1. Introduction
High-temporal resolution radar observations are crucial for the study of tornadoes because of the rapidly changing nature of tornadoes. Tornadoes can intensify and weaken over short time scales, and many weak tornadoes have entire life cycles that are much shorter than the volumetric update time of traditional, mechanically steered radars. Volumetric update times much less than 1 min are required to properly sample the evolution of tornadoes, with temporal resolution of 10 s or less necessary to capture some of the more rapidly evolving features (Bluestein et al. 2010). Mobile radars currently capable of volumetric updates of ~30 s or less include the rapid-scan Doppler on Wheels (Wurman and Randall 2001), the Mobile Weather Radar 2005 X-band Phased Array (MWR-05XP; Bluestein et al. 2010), the rapid X-band polarimetric radar (RaXPol; Pazmany et al. 2013), and the Atmospheric Imaging Radar (AIR; Isom et al. 2013; Kurdzo et al. 2017). In addition to high-temporal resolution, these radars have the advantage of close-range deployments (Wurman et al. 1996), which allow for much better spatial and low-level sampling of tornadoes and mesocyclones compared to stationary radars (e.g., Wurman and Gill 2000; Bluestein and Pazmany 2000; Bluestein et al. 2007a; Kosiba and Wurman 2013; Wurman and Kosiba 2013; Weiss et al. 2014). The finescale spatial resolution provided by mobile radars maximizes the benefits of high-temporal updates for tornado applications by allowing radars to observe rapidly evolving small-scale phenomena.
Recent tornado studies using rapid-update radars have illustrated the importance of short-lived tornadic subvortices on a tornado’s intensity and damage potential (Snyder and Bluestein 2014; Wakimoto et al. 2016). Rapid-scan radars are also uniquely qualified to interrogate the vertical evolution of rotation during tornadogenesis (Houser et al. 2015; French et al. 2013) and tornado decay (French et al. 2014). Houser et al. (2015) found that subtornadic rotation in the 24 May 2011 El Reno, Oklahoma, EF-5 tornado first developed below 1 km and then intensification of rotation to tornadic intensity occurred simultaneously throughout the column when rotation increased aloft (e.g., above 3 km). During decay, French et al. (2014) found that the 5 June 2009 Goshen County, Wyoming, tornado first dissipated near a height of 1.5 km and then proceeded to decay in an “inside out” manner at progressively higher and lower elevations. The level where dissipation first occurred was found to be just above where the tornado motion was most influenced by strong rear-flank gust front outflow. Houser et al. (2015) also noted a similar decay mechanism in the El Reno tornado with tornado dissipation first occurring in a stable layer between 1.5 and 3 km above the ground.
Rapid-update radars have observed that tornado intensity can be highly variable, sometimes periodic, over short time scales (e.g., Wurman and Kosiba 2013) and that tornado intensity in the low levels (e.g., below 1 km) may be uncorrelated from tornado intensity aloft (French et al. 2014; Mahre et al. 2018). Additional topics of recent investigation using high-temporal resolution radar data include the impact that rapidly evolving storm-scale processes, such as rear-flank gust front surges, have on the tornado life cycle (Houser et al. 2015) and the rapid evolution of the distribution of tornadic debris (Kurdzo et al. 2015; Houser et al. 2016; Mahre et al. 2018). However, the number of high-temporal resolution tornado cases is still extremely limited, and thus it remains unclear whether or not these past findings can be generalized. A range of tornado cases of varying tornado size and intensity are needed to better understand the diversity of tornado behavior and document whether similarities exist in tornadogenesis and tornado dissipation modes.
Toward the goal of a broader view of rapid tornado evolution, the present study interrogates close-range tornado observations collected by the AIR. The AIR is a mobile, X-band phased-array radar that collects high-temporal resolution, vertically continuous volumetric observations of tornadoes. On 27 May 2015, the AIR collected 5.5-s temporal resolution observations of an EF-2 rated tornado near Canadian, Texas, with maximum Doppler velocities of 65 m s−1 during the scanning period and at its widest, had a radius of maximum winds (RMW) of 500 m. The nearly simultaneous volumetric data collection within the Canadian tornado provides a unique opportunity to interrogate the vertical structure of a tornado during decay without having to account for translation or evolution of the vortex between elevation scans.
This study documents the vertical evolution of tornado intensity during tornado decay with the goal of identifying if specific layers within the tornado are important for triggering tornado dissipation. Studies by French et al. (2014) and Houser et al. (2015) have previously documented tornado decay with high-temporal resolution and have provided early insight into the vertical behavior of tornadoes during dissipation. However, the data in this case have better spatial resolution compared to French et al. (2014) and better low-level coverage than Houser et al. (2015). Understanding where tornadoes first weaken and how decay evolves in time and space could inform future investigations into the vorticity budget of the tornado using models, multiple-Doppler analyses, or a combination of both (e.g., Marquis et al. 2012, 2016). In addition to contributing to a better understanding of why tornadoes dissipate, rapid observations that determine how late-life cycle tornadoes behave have the potential to inform forecasters when to let warnings expire if phased-array radars are implemented operationally in the future.
