanalysis graphics for (top) SBCAPE (solid contours, J kg−1) and surface-based convective inhibition (SBCIN) (dashed contours and shading, J kg−1) and (bottom) 0–1-km SRH (m2 s−2) at 2100 UTC 5 Feb 2008 centered on the southeast United States. The wind barbs represent (top) surface wind (kt; 1 kt = 0.5144 m s−1) and (bottom) predicted storm motion (kt). The black dot in southeastern Arkansas indicates the approximate location of the first reported tornado in the outbreak.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
SPC mesoanalysis graphics for (top) SBCAPE (solid contours, J kg−1) and surface-based convective inhibition (SBCIN) (dashed contours and shading, J kg−1) and (bottom) 0–1-km SRH (m2 s−2) at 2100 UTC 5 Feb 2008 centered on the southeast United States. The wind barbs represent (top) surface wind (kt; 1 kt = 0.5144 m s−1) and (bottom) predicted storm motion (kt). The black dot in southeastern Arkansas indicates the approximate location of the first reported tornado in the outbreak.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
b. Cases
Case A occurred near Nashville between 0645 and 0725 UTC. The initial radar echoes for this case first appeared in Panola County, Mississippi (an azimuth of 237° and range of 372 km from KOHX), near 0220 UTC, where a collection of cells (4–5 individual echoes) developed. The reflectivity Z in these cells increased, both in intensity and spatial coverage, as they tracked toward the east and northeast into Tennessee and organized into a secondary line, just ahead of the primary quasi-linear convective system (QLCS). By 0410 UTC, one cell slightly ahead of the secondary line became better organized as cyclonic rotation developed at the 1.8° and 2.4° radar tilts from the Memphis radar (KNQA; approximately 2600–3300 m AGL). The developing mesocyclone in this cell began to propagate downward according to radar (Trapp and Davies-Jones 1997; Trapp et al. 1999). Case A produced EF0 and EF1 tornadoes in Sumner County, Tennessee, just north of Nashville and KOHX near 0714 and 0722 UTC, respectively. The supercell characteristics of the storm became more prominent as it moved into Kentucky (increase in Z in the forward flank, strengthening cyclonic rotational velocity VROT, and a better defined appendage indicating enhanced rotation), where an EF3 tornado subsequently formed at 0740 UTC, during merger with the primary squall line.
Radar volumes at 0650 and 0712 UTC were utilized for case A in the SDD procedure, when the supercell passed KOHX at a distance of 16–25 km, moving from 242° at 24 m s−1 (Fig. 3). A subjective visual comparison of Z and radial velocity VR revealed the storm was nearly steady between these two times, making it an ideal case for the SDD technique. This visual comparison compared the shape, structure, and strength of the Z and VR fields at all elevation angels. The height of the 50-dBZ echo remained generally unchanged during the analysis, ranging from 6.5 to 7 km AGL. Additionally, an examination of VROT also revealed a nearly steady mesocyclone; VROT values alternated from 8 to 12 m s−1 during the analysis interval. However, it should be noted that VROT increased to near 24 m s−1 shortly after the analysis time, when the storm produced an EF1 tornado at 0722 UTC.
Base reflectivity from KOHX at 0.5° showing the two volume scans used for the SDD analysis performed on case A: (top) 0650 and (bottom) 0712 UTC. The black line from the radar to the core of case A indicates the change in azimuth angle between the two times was near 64°.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Base reflectivity from KOHX at 0.5° showing the two volume scans used for the SDD analysis performed on case A: (top) 0650 and (bottom) 0712 UTC. The black line from the radar to the core of case A indicates the change in azimuth angle between the two times was near 64°.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Base reflectivity from KOHX at 0.5° showing the two volume scans used for the SDD analysis performed on case A: (top) 0650 and (bottom) 0712 UTC. The black line from the radar to the core of case A indicates the change in azimuth angle between the two times was near 64°.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Case A occurred in a nocturnal environment where SBCAPE decreased (near and below 500 J kg−1) but 0–1-km SRH remained elevated and even increased over time (near 400 m2 s−2) (Fig. 4). A representative sounding for the environment near case A is shown in Fig. 5 (top panel). This is a Rapid Update Cycle (RUC) model sounding centered on Nashville at 0700 UTC, modified by adjusting the lowest level temperature and dewpoint using the 0700 UTC surface observation from KBNA. This sounding shows some cooling at the surface and the development of an inversion, limiting the SBCAPE. However, there is energy available if a parcel is able to ascend to the LFC above the inversion near 810 hPa (about 1.8 km AGL), where it could realize CAPE near 700 J kg−1. The LCL of this sounding is near 920 hPa (about 0.6 km AGL).
As in Fig. 2, but at 0700 UTC 6 Feb 2008. The black dot indicates the approximate location of case A.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
As in Fig. 2, but at 0700 UTC 6 Feb 2008. The black dot indicates the approximate location of case A.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
As in Fig. 2, but at 0700 UTC 6 Feb 2008. The black dot indicates the approximate location of case A.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Modified RUC soundings centered on (top) Nashville, TN, at 0700 UTC and (bottom) Memphis, TN, at 0000 UTC 6 Feb 2008. The solid black and gray lines represent the environmental temperature and dewpoint, respectively. The dashed line represents the path a parcel would take if lifted from the surface. The CAPE in the KBNA and KMEM soundings are near 700 and 1400 J kg−1, respectively.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Modified RUC soundings centered on (top) Nashville, TN, at 0700 UTC and (bottom) Memphis, TN, at 0000 UTC 6 Feb 2008. The solid black and gray lines represent the environmental temperature and dewpoint, respectively. The dashed line represents the path a parcel would take if lifted from the surface. The CAPE in the KBNA and KMEM soundings are near 700 and 1400 J kg−1, respectively.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Modified RUC soundings centered on (top) Nashville, TN, at 0700 UTC and (bottom) Memphis, TN, at 0000 UTC 6 Feb 2008. The solid black and gray lines represent the environmental temperature and dewpoint, respectively. The dashed line represents the path a parcel would take if lifted from the surface. The CAPE in the KBNA and KMEM soundings are near 700 and 1400 J kg−1, respectively.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Case B was analyzed just after sunset near Memphis between 2320 and 0030 UTC. Its initial radar echoes first appeared in Ashley County, Arkansas (an azimuth of 215° and range of 296 km from KNQA), around 2000 UTC. This initial convection tracked along the Mississippi River, where its Z field strengthened and expanded and rotation became apparent in VR by 2200 UTC. As the supercell moved into Tennessee (2330 UTC), it produced an EF2 tornado beginning just north of Southaven, Mississippi (an azimuth of 197° and range of 43 km from KNQA). This tornado was the first of a series of seven tornadoes associated with case B. It also produced a long track EF3 and EF4 tornado that devastated Union University near Jackson, Tennessee, beginning at 0022 UTC.
Radar volumes at 2350 and 0020 UTC were utilized for case B in the SDD procedure, when the mesocyclone passed KNQA at a distance of 25–38 km, moving from 230° at 26 m s−1 (Fig. 6). In this case, the storm did not appear as steady as case A, making it less ideal for the SDD technique. The height of the 50-dBZ echo was unsteady during the analysis time; increasing from near 4.5 km to 8 km AGL just prior to the first volume scan, and then decreasing to near 5.5 km AGL just before the second volume scan. The VROT was also quite unsteady, decreasing from 20 m s−1 to <5 m s−1 during the time surrounding the first volume scan, while slightly increasing to near 10 m s−1 with the second volume scan. Shortly after the second volume scan, VROT increased to 30 m s−1 during the development of an EF3 tornado. Additionally, the forward flank passed KNQA at a closer distance (10–15 km), putting it too close to the (synthetic) radar baseline, which does not allow for accurate wind retrievals because of the geometry along the baseline. Based on the visual comparisons, the results of case A should contain less error than the results of case B. As such, case A is considered the primary case for this study. The results of case B are primarily used for comparisons with case A.
Base reflectivity from KNQA at 1.3° showing the two volume scans used for the SDD analysis performed on case B: (top) 2350 and (bottom) 0018 UTC. The black line from the radar to the hook echo of case B indicates the change in azimuthal angle between the two times was near 75°.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Base reflectivity from KNQA at 1.3° showing the two volume scans used for the SDD analysis performed on case B: (top) 2350 and (bottom) 0018 UTC. The black line from the radar to the hook echo of case B indicates the change in azimuthal angle between the two times was near 75°.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Base reflectivity from KNQA at 1.3° showing the two volume scans used for the SDD analysis performed on case B: (top) 2350 and (bottom) 0018 UTC. The black line from the radar to the hook echo of case B indicates the change in azimuthal angle between the two times was near 75°.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
3. Data and methodology
a. Wind retrieval
Doppler radar data were the primary data source used in this study. Radar data from the National Weather Service’s Weather Surveillance Radar-1988 Doppler (WSR-88D) radars were obtained from the National Climatic Data Center (NCDC) in the raw, unedited, level II format. The level II data includes Z, VR, and spectrum width. The radar volume scans were edited using the National Center for Atmospheric Research (NCAR) SOLOII (Oye et al. 1995) software to remove second trip signals and ground clutter, and to also unfold aliased VR. Reflectivity at the lowest elevation angle was thresholded around −5 dBZ to alleviate possible ground clutter contamination (i.e., Z ≤ −5 dBZ was removed). Ground clutter that escaped the thresholding technique was manually removed by deleting individual outlier pixels with a near-zero VR within the SOLOII window. Using the REORDER (Oye et al. 1995) software package, each edited radar volume was interpolated and gridded to a Cartesian coordinate system using the Cressman weighting function (Cressman 1959). The data were gridded with a grid spacing of 0.5 km × 0.5 km × 0.5 km in the x, y, and z directions, with a radius of influence of 1 km and a vertical extent to 15 km with the origin [0, 0, 0] located at the radar site.
Klimowski and Marwitz (1992) describe the SDD procedure in detail and this study follows their procedure closely. The storm must move perpendicular to the radar line of sight at a relatively close distance (20–40 km) and travel far enough so that the azimuth angle α to the storm core changes by at least 30°. Additionally, the storm flow fields associated with the feature cannot significantly change (i.e., the steady-state assumption is invoked). The goal is to find two volume scans separated by a minimum temporal interval so that the feature in question does not change (strengthen, weaken, etc.) significantly, and also satisfying the requirement that the azimuth angle is nearly perpendicular to the storm motion vector. In other words, one needs to find a storm/feature that is moving quickly and perpendicular to the radars line of sight. In this study, similar to Klimowski and Marwitz, steadiness was estimated by visually comparing the reflectivity and velocity signatures of several volume scans (Nelson and Knight 1987; Miller et al. 1990).
After two separate volume scans meeting the aforementioned requirements were identified, the translation vector between the supercell at time t and t + Δts (Δts being the time between radar volume scans) was determined. By matching reflectivity and/or velocity features common to both volume scans, the translation vector is estimated. Then, using the known translation vector, the storm at t + Δts is essentially moved “back” to match up with the storm at time t. This procedure is accomplished via REORDER when converting to the Cartesian coordinate space; the actual radar coordinates are edited in order to change the location of the radar. The radar coordinates for the second volume scan are changed so the supercell at both volume scan times (t and t + Δts) are matched and the correlation between the reflectivity fields is maximized. A correlation analysis of the matched Z fields yielded r values of 0.784 and 0.652, and standard differences of 5.69 and 11.62 dBZ, for cases A and B, respectively. The translated radar effectively becomes the second (“synthetic”) radar, and the rest of the wind retrieval process is similar to the standard dual-Doppler retrieval technique.
The (synthetic) dual-Doppler geometry is shown in Fig. 7 (Lhermitte and Miller 1970). The length of the baseline b is defined as cΔts, where c is storm speed and Δts remains the time between radar volume scans. The other variables are defined as follows: Ri is the initial radar position, Rt is the translated radar position, β is the radar azimuth angle, and θ is the radar elevation angle. Although error may come from many sources during a SDD analysis, the main source is usually changes in airflow (storm evolution) between the volume scans. Additional errors can be introduced through uncertainties in calculating the translation vector, the magnitude of which is the length of the radar baseline. Also, inherent errors associated with the dual-Doppler technique would also apply to the SDD retrieval since the techniques are similar (Doviak et al. 1976; Klimowski and Marwitz 1992). However, one advantage of the SDD analysis is that the same radar is used for each volume scan, whereas, a standard multi-Doppler analysis must use at least two different radars, thus two sets of radar assumptions and/or timing errors must be taking into account. The wind retrieval was performed using NCAR’s Custom Editing and Display of Reduced Information in Cartesian Space (CEDRIC; Mohr et al. 1986). Following retrieval of the horizontal winds, the vertical wind component was found by integrating mass continuity using a variational integration scheme (O’Brien 1970; Doviak et al. 1976; Ray et al. 1985; Miller and Anderson 1991; Matejka and Bartels 1998).
Geometry of the synthetic dual-Doppler analysis technique (Lhermitte and Miller 1970). The symbols are defined in the text.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Geometry of the synthetic dual-Doppler analysis technique (Lhermitte and Miller 1970). The symbols are defined in the text.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Geometry of the synthetic dual-Doppler analysis technique (Lhermitte and Miller 1970). The symbols are defined in the text.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
b. Pressure perturbation retrieval
Once the three-dimensional wind field is found, rearrangement of the momentum equations yields an estimate of the pressure perturbation p′ field. Gal-Chen (1978) pioneered the retrieval of dynamic and thermodynamic variables from observed winds while many others have used and modified the procedure (Hane et al. 1981; Pasken and Lin 1982; Brandes 1984; Roux 1985; Hane and Ray 1985; Cai and Wakimoto 2001). The method performed here is similar to that done by others.