In addition to investigating tornado decay, this study also documents two instances of small-scale vortices being shed from the parent tornado, which occur during a brief tornado weakening period and may have contributed to the transport of angular momentum away from the tornado. A detailed evolution of tornado tilt through multiple layers of the tornado is presented. French et al. (2014) previously found no relationship between tornado tilt and vortex intensity; however, this study calculates tornado tilt with much finer vertical resolution in order to investigate much smaller layers within the tornado rather than just the overall tornado inclination. Finally, tornado intensity is compared to vortex diameter to interrogate the role of conservation of angular momentum on tornado intensity and the vertical cross correlation of tornado intensity is analyzed.
2. Methods
a. Atmospheric imaging radar and case overview
The AIR is a mobile, rapid-scan, X-band, phased-array imaging radar designed and constructed by the Advanced Radar Research Center (Isom et al. 2013). The AIR transmits a 20° in elevation by 1° in azimuth horizontally polarized fan beam and receives data using a 36-element array with a vertically oriented baseline. The 36 channels of received data are combined using digital beamforming (Mead et al. 1998) to reconstruct 20 (total) 1° × 1° beams resulting in a simultaneous RHI with each pulse, which is mechanically steered in azimuth to collect volumetric data. In the present study, Fourier (nonadaptive) digital beamforming is used.
The 27 May 2015 Canadian, Texas, EF-2 tornado had a life span of approximately 10 min, forming around 2157 UTC and visually dissipating at 2207 UTC. Beginning at 2203 UTC, the AIR collected approximately 4.5 min of data in 20° by 80° volumetric sectors with 5.5-s temporal resolution. The range to the tornado was between 5 and 6 km for the deployment with the closest distance occurring at the beginning of the sampling period. Corresponding beam heights range from 50 to 2000 m above radar level (ARL) for the center of the beams between 0° and 20° elevation. The presented data are oversampled to 0.5° in both azimuth and elevation and have a 37.5-m gate spacing in range that is oversampled to 30 m (Kurdzo et al. 2014). For this case, the Nyquist velocity was 25 m s−1 and the radar truck was level during data collection. Velocity aliasing was manually corrected during post processing. High spectrum width, low reflectivity data points were subjectively removed if their Doppler velocity looked suspect for all permutations of dealiasing. Spectrum width was used to identify large gradients in Doppler velocity and aid in the identification of the zero isodop. However, in some instances ambiguities in the placement of zero isodop exist, especially regarding the distinction between folded outbound velocities and weak inbound velocities in the core of the tornado. For times with particularly high uncertainty, the position of the zero isodop was selected to best preserve temporal and vertical continuity. Nevertheless, changing the location of the zero isodop a few radials in either direction would have little or no impact on the presented results because of the displacement of the strongest Doppler velocities from the center of the tornado.
b. Time–height plots
The 5.5-s volumetric updates of the AIR can be leveraged to interrogate the vertical evolution of tornado characteristics. Time–height plots of differential velocity 
3. Results
a. High-temporal resolution observations
The maximum sampled intensity of the Canadian, Texas, tornado occurred at 2203:52 UTC, in the first volume of data collected by the AIR (Figs. 1a,b). At this time, the tornado had maximum 
PPI plots of (a),(c),(e),(g) reflectivity (dBZ) and (b),(d),(f),(h) radial velocity (m s−1) at 1° elevation at (a),(b) 2203:52; (c),(d) 2204:13; (e),(f) 2204:56; and (g),(h) 2205:18 UTC. White pixels here and elsewhere represent censored data.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
PPI plots of (a),(c),(e),(g) reflectivity (dBZ) and (b),(d),(f),(h) radial velocity (m s−1) at 1° elevation at (a),(b) 2203:52; (c),(d) 2204:13; (e),(f) 2204:56; and (g),(h) 2205:18 UTC. White pixels here and elsewhere represent censored data.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
PPI plots of (a),(c),(e),(g) reflectivity (dBZ) and (b),(d),(f),(h) radial velocity (m s−1) at 1° elevation at (a),(b) 2203:52; (c),(d) 2204:13; (e),(f) 2204:56; and (g),(h) 2205:18 UTC. White pixels here and elsewhere represent censored data.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
PPI plots of (a),(c),(e),(g) reflectivity (dBZ) and (b),(d),(f),(h) radial velocity (m s−1) at (a),(b) 1°; (c),(d) 4°; (e),(f) 7°; and (g),(h) 10° elevation valid at 2203:52 UTC. The elevations roughly correspond with 100, 400, 750, and 1000 m ARL.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
PPI plots of (a),(c),(e),(g) reflectivity (dBZ) and (b),(d),(f),(h) radial velocity (m s−1) at (a),(b) 1°; (c),(d) 4°; (e),(f) 7°; and (g),(h) 10° elevation valid at 2203:52 UTC. The elevations roughly correspond with 100, 400, 750, and 1000 m ARL.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
PPI plots of (a),(c),(e),(g) reflectivity (dBZ) and (b),(d),(f),(h) radial velocity (m s−1) at (a),(b) 1°; (c),(d) 4°; (e),(f) 7°; and (g),(h) 10° elevation valid at 2203:52 UTC. The elevations roughly correspond with 100, 400, 750, and 1000 m ARL.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
As noted by Kurdzo et al. (2017), an interesting feature of the Canadian tornado is the weak radial velocities in the core of the tornado (Figs. 2d,f) and large separation of the maximum inbound and maximum outbound radial velocities at early times, especially between 3° and 10° in elevation (400–1000 m ARL). A similar pattern of large separation between maximum inbound and outbound radial velocities with weak core Doppler velocities was observed by RaXPol in Wakimoto et al. (2016) when the 2013 El Reno tornado was exhibiting multiple vortex structure [cf. Fig. 15 in Wakimoto et al. (2016)]. However, in the Canadian case these observations were seen at a beam height of ~400–600 m compared to observations in Wakimoto et al. (2016), which were at 0.5° (below 100 m). At 1° elevation, the Canadian tornado instead exhibited a narrower core diameter1 of ~150 m (Fig. 2b). It must be noted that the AIR and RaXPol are both low-power radars, which may have less sensitivity in areas of lower power, like the WEH. While the power returns in the WEH of the Canadian tornado were well above the noise floor of the AIR, to the authors’ knowledge, the weak Doppler velocity structure in the core of tornadoes has not been observed by a higher-power radar; therefore, it is possible that the these observations are a radar artifact unique to low-power radars and not representative of the actual wind field.
Up to approximately 900 m ARL, the tornado widened to 600 m in diameter and weakened in intensity (Figs. 2d,f). Similar vertical structure to what is seen in the Canadian tornado can also be seen in model simulations of medium-swirl tornadoes with a narrow diameter near the surface and a larger diameter aloft [e.g., cf. Fig. 5 in Lewellen et al. (2000)]. In the Lewellen et al. (2000) simulations, this widening of the vortex is associated with a transition from single-cell to two-cell vortex flow. Axisymmetric wind retrievals of the Canadian tornado reveal a similar low-swirl corner flow with vortex breakdown characteristics aloft at this time; however, uncertainties in the retrieved vertical velocities arising from poor sampling of the low-level mass field preclude the inclusion of these analyses. Above ~900 m, the tornado narrowed again (Fig. 2h) and maintained a similar 350-m diameter through the highest observed elevations (not shown). The WEH narrowed in response to the tightening of the vortex aloft (Fig. 2g) and is similar in structure to the bulging WEC structure noted in Bluestein et al. (2007b). The bulging structure of the vortex and WEH continued for approximately 30 s and is further discussed in section 3c.
A broken ring of high (>10 m s−1) spectrum width was also observed at early times near the radius where the largest radial velocities were located (Fig. 3a). Similar spectrum width ring structure has also been observed by Wurman (2002) and Snyder and Bluestein (2014) in multiple-vortex tornadoes. Locally enhanced spectrum width near tornadic subvortices was likely due to high wind shear and turbulence. Additionally, strong radial gradients in tangential velocity are one requirement for subvortices to form (e.g., Rotunno 1978; Gall 1983; Lewellen et al. 2000), which would result in high spectrum width rings in an axisymmetric framework. Video of the Canadian tornado is available at https://drive.google.com/drive/folders/1sG9PEEBk9obq7cGv6vqlqffyj5yb4d1S?usp=sharing) and enhanced shear regions along the zero isodop in Fig. 2b also indicate that multiple vortices were present near this time.