Equation (4) is a partial differential equation known as a Poisson equation, which is subject to Neumann boundary conditions when being solved. The pressure perturbation retrieval is performed within CEDRIC. The resulting pressure perturbation fields are then filtered to minimize possible error. Given that a steady-state assumption is invoked in the SDD analysis and the fact that there are no subsequent analysis times, time derivatives are set to zero. This has the potential to introduce error into the p′ retrievals but especially any possible buoyancy retrieval (see the appendix). However, by definition of the steady-state assumption, storm features did not change significantly during the analysis period and this is especially true for observed horizontal flow features of case A, thereby minimizing error caused by the elimination of time derivatives. However, it is admitted that errors are expected to be larger in the vicinity of the mesocyclone, where flow features are likely evolving more quickly than the rest of the storm (Gal-Chen 1978; Hane et al. 1981; Hane and Ray 1985).
c. Vorticity production
4. Results
a. Case A
Reflectivity Z (dBZ) overlain with the horizontal storm-relative wind vectors is shown in Fig. 8 at 0.5, 1.5, 3.0, and 4.0 km AGL. The storm-relative flow indicates strong inflow from the southeast and rotation within a well-defined mesocyclone, particularly in the lower levels at 0.5 and 1.5 km. Placement of the circulation is consistent with the Z distribution. Relative storm inflow is strong at 0.5 and 1.5 km and matches up well with the “inflow notch” in the Z field. This storm has an elongated region of high Z, and its east–west dimension (as defined by the 40-dBZ contour) of 10–12 km is small when compared to case B or other supercells that formed during the outbreak. An asymmetry in the storm-relative flow is present (stronger southerly flow east of the circulation center) leading to convergence along the rear-flank gust front at 0.5 and 1.5 km. This convergent boundary is manifested as an area of enhanced vertical motion in the low levels (Figs. 9a,b). Overall, there does not seem to be much (if any) suspect data with this analysis and the flow features appear to be physically consistent and are similar to the known features of a supercell storm (Lemon and Doswell 1979; Thompson 1998; Klimowski et al. 2003; Bunkers et al. 2006).
Reflectivity (shading and contours; dBZ) and storm-relative wind flow (vectors; m s−1) at (a) 0.5, (b) 1.5, (c) 3.0, and (d) 4.0 km AGL for case A. The horizontal and vertical dashed lines represent locations of vertical cross sections. The thick contour line represents 40 dBZ and is represented in the subsequent figures.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Reflectivity (shading and contours; dBZ) and storm-relative wind flow (vectors; m s−1) at (a) 0.5, (b) 1.5, (c) 3.0, and (d) 4.0 km AGL for case A. The horizontal and vertical dashed lines represent locations of vertical cross sections. The thick contour line represents 40 dBZ and is represented in the subsequent figures.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Reflectivity (shading and contours; dBZ) and storm-relative wind flow (vectors; m s−1) at (a) 0.5, (b) 1.5, (c) 3.0, and (d) 4.0 km AGL for case A. The horizontal and vertical dashed lines represent locations of vertical cross sections. The thick contour line represents 40 dBZ and is represented in the subsequent figures.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical motion (m s−1; shading and contours; contour lines with an interval of 10 m s−1) at (a) 0.5, (b) 1.5, (c) 3.0, and (d) 4.0 km AGL for case A. The shaded contours are positive values and the unshaded, dotted contour is vertical motion ≤−5 m s−1. The thick black line is a 40-dBZ contour. Vertical motion peaks at 34.5 m s−1 at 3.0 km.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical motion (m s−1; shading and contours; contour lines with an interval of 10 m s−1) at (a) 0.5, (b) 1.5, (c) 3.0, and (d) 4.0 km AGL for case A. The shaded contours are positive values and the unshaded, dotted contour is vertical motion ≤−5 m s−1. The thick black line is a 40-dBZ contour. Vertical motion peaks at 34.5 m s−1 at 3.0 km.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical motion (m s−1; shading and contours; contour lines with an interval of 10 m s−1) at (a) 0.5, (b) 1.5, (c) 3.0, and (d) 4.0 km AGL for case A. The shaded contours are positive values and the unshaded, dotted contour is vertical motion ≤−5 m s−1. The thick black line is a 40-dBZ contour. Vertical motion peaks at 34.5 m s−1 at 3.0 km.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical cross sections of Z and storm-relative winds (Fig. 10) were taken along the dotted lines in Fig. 8 (x = −21 km and y = 20 km). A clearly defined updraft is exhibited on the southern flank of the supercell, originating below a Z overhang, in the north–south cross section (Fig. 10a). Downward motion is located within the forward flank, maximizing near the surface, just below an elevated +55-dBZ echo (likely a hail core). The 30- and 40-dBZ contours reach maximum heights of near 9 and 7 km AGL, respectively. This is lower in altitude than that seen in the more classic supercell (e.g., Gilmore and Wicker 2002; Steiger et al. 2007) indicating the shallow nature of this storm. Additionally, precipitation transport, as indicated by the storm-relative flow, primarily occurs to the north and east (above 4 km; Fig. 10b), away from the Z core. This likely contributes to the elongation of Z toward the northeast seen in the horizontal sections (Fig. 8). Overall, the SDD technique yielded a clean and physically consistent analysis for case A, both in the horizontal and vertical, resulting in increased confidence for the SDD procedure.
(a) North–south and (b) east–west vertical cross sections of Z (contours; dBZ) and storm-relative wind flow (vectors; m s−1) for case A taken at x = −21 km and y = 20 km, respectively.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
(a) North–south and (b) east–west vertical cross sections of Z (contours; dBZ) and storm-relative wind flow (vectors; m s−1) for case A taken at x = −21 km and y = 20 km, respectively.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
(a) North–south and (b) east–west vertical cross sections of Z (contours; dBZ) and storm-relative wind flow (vectors; m s−1) for case A taken at x = −21 km and y = 20 km, respectively.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical motion w (m s−1) at 0.5, 1.5, 3.0, and 4.0 km AGL is shown in Fig. 9. In this and all similar figures, the thick black line represents a 40-dBZ contour. The updraft is rooted at low levels along the convergent boundary between the strong inflow and weak downdraft outflow. Convergence maximizes below 2.0 km (Fig. 11), coincident with an increase in updraft strength. The updraft peaks near 34.5 m s−1 at 3.0 km for this analysis (Figs. 9c and 11). This is a particularly low level for the maximum updraft in a supercell and is actually more akin to the “mini” or “shallow” supercells typically observed embedded within squall lines or tropical cyclone rainbands (McCaul and Weisman 1996; Weisman and Trapp 2003; Eastin and Link 2009). The updraft core, as defined by the +10 m s−1 closed contour, is elongated from the south-southwest to north-northeast, similar to the elongation seen in the Z field. Along the southern end of this elongation at 0.5 and 1.5 km, w ≥ 2 m s−1 tends to demark the convergent boundary along the rear-flank gust front. A secondary updraft maximum (≥10 m s−1) is located to the south-southwest of the main updraft at 0.5 and 1.5 km; this is likely the formation of a new updraft along the flanking line, an area known for updraft regeneration in supercells (Lemon 1976; Bluestein 1986). Subsequent radar data (near 0725–0730 UTC) does indicate the weakening/occlusion of the initial mesocyclone, followed by new mesocyclone development. The east–west diameter of the updraft, defined by the +20 m s−1 closed contour, is 2–3 km, while the north–south diameter is roughly 2 times that distance, again demonstrating the elongation. The maximum updraft decreases slightly between its maximum at 3.0 and 5.5 km before a more rapid decrease is observed above 5.5 km. Vertical motion decreases below 15 m s−1 near 8 km (Fig. 11), indicative of the shallow nature of this supercell. Downdrafts are generally weak and not as well defined when compared to the updraft. An elongated, continuous band of downward motion ≤−5 m s−1 is located within the forward flank of the storm within high Z and extends to the south at 0.5 and 1.5 km. This continuous band breaks up at 3.0 km. Downward motion associated with the rear-flank downdraft (RFD) is weak to nonexistent; the forward-flank downdraft (FFD) appears to dominate.