PPI plots of (a),(c) spectrum width (m s−1) and (b),(d) radial velocity (m s−1), at (a),(b) 1.5° and (c),(d) 5° elevation valid at 2204:13 UTC.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
PPI plots of (a),(c) spectrum width (m s−1) and (b),(d) radial velocity (m s−1), at (a),(b) 1.5° and (c),(d) 5° elevation valid at 2204:13 UTC.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
PPI plots of (a),(c) spectrum width (m s−1) and (b),(d) radial velocity (m s−1), at (a),(b) 1.5° and (c),(d) 5° elevation valid at 2204:13 UTC.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
The AIR captured at least two instances of cyclonic vortices being shed by the main circulation in rapid succession over a period of one minute (Fig. 4). The vortices first appeared on the eastern side of the tornado near the RMW and translated eastward, well beyond the RMW of the tornado where tornado subvortices are normally located (e.g., Wurman 2002). After being shed, the vortices went through a brief intensification period and reached maximum 
PPI plots of radial velocity (m s−1) at (a) 2204:02, (b) 2204:13, (c) 2204:24, (d) 2204:35, (e) 2204:45, (f) 2204:56, (g) 2205:07, and (h) 2205:18 UTC at 3.5° elevation. Arrows indicate the location of the shed vortices. Maximum 
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
PPI plots of radial velocity (m s−1) at (a) 2204:02, (b) 2204:13, (c) 2204:24, (d) 2204:35, (e) 2204:45, (f) 2204:56, (g) 2205:07, and (h) 2205:18 UTC at 3.5° elevation. Arrows indicate the location of the shed vortices. Maximum
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
PPI plots of radial velocity (m s−1) at (a) 2204:02, (b) 2204:13, (c) 2204:24, (d) 2204:35, (e) 2204:45, (f) 2204:56, (g) 2205:07, and (h) 2205:18 UTC at 3.5° elevation. Arrows indicate the location of the shed vortices. Maximum
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
The maximum 
Time–height plot of maximum 
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Time–height plot of maximum
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Time–height plot of maximum
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Following the brief restrengthening of the tornado at low levels, the Canadian tornado began a final weakening phase that culminated in tornado dissipation (Fig. 5). At all observed levels above 1.25 km ARL, 
After a minute of marginal tornado intensity, the vortex decayed beginning around 2207:27 UTC in the lowest 1.25 km ARL while it maintained 
b. Tornado tilt
The Canadian, Texas, tornado became increasingly tilted through all depths beginning at 2205:39 UTC and lasting through tornado dissipation (Fig. 6a). The increase in tornado tilt corresponded with the initial stage of tornado decay in Fig. 5. Increase in tornado tilt was maximized in the 0.1–0.5- and 0.1–1.0-km layers, where tilt increased from approximately 25° to 55° and from 35° to 55°, respectively. The increase in tilt was the result of the low-level vortex becoming increasingly displaced to the west of the midlevel vortex (Fig. 6b). While the Canadian tornado at the lowest elevation angle (black) moved steadily westward, higher up, between 1 and 2 km ARL (red), the Canadian tornado looped cyclonically and moved only slightly westward during dissipation. This tornado motion likely resulted in the tornado becoming occluded as the low levels became displaced to the west of the mesocyclone above 1 km as well as to the west of the position of the rear-flank gust front (not shown).
(a) Tornado tilt (°) between the lowest elevation angle (~0.1 km) and 0.5 km (blue), 1.0 km (orange), 1.5 km (green), and 2.0 km (red). (b) Tornado center at the lowest elevation angle (black) and mean tornado center in the 0.1–1.0-km layer (blue) and 1.0–2.0-km layer (red). Storm motion is from right to left at all levels.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
(a) Tornado tilt (°) between the lowest elevation angle (~0.1 km) and 0.5 km (blue), 1.0 km (orange), 1.5 km (green), and 2.0 km (red). (b) Tornado center at the lowest elevation angle (black) and mean tornado center in the 0.1–1.0-km layer (blue) and 1.0–2.0-km layer (red). Storm motion is from right to left at all levels.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
(a) Tornado tilt (°) between the lowest elevation angle (~0.1 km) and 0.5 km (blue), 1.0 km (orange), 1.5 km (green), and 2.0 km (red). (b) Tornado center at the lowest elevation angle (black) and mean tornado center in the 0.1–1.0-km layer (blue) and 1.0–2.0-km layer (red). Storm motion is from right to left at all levels.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
The direction of tilt in the 0.1–2.0 km ARL (0.5°–20°) layer was to the northeast for the entirety of the deployment (Fig. 7), which is similar to the 0–3-km environmental wind shear vector (not shown). Similar northeastward tilt has been noted previously (e.g., Tanamachi et al. 2012, 2013; French et al. 2014) and appears to be a common orientation for supercellular tornadoes in the Northern Hemisphere. The tilt direction changed from toward the NNE to toward the ENE during the tornado decay period. This clockwise shift in tilt direction corresponded with the westward motion of the low-level vortex relative to the upper levels of the vortex (Fig. 6b). The largest change in tilt direction started at 2205:39 UTC (Fig. 7), when the first tornado decay period began (Fig. 5) and the aforementioned increase in tilt occurred.
Direction of vortex tilt between 0.1 and 2.0 km ARL in clockwise degrees from north. Radial axis is time in UTC.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Direction of vortex tilt between 0.1 and 2.0 km ARL in clockwise degrees from north. Radial axis is time in UTC.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Direction of vortex tilt between 0.1 and 2.0 km ARL in clockwise degrees from north. Radial axis is time in UTC.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
The simultaneous RHIs collected by the AIR also allows for the interrogation of the vertical shape of the tornado in greater detail (Fig. 8). Early on in the deployment when the tornado was at its observed maximum diameter and intensity, the tornado was nearly upright with minimal vortex tilt below the cloud base (Figs. 8a,d). As previously mentioned, the tornado became increasingly tilted late in its life cycle (Figs. 8b,d). A rapid increase in tilt was observed between 400 and 700 m ARL during the 2206:01–2206:44 UTC period (Figs. 8c,d). The formation of the region of vortex tilt in excess of 70° corresponded temporally with the downward tornado decay observed in Fig. 5 and was located at the bottom of the of the vertical layer where downward decay occurred. Interestingly, the layer with the largest tilt, between 400 and 700 m ARL, exhibited the slowest decrease in 
Photographs at (a) 2204, (b) 2206, and (c) 2207 UTC and (d) a time–height plot of filtered vortex tilt (°). Vertical lines i, ii, and iii in (d) correspond with (a)–(c), respectively. Photographs adapted from Kurdzo et al. (2017).