Vertical profiles of maximum reflectivity (solid line; dBZ), vertical motion (dashed line; m s−1), and convergence (dotted line; −∇ ∙· V s−1, scaled by 103) for case A from 0.5 to 10 km AGL. Values given are the maximum values for the domain indicated in Fig. 8.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical profiles of maximum reflectivity (solid line; dBZ), vertical motion (dashed line; m s−1), and convergence (dotted line; −∇ ∙· V s−1, scaled by 103) for case A from 0.5 to 10 km AGL. Values given are the maximum values for the domain indicated in Fig. 8.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical profiles of maximum reflectivity (solid line; dBZ), vertical motion (dashed line; m s−1), and convergence (dotted line; −∇ ∙· V s−1, scaled by 103) for case A from 0.5 to 10 km AGL. Values given are the maximum values for the domain indicated in Fig. 8.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Pressure perturbations p′ (mb) at 0.5, 1.5, 3.0, and 4.0 km AGL are shown in Fig. 12. A pressure minimum of −3.1 mb was observed at 3.0 km, which is the level of the updraft maximum. This mesolow pressure at 3.0 km (in part associated with the cyclonically rotating updraft), suggests that an upward-directed vertical pressure gradient was established, helping maintain the updraft at the relatively low level previously described. Near the updraft location, the vertical perturbation pressure gradient force (VPGF) maximized in the lowest 2 km (Fig. 13), consistent with the low-level updraft. One should take note of the negative p′ extension toward the east, away from the mesolow at 3.0 km. This negative extension also leads to an upward-directed VPGF along the right flank of the storm in the lowest 2 km (Fig. 13) and is the favored location for storm movement, as described by Rotunno and Klemp (1982) (the updraft grows preferentially on the right flank). The p′ distribution within the rear-flank (Fig. 12) suggests a downward-directed VPGF in this location, such as that indicated in Fig. 13. The magnitude of the VPGF in the lowest levels of the rear flank is weaker than that observed near the updraft; however, the magnitude increases as you go up in elevation in the storm. This weaker low-level forcing might explain why the low-level RFD appeared weak to nonexistent at the analysis time.
As in Fig. 9, but for pressure perturbation (mb; shading and contours; contour lines with an interval of 1 mb). The shaded contours are negative perturbations and the unshaded, dotted contour represents perturbations ≥+1.0 mb. A local minimum of −3.1 mb occurs at 3.0 km within the updraft.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
As in Fig. 9, but for pressure perturbation (mb; shading and contours; contour lines with an interval of 1 mb). The shaded contours are negative perturbations and the unshaded, dotted contour represents perturbations ≥+1.0 mb. A local minimum of −3.1 mb occurs at 3.0 km within the updraft.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
As in Fig. 9, but for pressure perturbation (mb; shading and contours; contour lines with an interval of 1 mb). The shaded contours are negative perturbations and the unshaded, dotted contour represents perturbations ≥+1.0 mb. A local minimum of −3.1 mb occurs at 3.0 km within the updraft.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical perturbation pressure gradient force (m s−2; scaled by 103; contour interval of 0.5 m s−2) at (a) 1.0, (b) 2.0, (c) 3.0, and (d) 4.0 km AGL for case A. Positive and negative values are drawn as thin, solid lines and dashed lines, respectively. The bold, solid line represents the 0 m s−2 VPGF contour. The shading represents Z > 40, 45, and 50 dBZ.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical perturbation pressure gradient force (m s−2; scaled by 103; contour interval of 0.5 m s−2) at (a) 1.0, (b) 2.0, (c) 3.0, and (d) 4.0 km AGL for case A. Positive and negative values are drawn as thin, solid lines and dashed lines, respectively. The bold, solid line represents the 0 m s−2 VPGF contour. The shading represents Z > 40, 45, and 50 dBZ.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical perturbation pressure gradient force (m s−2; scaled by 103; contour interval of 0.5 m s−2) at (a) 1.0, (b) 2.0, (c) 3.0, and (d) 4.0 km AGL for case A. Positive and negative values are drawn as thin, solid lines and dashed lines, respectively. The bold, solid line represents the 0 m s−2 VPGF contour. The shading represents Z > 40, 45, and 50 dBZ.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
A high correlation between positive ζ and updraft is indicated in Fig. 14 (dashed–dotted line represents w ≥≥ 10 m s−1), verifying the existence of the cyclonically rotating updraft. Similar to the updraft and Z field, ζ is elongated from the south-southwest to north-northeast. Vertical vorticity magnitudes increase with height to a maximum of 0.023 s−1 at 2.0 km (not shown); they exceed 0.02 s−1 until 3.5 km. Interestingly, the ζ maximum is 1 km lower than the p′ minimum and updraft maximum, which suggests low-level ζ enhancement via vorticity stretching. Tilting of streamwise vorticity into the vertical also appears to contribute to the mesocyclone, as shown in Fig. 15. The storm-relative flow entering the base of the updraft within the southeast flank is nearly aligned with the horizontal vorticity vector. Horizontal vorticity in the low levels is maximized along and just ahead of the rear-flank gust front, as inferred by the 3 m s−1 vertical motion contour (Fig. 15), suggesting the existence of baroclinic vorticity generation. These processes are consistent with prior observations and numerical simulations of supercell thunderstorms (Klemp et al. 1981; Davies-Jones 1984; Droegemeier et al. 1993; McCaul and Weisman 1996). Also, the maximum updraft and maximum ζ center at each vertical level are slightly offset such that the updraft maximum is found just east of the vorticity maximum.
As in Fig. 9, but for vertical vorticity (s−1; shading and contours; contour lines with an interval of 0.005 s−1). The shaded contours are positive ζ and the unshaded, dotted contour represents ζ ≤ −0.01 s−1. The dashed–dotted line is w ≥ 10 m s−1.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
As in Fig. 9, but for vertical vorticity (s−1; shading and contours; contour lines with an interval of 0.005 s−1). The shaded contours are positive ζ and the unshaded, dotted contour represents ζ ≤ −0.01 s−1. The dashed–dotted line is w ≥ 10 m s−1.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
As in Fig. 9, but for vertical vorticity (s−1; shading and contours; contour lines with an interval of 0.005 s−1). The shaded contours are positive ζ and the unshaded, dotted contour represents ζ ≤ −0.01 s−1. The dashed–dotted line is w ≥ 10 m s−1.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Positive horizontal vorticity (bold, solid contours beginning at 0.005 s−1 with a contour interval of 0.005 s−1), positive vertical motion (dotted contours at 3, 10, and 20 m s−1), storm-relative wind vectors (light gray vectors), and horizontal vorticity vectors (black vectors) at 1.0 km AGL for case A. The shading represents Z > 40, 45, and 50 dBZ.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Positive horizontal vorticity (bold, solid contours beginning at 0.005 s−1 with a contour interval of 0.005 s−1), positive vertical motion (dotted contours at 3, 10, and 20 m s−1), storm-relative wind vectors (light gray vectors), and horizontal vorticity vectors (black vectors) at 1.0 km AGL for case A. The shading represents Z > 40, 45, and 50 dBZ.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Positive horizontal vorticity (bold, solid contours beginning at 0.005 s−1 with a contour interval of 0.005 s−1), positive vertical motion (dotted contours at 3, 10, and 20 m s−1), storm-relative wind vectors (light gray vectors), and horizontal vorticity vectors (black vectors) at 1.0 km AGL for case A. The shading represents Z > 40, 45, and 50 dBZ.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Figure 16 shows the horizontal distribution of the vertical vorticity production terms from the vorticity equation (6) at 0.5, 1.5, and 3.0 km AGL (dashed–dotted line represents ζ ≥ 0.015 s−1). The ζDIV (stretching, Fig. 18c) term is strongly positive at all levels below 3.5 km, reaching a maximum value of 24.5 × 10−5 s−2 at 1.5 km within the main ζ center. A high correlation exists between ζ and ζDIV at 1.5 km; this suggests vorticity stretching is a significant contributor to the ζ maximum at 2.0 km. The magnitude of the ζTILT (Fig. 18d) term is about half of the ζDIV term, reaching a maximum of 13 × 10−5 s−2 at 1.5 km; however, this maximum is found on the southeastern flank of the storm, collocated with the horizontal vorticity maximum (Fig. 15), where southerly flow is entering the base of the updraft (i.e., tilting of streamwise vorticity). The ζυ (vertical advection, Fig. 16b) term is primarily negative below 2.0 km as a result of increasing ζ with height. Positive contributions from ζυ appear at and above 2.5 km and it reaches a maximum magnitude at 4.0 km (not shown) comparable to the stretching term (25.5 × 10−5 s−2). The ζh (horizontal advection, Fig. 16a) term exhibits a positive–negative couplet across the main ζ center; the positive center is on the southeast-to-east flank and the negative center is on the west flank of the ζ center. This positioning rotates with height; at 4.5 km, positive ζh is located on the northeast flank of the ζ center (not shown). This is consistent with the observed northeastern tilt of the Z pattern and mesocyclone. Overall, low-level vorticity production in this storm was dominated by stretching, with additional contributions from tilting. Midlevel vorticity production resulted primarily from the vertical redistribution of vorticity from the lower levels, as indicated by ζυ increasing with height and reaching a maximum at 4.0 km.
Horizontal distributions of vertical vorticity production terms (10−5 s−2; shaded contours with a contour interval of 5 s−2): (a) ζh (storm relative horizontal advection), (b) ζυ (vertical advection), (c) ζDIV (stretching), and (d) ζTILT (tilting), at 0.5, 1.5, and 3.0 km AGL for case A. The shaded contours are positive values and the unshaded, dotted contour is values ≤−5 × 10−5 s−2. The dashed–dotted line represents ζ (vertical vorticity) ≥0.015 s−1.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Horizontal distributions of vertical vorticity production terms (10−5 s−2; shaded contours with a contour interval of 5 s−2): (a) ζh (storm relative horizontal advection), (b) ζυ (vertical advection), (c) ζDIV (stretching), and (d) ζTILT (tilting), at 0.5, 1.5, and 3.0 km AGL for case A. The shaded contours are positive values and the unshaded, dotted contour is values ≤−5 × 10−5 s−2. The dashed–dotted line represents ζ (vertical vorticity) ≥0.015 s−1.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Horizontal distributions of vertical vorticity production terms (10−5 s−2; shaded contours with a contour interval of 5 s−2): (a) ζh (storm relative horizontal advection), (b) ζυ (vertical advection), (c) ζDIV (stretching), and (d) ζTILT (tilting), at 0.5, 1.5, and 3.0 km AGL for case A. The shaded contours are positive values and the unshaded, dotted contour is values ≤−5 × 10−5 s−2. The dashed–dotted line represents ζ (vertical vorticity) ≥0.015 s−1.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
b. Case B
Figure 17 shows Z overlain with the horizontal storm-relative wind vectors at 0.5, 1.5, 3.0, and 4.0 km for case B. The forward flank of this storm was not within suitable (synthetic) dual-Doppler geometry. This is partly due to the distance between the radar, but also the change in angle between volume scans in this region was smaller (30°–40°) than in the updraft core. Thus, the wind retrieval was thresholded to only include the mesocyclone core as defined by the box in Fig. 17. The storm-relative flow indicates strong cyclonic flow within the mesocyclone at all levels. Strong inflow and convergence along the rear-flank gust front is also very prominent at 0.5 and 1.5 km (Figs. 17a,b); similar to case A, this is essentially the start of the flanking line. The Z structure appears more similar to the “classic” supercell, with a well-defined hook echo in the lower levels, an expansive forward flank, and a V notch in the Z field. The structure of this storm is quite different than what was seen in case A.
As in Fig. 8, but for case B. Here, the wind retrieval was thresholded to only include the area near the updraft core, as defined by the dashed box.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
As in Fig. 8, but for case B. Here, the wind retrieval was thresholded to only include the area near the updraft core, as defined by the dashed box.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
As in Fig. 8, but for case B. Here, the wind retrieval was thresholded to only include the area near the updraft core, as defined by the dashed box.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical profiles of maximum Z, maximum w in the updraft, minimum w in the RFD, and minimum p′ in the updraft core are shown in Fig. 18. The maximum updraft increases rapidly in the lowest levels of the storm until it reaches 4 km AGL. Between 4 and 6.5 km, the updraft reaches its maximum value, but also remains nearly steady, ranging from 52 to 54 m s−1. The updraft rapidly decreases above 6.5 km. Within the RFD, downward motion attained a minimum near −14 m s−1. This is in stark contrast to case A, where evidence of the RFD was minimal. Although not shown, it is important to note that the downward VPGF within the RFD was much stronger here than in case A, obtaining a minimum of −3.52 m s−2 near 2 km AGL. The strong p′ forcing in the RFD was also located closer to the surface than that observed in case A. Interestingly, reflectivity is maximized much lower in case B (64 dBZ at 3 km AGL) than what was analyzed in case A (60 dBZ at 6 km AGL; Fig. 11). The retrieved p′ minimum within the updraft core was near −3.0 mb at 5 km, similar in magnitude to case A.
Vertical profiles of maximum reflectivity (solid line; dBZ), maximum vertical motion (dashed line; m s−1), minimum vertical motion in the RFD (dashed–dotted line; m s−1), and minimum pressure perturbation in the updraft core (dotted line; ×10 mb) for case B from 0.5 to 10 km AGL.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical profiles of maximum reflectivity (solid line; dBZ), maximum vertical motion (dashed line; m s−1), minimum vertical motion in the RFD (dashed–dotted line; m s−1), and minimum pressure perturbation in the updraft core (dotted line; ×10 mb) for case B from 0.5 to 10 km AGL.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
Vertical profiles of maximum reflectivity (solid line; dBZ), maximum vertical motion (dashed line; m s−1), minimum vertical motion in the RFD (dashed–dotted line; m s−1), and minimum pressure perturbation in the updraft core (dotted line; ×10 mb) for case B from 0.5 to 10 km AGL.