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Photographs at (a) 2204, (b) 2206, and (c) 2207 UTC and (d) a time–height plot of filtered vortex tilt (°). Vertical lines i, ii, and iii in (d) correspond with (a)–(c), respectively. Photographs adapted from Kurdzo et al. (2017).
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Photographs at (a) 2204, (b) 2206, and (c) 2207 UTC and (d) a time–height plot of filtered vortex tilt (°). Vertical lines i, ii, and iii in (d) correspond with (a)–(c), respectively. Photographs adapted from Kurdzo et al. (2017).
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
c. Vertical cross sections
A simultaneously collected RHI through the concentric WEHs (Fig. 9a) reveals that the inner bands of enhanced reflectivity were disconnected from any regions of higher reflectivity aloft and extended below the lowest observed elevation (Fig. 9b). Conversely, the outer bounds of the larger WEH were part of a continuous region of high reflectivity that extended through 20° in elevation. While the outer WEC extended higher than the 750 m illustrated in Fig. 9b, the tornado tilts out of the RHI plane, precluding observation of the feature through a deeper layer. Regardless, the contrasting vertical structure of the WECs in the low levels supports the hypothesis that the two WECs represented different types of scatterers. It is speculated that the inner WEH comprised light debris that were lofted from the surface and the outer WEH represented descending rainbands at the periphery of the tornado. However, in the absence of polarimetric data, this hypothesis cannot be confirmed.
(a) PPI plot of reflectivity (dBZ) at 1° elevation and (b) RHI of reflectivity (dBZ) valid at 2203:52 UTC. Black line in (a) represents the location of the RHI in (b).
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
(a) PPI plot of reflectivity (dBZ) at 1° elevation and (b) RHI of reflectivity (dBZ) valid at 2203:52 UTC. Black line in (a) represents the location of the RHI in (b).
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
(a) PPI plot of reflectivity (dBZ) at 1° elevation and (b) RHI of reflectivity (dBZ) valid at 2203:52 UTC. Black line in (a) represents the location of the RHI in (b).
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
To better observe the vertical structure of the WEH, azimuth height indicators (AHIs) consisting of a vertical cross section along a constant range are constructed (Fig. 10a). While these AHIs are not simultaneous, they only represent a span of ~5 s of data, and therefore are comparable to traditional RHIs collected by mechanically steered radars. Because the Canadian tornado was tilted to the northeast (Fig. 7), this geometry better captures a deeper layer of the WEC along the direction of tornado tilt. However, the tapered nature of the WEC above 1.25 km ARL and the disappearance of the WEC above 1.5 km (Fig. 10b) are artifacts of the tornado tilting out of the AHI plane within the persistent tilt layer between 1.25 and 1.5 km ARL. In contrast, the abrupt narrowing of the WEC at 1 km ARL is not an artifact of tornado tilt since the AHI bisects the WEH at its widest point beneath 1 km (Fig. 10c) and just above 1 km (Fig. 10d). While there was no observed vertical gradient in 
(a),(c),(d) PPI plots of reflectivity (dBZ) at 1°, 8°, and 10°, respectively; and (b) AHI of reflectivity (dBZ) valid at 2203:52 UTC. Horizontal lines i, ii, and iii in (b) correspond with (a), (b), and (d), respectively. Black lines in (a), (c), and (d) represent the location of the AHI in (b).
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
(a),(c),(d) PPI plots of reflectivity (dBZ) at 1°, 8°, and 10°, respectively; and (b) AHI of reflectivity (dBZ) valid at 2203:52 UTC. Horizontal lines i, ii, and iii in (b) correspond with (a), (b), and (d), respectively. Black lines in (a), (c), and (d) represent the location of the AHI in (b).
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
(a),(c),(d) PPI plots of reflectivity (dBZ) at 1°, 8°, and 10°, respectively; and (b) AHI of reflectivity (dBZ) valid at 2203:52 UTC. Horizontal lines i, ii, and iii in (b) correspond with (a), (b), and (d), respectively. Black lines in (a), (c), and (d) represent the location of the AHI in (b).