Citation: Monthly Weather Review 141, 2; 10.1175/MWR-D-12-00035.1
5. Discussion and conclusions
This study has presented one of the first comprehensive kinematic analyses of a cold season supercell storm using only one Doppler radar. Previous utilizations of the synthetic dual-Doppler technique on supercells (Klimowski and Marwitz 1992) were primarily to evaluate it. Major findings, particularly for case A, are listed below with further details given within the text:
The maximum updraft (34.5 m s−1) occurred at a very low height (3.0 km AGL) and was related to a low-level, shear-induced upward-directed vertical pressure gradient.
Low-level vorticity production appeared to be dominated by stretching; however, there are strong indications that tilting of baroclinically produced vorticity also contributed to the low-level mesocyclone.
The low-level pressure forcing within the rear flank resulted in a weak to nonexistent RFD. This finding might provide evidence why the tornadoes near the analysis time were weak and short lived.
In contrast, case B exhibited stronger downward pressure forcing within the rear flank, which likely contributed to the stronger RFD. Additionally, stronger and longer lived tornadoes occurred near the analysis time.
Pressure retrievals using multiple Doppler winds were examined by Hane and Ray (1985). They found a high–low couplet existed at each level across the updraft with a horizontal pressure gradient typically oriented along the environmental shear vector, as predicted by linear theory (Rotunno and Klemp 1982). A similar distribution is found in this study. A significant pressure minimum is found near or around the rotating updraft, attributed to the strong vertical vorticity and cyclostrophic balance in the mesocyclone. Hane and Ray (1985) found positive p′ at high levels in the RFD, creating a downward-directed vertical pressure gradient. It is believed that vertical gradients of nonhydrostatic pressure perturbations are important to the formation and evolution of the RFD (Markowski 2002). Low-level p′ forcing within the RFD of case A (Fig. 13) were weak, likely contributing to the weak to nonexistent downdraft. The opposite was true for case B, where the low-level downward forcing in the RFD was stronger than that observed in case A, likely contributing to stronger downdrafts.
Interestingly, the updraft maximized at a low level of 3 km for case A. Given the limited amount of CAPE (500–700 J kg−1), this low-level maximum is (somewhat) consistent with numerical simulations of supercells. Wicker and Cantrell (1996) showed that for a low CAPE environment (600 J kg−1) supercell updrafts peak around 5–6 km but in high CAPE environments (2200 J kg−1), updrafts attained maximum values near 10 km. A lower maximum updraft height was expected, but the height for case A was still 40%–50% lower than that shown in the Wicker and Cantrell study. However, a large part of the total updraft in supercell storms is typically derived from the VPGF distribution that arises due to interactions between the updraft and wind shear (Rotunno and Klemp 1982) and this arrangement near the updraft does agree with the low-level maximum. As previously mentioned, this low-level maximum is similar to what one might find in the “mini” or “shallow” supercells, such as those observed within squall lines or tropical cyclone rainbands (McCaul and Weisman 1996; Weisman and Trapp 2003; Eastin and Link 2009). However, this storm was still somewhat isolated and was not fully overtaken (absorbed) by the primary QLCS until it moved into Kentucky. It is important to note that up to this point, case A produced weak, short-lived tornadoes. Shortly after this analysis, as it moved into Kentucky, an EF3 tornado developed as the storm appeared to merge with the QLCS (high Z was no longer isolated in nature). Perhaps the approaching QLCS altered the near-storm environment. Although beyond the scope of this paper, it is speculated that this analysis captured case A in a transitioning phase due to interactions with the QLCS. The CAPE values of 1000–1400 J kg−1 near Memphis at the time of case B (2300–0000 UTC) were near double that of case A, so the relatively low-level updraft maximum (52–54 m s−1 between 4 and 6.5 km AGL) was unexpected. Unfortunately, the wind retrieval was of lower quality because of storm evolution during the analysis period, so the low-level maximum could very well be erroneous.
Although not shown, estimates of total buoyancy deviation were retrieved for case A (see the appendix for details on the retrieval) where a maximum deviation of 2.5°C near the updraft maximum at 3.0 km (Figs. 9c and 11) was found. In light of the sounding data (Fig. 5), the magnitude of this retrieval appears suspicious; however, the fact that a maximum positive temperature anomaly was retrieved near the updraft maximum is important to note. Shear-induced pressure perturbations (Rotunno and Klemp 1982, 1985; Barnes 1995; Cai and Wakimoto 2001) and the vertical distribution of buoyancy (McCaul and Weisman 1996, 2001; Wicker and Cantrell 1996; McCaul and Cohen 2002; Kirkpatrick et al. 2009) are vital for supercell maintenance in high shear, low CAPE environments. The negative p′ extension seen in Fig. 12 is due in part by the high shear the storm encountered, helping regenerate and maintain the updraft of the storm. Additionally, the acceleration caused by the upward-directed pressure gradient within and near the updraft can compensate for the lower CAPE. The maximum buoyancy deviation near the updraft maximum was also located near the LFC according to model soundings and the SPC mesoanalysis. This is important, since buoyancy concentrated near the LFC helps the updraft overcome strongly sheared environments (Kirkpatrick et al. 2009).
It is generally accepted that tilting of environmental horizontal vorticity via the updraft and subsequent stretching is the source of low to midlevel rotation for the mesocyclone (Barnes 1970; Brandes 1984; Rotunno and Klemp 1985; Davies-Jones et al. 2001). This vertical vorticity production is maximized when streamwise vorticity is tilted into the vertical and a positive correlation exists between vertical velocity and vertical vorticity (Davies-Jones 1984). These processes were analyzed in case A, even as it was beginning to be absorbed by the primary QLCS. Both ζDIV and ζTILT are maximized just below the level of maximum ζ, indicating tilting and stretching of horizontal (streamwise) vorticity.
The focus now has typically shifted to the source of near-surface rotation, specifically that required for tornadogenesis. It is believed that tilting of horizontal vorticity generated along the baroclinic zones of the supercell gust fronts may play a small role (Rotunno and Klemp 1985; Markowski and Richardson 2008). However, more importantly is the presence of descending relatively warm air within the RFD, advecting vertical vorticity toward the surface (Brandes 1978; Davies-Jones and Brooks 1993; Dowell and Bluestein 1997; Davies-Jones et al. 2001; Markowski 2002; Markowski and Richardson 2008). Unfortunately, because of radar limitations (beam height), the near-surface layer is not readily sampled. Additionally, it is difficult to resolve small-scale flows using the SDD technique. However, the nonhydrostatic pressure forcing and incipient RFD was weak to nonexistent during the analysis time of case A, perhaps offering an explanation as to why significant tornadoes were not associated with case A before or during the analysis time. It was not until it began merging with the primary squall line that a significant tornado (EF3) developed.
While the SDD analysis technique is not new, it is seldom applied to supercell thunderstorms. However, case A demonstrates the usefulness of the technique from a mesoscale perspective within a postevent reanalysis, given the requirements and assumptions are met. The feature needs to meet the steady-state assumption, as well as propagate at a close enough distance to a radar site, at a fast speed, to move through at least 30° of radar azimuth. It’s often difficult to meet these requirements and even if they are met, one should be aware that the SDD retrieval is not a perfect solution, something to be mindful of when viewing the results. However, this analysis technique provides a dependable first guess as to storm processes and should be given more consideration when multiple radars are not available and conditions are reasonably satisfied. Even though mobile radars are becoming more widespread, it is generally not possible to construct multiple Doppler analyses for a substantial number of tornado events, particularly those in the Southeast where tree and terrain blockage is significant. This analysis and others using the SDD technique (Bluestein and Hazen 1989; Klimowski and Marwitz 1992; Laird et al. 2001) show that it is a viable solution, for a wide range of meteorological phenomenon, when the need for wind retrievals is warranted but only a single radar is available.
Acknowledgments
We wish to thank Rich Thompson at the SPC for providing the mesoanalysis data and graphics. We would also like to acknowledge Dr. Cody Kirkpatrick and three anonymous reviewers for providing comments that greatly improved the original manuscript. This research originated from the corresponding author’s masters research, so we would be remiss if we did not acknowledge his committee: Drs. Larry Carey, Walt Petersen, and Tim Coleman. This research was supported by NOAA Grants NA08OAR4600896 and NA09OAR4600204.
APPENDIX
Buoyancy Retrieval
A full retrieval of the buoyancy deviation is quite sensitive to time derivatives of vertical velocity. As such, the estimates completed in this study likely contain errors due to the steady-state assumption invoked for the SDD analysis. Additionally, it can also be difficult to determine accurately from Doppler observations because of potential errors in the observations, long periods between volume scans, and quasi-horizontal scanning (Gal-Chen and Kropfli 1984; Crook 1994; Dowell et al. 2004). However, Doppler-retrieved estimates of buoyancy, while not ideal, have been used in the past to identify important features, such as midlevel temperature anomalies in updrafts (Brandes 1984; Hane and Ray 1985). Additionally, the buoyancy estimation as outlined above assumes an ice-free cloud. The storms in this study occurred in an environment where the freezing level was 3000–4000 m AGL, so ice in the form of graupel and hail is certainly present. This is confirmed by several hail reports just prior to the analysis time.
REFERENCES
Barnes, G. M., 1995: Updraft evolution: A perspective from cloud base. Mon. Wea. Rev., 123, 2693–2715.
Barnes, S. L., 1970: Some aspects of a severe, right-moving thunderstorm deduced from mesonetwork rawinsonde observations. J. Atmos. Sci., 27, 634–648.
Bluestein, H. B., 1986: Visual aspects of the flanking line in severe thunderstorms. Mon. Wea. Rev., 114, 788–795.
Bluestein, H. B., and D. Hazen, 1989: Doppler-radar analysis of a tropical cyclone over land: Hurricane Alicia (1983) in Oklahoma. Mon. Wea. Rev., 117, 2594–2611.
Bluestein, H. B., S. D. Hrebenach, C.-F. Chang, and E. A. Brandes, 1994: Synthetic dual-Doppler analysis of mesoscale convective systems. Mon. Wea. Rev., 122, 2105–2124.
Brandes, E. A., 1978: Mesocyclone evolution and tornadogenesis: Some observations. Mon. Wea. Rev., 106, 995–1011.
Brandes, E. A., 1984: Relationships between radar-derived thermodynamic variables and tornadogenesis. Mon. Wea. Rev., 112, 1033–1052.
Bunkers, M. J., B. A. Klimowski, J. W. Zeitler, R. L. Thompson, and M. L. Weisman, 2000: Predicting supercell motion using a new hodograph technique. Wea. Forecasting, 15, 61–79.
Bunkers, M. J., M. R. Hjelmfelt, and P. L. Smith, 2006: An observational examination of long-lived supercells. Part I: Characteristics, evolution, and demise. Wea. Forecasting, 21, 673–688.
Cai, H., and R. M. Wakimoto, 2001: Retrieved pressure field and its influence on the propagation of a supercell thunderstorm. Mon. Wea. Rev., 129, 2695–2713.
Cressman, G. P., 1959: An operational objective analysis system. Mon. Wea. Rev., 87, 367–374.
Crook, A., 1994: Numerical simulations initialized with radar-derived winds. Part I: Simulated data experiments. Mon. Wea. Rev., 122, 1189–1203.
Davies-Jones, R. P., 1984: Streamwise vorticity: The origin of updraft rotation in supercell storms. J. Atmos. Sci., 41, 2991–3006.
Davies-Jones, R. P., and H. E. Brooks, 1993: Mesocyclogenesis from a theoretical perspective. The Tornado: Its Structure, Dynamics, Prediction, and Hazards,Geophys. Monogr., Vol. 79, Amer. Geophys. Union, 105–114.
Davies-Jones, R. P., R. J. Trapp, and H. B. Bluestein, 2001: Tornadoes and tornadic storms. Severe Convective Storms, Meteor. Monogr., No. 50, Amer. Meteor. Soc., 167–221.
Doviak, R. J., P. S. Ray, R. G. Strauch, and L. J. Miller, 1976: Error estimation in wind fields from dual Doppler radar measurements. J. Appl. Meteor., 15, 868–878.
Dowell, D. C., and H. B. Bluestein, 1997: The Arcadia, Oklahoma, storm of 17 May 1981: Analysis of a supercell during tornadogenesis. Mon. Wea. Rev., 125, 2562–2582.
Dowell, D. C., F. Zhang, L. J. Wicker, C. Snyder, and N. A. Crook, 2004: Wind and temperature retrievals in the 17 May 1981 Arcadia, Oklahoma, supercell: Ensemble Kalman filter experiments. Mon. Wea. Rev., 132, 1982–2005.
Droegemeier, K. K., S. M. Lazarus, and R. Davies-Jones, 1993: The influence of helicity on numerically simulated convective storms. Mon. Wea. Rev., 121, 2005–2029.
Eastin, M. D., and M. C. Link, 2009: Miniature supercells in an offshore outer rainband of Hurricane Ivan (2004). Mon. Wea. Rev., 137, 2081–2104.
Gal-Chen, T., 1978: A method for the initialization of the anelastic equations: Implications for matching models with observations. Mon. Wea. Rev., 106, 587–606.
Gal-Chen, T., and R. A. Kropfli, 1984: Buoyancy and pressure perturbations derived from dual-Doppler radar observations of the planetary boundary layer: Applications for matching models with observations. J. Atmos. Sci., 41, 3007–3020.
Gilmore, M. S., and L. J. Wicker, 2002: Influences of the local environment on supercell cloud-to-ground lightning, radar characteristics, and severe weather on 2 June 1995. Mon. Wea. Rev., 130, 2349–2372.
Hane, C. E., and B. C. Scott, 1978: Temperature and pressure perturbations within convective clouds derived from detailed air motion information: Preliminary testing. Mon. Wea. Rev., 106, 654–661.
Hane, C. E., and P. S. Ray, 1985: Pressure and buoyancy fields derived from Doppler radar data in a tornadic thunderstorm. J. Atmos. Sci., 42, 18–35.
Hane, C. E., R. B. Wilhelmson, and T. Gal-Chen, 1981: Retrieval of thermodynamic variables within deep convective clouds: Experiments in three dimensions. Mon. Wea. Rev., 109, 564–576.