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
As previously mentioned, the Canadian tornado became increasingly tilted late in its life cycle (Fig. 8). A simultaneous RHI was taken through the tornado (Fig. 11) near time ii in Figs. 8b,d. The tilt of the tornado at this time was 45°–50° from vertical in the layer between 400 and 750 m ARL. Differential velocity was computed in both the vertical (Fig. 11b) and horizontal (Figs. 11c,d) planes. As expected from a tornado that was tilted approximately 45° from vertical, the magnitude of vertical shear was similar to the magnitude of horizontal shear with 
RHI plots of (a) radial velocity (m s−1) and (b) reflectivity (dBZ) and PPI plots of radial velocity (m s−1) at (c) 2° and (d) 6.5° elevation valid at 2206:44 UTC. The tornado is very tilted through the 400–700-m layer (see Fig. 8d) as evidenced by the shift in vortex position between (c) and (d). Strong horizontal vorticity is evident through this layer with slightly larger 
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
RHI plots of (a) radial velocity (m s−1) and (b) reflectivity (dBZ) and PPI plots of radial velocity (m s−1) at (c) 2° and (d) 6.5° elevation valid at 2206:44 UTC. The tornado is very tilted through the 400–700-m layer (see Fig. 8d) as evidenced by the shift in vortex position between (c) and (d). Strong horizontal vorticity is evident through this layer with slightly larger
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
RHI plots of (a) radial velocity (m s−1) and (b) reflectivity (dBZ) and PPI plots of radial velocity (m s−1) at (c) 2° and (d) 6.5° elevation valid at 2206:44 UTC. The tornado is very tilted through the 400–700-m layer (see Fig. 8d) as evidenced by the shift in vortex position between (c) and (d). Strong horizontal vorticity is evident through this layer with slightly larger
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
4. Discussion
a. Persistent vortex tilt layer and tornado decay
The persistent region of enhanced vortex tilt between 1.25 and 1.5 km ARL (Fig. 8d) coincided with the level that separated the temporally displaced mid- and low-level vortex decay near the end of the Canadian tornado life cycle (Fig. 5). Additionally, the maximum tornado intensity observed at the beginning of the deployment as well as the brief restrengthening of the tornado that occurred shortly thereafter was confined to within the lowest 1 km (Fig. 5). It is hypothesized from this body of circumstantial evidence that the vertical layer around ~1.25 km ARL was particularly relevant to the Canadian tornado.
The Dodge City, Kansas, 0000 UTC sounding was launched approximately 1 h after the Canadian tornado occurred, and despite the approximately 200 km that separate the locations, the surface conditions at Dodge City (DDC) were nearly identical to those in Canadian, Texas, at the time of the tornado (Fig. 12a). Additionally, environmental temperatures at 700 (Fig. 12b) and 500 mb (1 mb = 1 hPa) (Fig. 12c) were also similar at DDC and Canadian, Texas, and the regional environment was relatively uniform. Thus, it is assumed that the DDC sounding (Fig. 12b) was at least somewhat representative of the environmental thermodynamic profile in Canadian, Texas. The level of free convection (LFC) at DDC was at approximately 760 mb, corresponding to ~1500 m above ground level (AGL) at the sounding site. Replacing the surface conditions of the DDC sounding with the ASOS observations in Canadian, Texas, at the time of the tornado (not shown) lowers the LFC to 800 mb or 1250 m ARL, which is within the layer of enhanced vortex tilt.
Storm Prediction Center mesoanalysis of (a) surface temperature (red lines, °F), mean sea level pressure (black lines, mb), and dewpoint temperature (shaded, °F), (b) 700-mb height (black lines, m), temperature (red lines, °C), and 700–500-mb mean RH (shaded, %), and (c) 500-mb height (black lines, m) and temperature (red lines, °C) valid at 2200 UTC 27 May 2015. The star indicates the location of Canadian, TX, and the circle indicates the location of Dodge City, KS. Archived mesoananalyses are available online at http://www.spc.noaa.gov/exper/ma_archive/. (d) Skew T–logp diagram of the vertical sounding launched at Dodge City, KS, valid at 0000 UTC 28 May 2015. Archived soundings are available online at http://weather.uwyo.edu/upperair/sounding.html.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Storm Prediction Center mesoanalysis of (a) surface temperature (red lines, °F), mean sea level pressure (black lines, mb), and dewpoint temperature (shaded, °F), (b) 700-mb height (black lines, m), temperature (red lines, °C), and 700–500-mb mean RH (shaded, %), and (c) 500-mb height (black lines, m) and temperature (red lines, °C) valid at 2200 UTC 27 May 2015. The star indicates the location of Canadian, TX, and the circle indicates the location of Dodge City, KS. Archived mesoananalyses are available online at http://www.spc.noaa.gov/exper/ma_archive/. (d) Skew T–logp diagram of the vertical sounding launched at Dodge City, KS, valid at 0000 UTC 28 May 2015. Archived soundings are available online at http://weather.uwyo.edu/upperair/sounding.html.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Storm Prediction Center mesoanalysis of (a) surface temperature (red lines, °F), mean sea level pressure (black lines, mb), and dewpoint temperature (shaded, °F), (b) 700-mb height (black lines, m), temperature (red lines, °C), and 700–500-mb mean RH (shaded, %), and (c) 500-mb height (black lines, m) and temperature (red lines, °C) valid at 2200 UTC 27 May 2015. The star indicates the location of Canadian, TX, and the circle indicates the location of Dodge City, KS. Archived mesoananalyses are available online at http://www.spc.noaa.gov/exper/ma_archive/. (d) Skew T–logp diagram of the vertical sounding launched at Dodge City, KS, valid at 0000 UTC 28 May 2015. Archived soundings are available online at http://weather.uwyo.edu/upperair/sounding.html.
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
The transition from forced to free ascent by parcels within the tornado may have caused a change in parcel trajectory and might explain the tilted nature of the vortex in this layer, especially given that the layer between 1 and 2 km exhibited considerable environmental wind shear (Fig. 12d). In this scenario, ascent, which was forced beneath the LFC, would have been slower than above the LFC where parcels were positively buoyant. As a result, more horizontal advection would have occurred beneath the LFC than above it. The vertical gradient in vertical velocity necessary for increased tilt in this layer is similar to what was inferred by French et al. (2014) when discussing the increased vertical vorticity stretching that may occur near the LFC. Differential horizontal advection in this layer would be enhanced if the near-storm environment beneath the LFC were particularly stable. A stable layer would also have contributed to the observed vertical discontinuity in tornado decay similar to what was observed by Houser et al. (2015). Within the stable layer, parcels would not be accelerated upward by the buoyancy force, which in turn weakens vertical gradients in vertical velocity and subsequently limits the stretching of vertical vorticity. If negatively buoyant air were to be ingested into the updraft, within and above a stable layer the vertical gradient in vertical velocity could even become negative, rapidly reducing vortex strength through horizontal divergence. However, a near-storm sounding would be necessary to assess whether a small stable layer was present.
Changes in inflow characteristics associated with tornado occlusion may impact vortex strength above the LFC on a different time scale than below the LFC. The gradual, downward vortex weakening in Fig. 5 might first occur just above the LFC where the buoyancy force plays an important role in accelerating parcels and stretching vorticity. Vortex decay below the LFC may lag midlevel decay because ascent is forced by a dynamic pressure gradient force that is modulated by midlevel vortex strength. This possible explanation would be in agreement with the tornado occlusion mechanism for tornado decay hypothesized in French et al. (2014), where tornadoes first weaken just above the level of occlusion and then dissipate in an “inside out” fashion. Additionally, the 5 June 2009 Goshen County, Wyoming, tornado was observed to intensify much more rapidly above the LFC than below the LFC (French et al. 2013), which further supports the theory that the LFC demarks a boundary in tornado structure and evolution.
While the LFC appears to separate two dissipation modes, there are many additional factors that may contribute to tornado intensity. A comparison of tornado diameter to tornado intensity (Fig. 13) reveals that tornado intensity was anticorrelated with the distance between the location of maximum and minimum Doppler velocity within the tornado. During the brief restrengthening period between 2204:13 and 2205:18 UTC, tornado diameter decreased from approximately 700 m to between 100 and 200 m. The tightening of the vortex was confined to the lowest 750 m ARL, the same depth over which tornado intensity increased. Conversely, the near-simultaneous weakening of the tornado above 1250 m ARL at 2205:28 UTC occurred at the same time vortex diameter doubled from ~300–400 to ~700–800 m within the same vertical layer. Tornado intensity at this time approximately halved from 
Time–height plots of (a) maximum 
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Time–height plots of (a) maximum
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Time–height plots of (a) maximum
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
b. Vertical correlation in vortex intensity
Differential velocity (
Cross correlation at lag 0 of 
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Cross correlation at lag 0 of
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
Cross correlation at lag 0 of
Citation: Monthly Weather Review 147, 3; 10.1175/MWR-D-18-0297.1
The high correlation of vortex intensity at all heights with the low-level vortex in the Canadian tornado contrasts with other recent high-temporal observations of tornadoes. French et al. (2014) noted that low-level vortex intensity in the 5 June 2009 Goshen County, Wyoming, EF2 tornado was often anticorrelated with midlevel intensity. This was largely due to the oscillatory nature of low-level vortex strength. However, French et al. (2014) were comparing average tornado intensity below 2 km with average intensity above 2 km, which sometimes extended above 5 km ARL, whereas all of the observations in the Canadian case were below 2 km. Similar to the Goshen County tornado, the 16 May 2015 Tipton, Oklahoma, tornado also exhibited periodic changes in low-level vortex intensity, which often decorrelated vortex intensity in the lowest 500–700 m from vortex intensity above 700 m (Mahre et al. 2018). Periodic changes in low-level tornado intensity have also been observed by Wurman et al. (2013) using high-temporal resolution mobile radar data. Additionally, the Tipton tornado experienced one instance of pronounced upward tornado intensification, although this may be due to debris-induced biases (Mahre et al. 2018). The Canadian tornado did not exhibit strong periodicity in low-level vortex intensity during the observed period; however, these observations were during the late-mature to dissipation stages whereas the Tipton observations were during the mature stage, so caution is required when comparing the two cases. Perhaps there is some mechanism for periodic oscillations in tornado intensity that is less likely to occur late in the tornado life cycle (e.g., because the tornado and mesocyclone are less vertically stacked). However, it can be concluded that during the late mature and decay phases of the Canadian tornado, changes in vortex intensity in the lowest 2 km were largely nonperiodic and were highly correlated with tornado intensity at the lowest analysis level.