Hayes, J. L., 2009: Service assessment: Super Tuesday tornado outbreak of February 5–6, 2008. National Weather Service, 48 pp.
Kirkpatrick, C., E. W. McCaul, and C. Cohen, 2009: Variability of updraft and downdraft characteristics in a large parameter space study of convective storms. Mon. Wea. Rev., 137, 1550–1561.
Klemp, J. B., R. B. Wilhelmson, and P. S. Ray, 1981: Observed and numerically simulated structure of a mature supercell thunderstorm. J. Atmos. Sci., 38, 1558–1580.
Klimowski, B. A., and J. D. Marwitz, 1992: The synthetic dual-Doppler analysis technique. J. Atmos. Oceanic Technol., 9, 728–745.
Klimowski, B. A., M. J. Bunkers, M. R. Hjelmfelt, and J. N. Covert, 2003: Severe convective windstorms over the northern high plains of the United States. Wea. Forecasting, 18, 502–519.
Kraus, M. J., 1974: Doppler radar investigation of flow patterns within thunderstorms. AFCRL Environmental Research Paper 481, Hanscom AFM, MA, AFCRL-TR-74-0290, 86 pp.
Laird, N. F., L. J. Miller, and D. A. R. Kristovich, 2001: Synthetic dual-Doppler analysis of a winter mesoscale vortex. Mon. Wea. Rev., 129, 312–331.
Lemon, L. R., 1976: The flanking line, a severe thunderstorm intensification source. J. Atmos. Sci., 33, 686–694.
Lemon, L. R., and C. A. Doswell, 1979: Severe thunderstorm evolution and mesocyclone structure as related to tornadogenesis. Mon. Wea. Rev., 107, 1184–1197.
Lhermitte, R. M., and L. J. Miller, 1970: Doppler radar methodology for the observations of convective storms. Preprints, 14th Conf. on Radar Meteorology, Tucson, AZ, Amer. Meteor. Soc., 133–138.
Markowski, P. M., 2002: Hook echoes and rear-flank downdrafts: A review. Mon. Wea. Rev., 130, 852–876.
Markowski, P. M., and Y. P. Richardson, 2008: Tornadogenesis: Our current understanding, forecasting considerations, and questions to guide future research. Atmos. Res., 93, 3–10.
Matejka, T., and D. L. Bartels, 1998: The accuracy of vertical air velocities from Doppler radar data. Mon. Wea. Rev., 126, 92–117.
McCaul, E. W., and M. L. Weisman, 1996: The dependence of simulated storm structure on variations in the shapes of environment buoyancy and shear profiles. Preprints, 18th Conf. on Severe Local Storms, San Francisco, CA, Amer. Meteor. Soc., 718–722.
McCaul, E. W., and M. L. Weisman, 2001: The sensitivity of simulated supercell structure and intensity to variations in the shapes of environmental buoyancy and shear profiles. Mon. Wea. Rev., 129, 664–687.
McCaul, E. W., and C. Cohen, 2002: The impact on simulated storm structure and intensity of variations in the mixed layer and moist layer depths. Mon. Wea. Rev., 130, 1722–1748.
Miller, L. J., and W. Anderson, 1991: Multiple Doppler radar wind synthesis in CEDRIC. CEDRIC manual, Appendix F, NCAR, 12 pp.
Miller, L. J., J. D. Tuttle, and G. B. Foote, 1990: Precipitation production in a large Montana hailstorm: Airflow and particle growth trajectories. J. Atmos. Sci., 47, 1619–1646.
Mohr, C. G., L. J. Miller, R. L. Vaughan, and H. W. Frank, 1986: The merger of mesoscale data sets into a common Cartesian format for efficient and systematic analysis. J. Atmos. Oceanic Technol., 3, 143–161.
Nelson, S. P., and N. C. Knight, 1987: The hybrid multicellular–supercellular storm—An efficient hail producer. Part I: An archetypal example. J. Atmos. Sci., 44, 2042–2059.
O’Brien, J. J., 1970: Alternative solutions to the classical vertical velocity problem. J. Appl. Meteor., 9, 197–203.
Oye, R., C. Mueller, and S. Smith, 1995: Software for radar translation, visualization, editing, and interpolation. Preprints, 27th Conf. on Radar Meteorology, Vail, CO, Amer. Meteor. Soc., 359–361.
Pasken, R., and Y. J. Lin, 1982: Pressure perturbations within a tornadic storm derived from dual-Doppler wind data. Preprints, 12th Conf. on Severe Local Storms, San Antonio, TX, Amer. Meteor. Soc., 257–260.
Peace, R. L., R. A. Brown, and H. G. Camnitz, 1969: Horizontal motion field observations with a single pulse Doppler radar. J. Atmos. Sci., 26, 1096–1103.
Ray, P. S., D. P. Jorgensen, and S. L. Wang, 1985: Airborne Doppler radar observations of a convective storm. J. Climate Appl. Meteor., 24, 687–698.
Roberts, R. D., and J. W. Wilson, 1995: The genesis of three nonsupercell tornadoes observed with dual-Doppler radar. Mon. Wea. Rev., 123, 3408–3436.
Rotunno, R., and J. B. Klemp, 1982: The influence of shear-induced pressure gradient on thunderstorm motion. Mon. Wea. Rev., 110, 136–151.
Rotunno, R., and J. B. Klemp, 1985: On the rotation and propagation of simulated supercell thunderstorms. J. Atmos. Sci., 42, 271–292.
Roux, F., 1985: Retrieval of thermodynamic fields from multiple-Doppler radar data using the equations of motion and thermodynamic equation. Mon. Wea. Rev., 113, 2142–2157.
Steiger, S. M., R. E. Orville, and L. D. Carey, 2007: Total lightning signatures of thunderstorm intensity over North Texas. Part I: Supercells. Mon. Wea. Rev., 135, 3281–3302.
Thompson, R. L., 1998: Eta model storm-relative winds associated with tornadic and nontornadic supercells. Wea. Forecasting, 13, 125–137.
Trapp, R. J., and R. Davies-Jones, 1997: Tornadogenesis with and without a dynamic-pipe effect. J. Atmos. Sci., 54, 113–133.
Trapp, R. J., E. D. Mitchell, G. A. Tipton, D. W. Effertz, A. I. Watson, D. L. Andra, and M. A. Magsig, 1999: Descending and nondescending tornadic vortex signatures detected by WSR-88Ds. Wea. Forecasting, 14, 625–639.
Weisman, M. L., and J. B. Klemp, 1984: The structure and classification of numerically simulated convective storms in directionally varying wind shears. Mon. Wea. Rev., 112, 2479–2498.
Weisman, M. L., and R. J. Trapp, 2003: Low-level mesovortices within squall lines and bow echoes. Part I: Overview and dependence on environmental shear. Mon. Wea. Rev., 131, 2779–2803.
Wicker, L. J., and L. Cantrell, 1996: The role of vertical buoyancy distributions in miniature supercells. Preprints, 18th Conf. on Severe Local Storms, San Francisco, CA, Amer. Meteor. Soc., 225–229.
Since the writing of this paper, fatalities resulting from the 27 April 2011 tornado outbreak have far surpassed the Super Tuesday outbreak.