5. Conclusions
On 27 May 2015, the AIR collected 4.5 min of sector volumes with 5.5-s temporal resolution during the late-mature and decay stages of the Canadian, Texas, tornado. At the beginning of the deployment, maximum 
A weakening of the vortex occurred soon after the beginning of the dataset with 
The decay of the vortex first occurred simultaneously in the column above 1.25 km ARL and dissipation progressed in a downward direction over the next minute. An increase in tornado diameter occurred simultaneously with the observed decreases in 
Apart from the brief upward intensification and downward decay periods, vortex intensity throughout the analyzed depth was highly correlated with intensity at the lowest analysis level and changes in intensity occurred nearly simultaneously in the vertical. The largely simultaneous vertical changes in vortex intensity were similar to what was observed in the 24 May 2011 El Reno, Oklahoma, tornado (Houser et al. 2015), but dissimilar to the anticorrelated low-level and midlevel tornado strength exhibited by the 5 June 2009 Goshen County, Wyoming, tornado (French et al. 2014) and the 16 May 2015 Tipton, Oklahoma, tornado (Mahre et al. 2018).
During decay, the Canadian tornado also became increasingly tilted in the low levels. The increase in tilt is the result of the low levels of the tornado becoming increasingly displaced to the west of the vortex above 1 km, which potentially resulted in the tornado becoming occluded. Additionally, a persistent region of enhanced tilt was noted between 1.25 and 1.5 km ARL throughout the entire deployment. This height was the same that separated the midlevel and low-level decay at the end of the Canadian tornado life cycle, and roughly corresponds with the level of free convection, which may help explain both the enhanced tilt and early vortex decay in this region.
While the Canadian tornado is only one case in a growing list of high-temporal resolution tornado observations, it illustrates how rapidly tornado structure can change. Additionally, when considered alongside other recent high-temporal tornado datasets, the Canadian case also exemplifies the diversity of tornado behavior that exists within even a small sample of cases. While similarities are noted between the behavior of the Canadian tornado and tornadoes in other high-temporal datasets, many differences between the tornadoes were highlighted in this case. More high-temporal observations are needed in order to properly document the range of tornado behavior at short time scales and begin to understand the mechanisms that drive different rapid tornado evolutionary modes. It is our hope that a C-band, dual-polarimetric version of the AIR currently under development (PAIR; Salazar-Cerreño et al. 2017) will facilitate future high-temporal resolution observations of tornadoes in addition to providing information on the evolution of tornadic debris and hydrometeor distributions in supercells.
Acknowledgments
This study was supported by the National Science Foundation under Grants AGS-1303685 and AGS-1823478. The authors thank Javier Lujan and Andrew Byrd for continued assistance with AIR fieldwork, as well as Feng Nai for help processing the data. We thank the ARRC engineering staff for continued upkeep of the AIR. The authors would also like to thank David Reimer for providing the supplemental video used as a reference in our analyses. Finally, the authors would like to thank the three anonymous reviewers for their contributions to improving all aspects of this paper.
REFERENCES
Alexander, C., and J. Wurman, 2008: Updated mobile radar climatology of supercell tornado structure and dynamics. 24th Conf. on Severe Local Storms, Savannah, GA, Amer. Meteor. Soc., 19.4, https://ams.confex.com/ams/24SLS/techprogram/paper_141821.htm.
Atkins, N. T., A. McGee, R. Ducharme, R. M. Wakimoto, and J. Wurman, 2012: The LaGrange tornado during VORTEX2. Part II: Photogrammetric analysis of the tornado combined with dual-Doppler radar data. Mon. Wea. Rev., 140, 2939–2958, https://doi.org/10.1175/MWR-D-11-00285.1.
Bluestein, H. B., and A. 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, https://doi.org/10.1175/1520-0477(2000)081<2939:OOTAOC>2.3.CO;2.
Bluestein, H. B., W.-C. Lee, M. Bell, C. C. Weiss, and A. L. Pazmany, 2003: Mobile Doppler radar observations of a tornado in a supercell near Bassett, Nebraska, on 5 June 1999. Part II: Tornado-vortex structure. Mon. Wea. Rev., 131, 2968–2984, https://doi.org/10.1175/1520-0493(2003)131<2968:MDROOA>2.0.CO;2.
Bluestein, H. B., M. M. French, R. L. Tanamachi, S. Frasier, K. Hardwick, F. Junyent, and A. Pazmany, 2007a: Close-range observations of tornadoes in supercells made with a dual-polarization, X-band, mobile Doppler radar. Mon. Wea. Rev., 135, 1522–1543, https://doi.org/10.1175/MWR3349.1.
Bluestein, H. B., C. C. Weiss, M. M. French, E. M. Holthaus, R. L. Tanamachi, S. Frasier, and A. L. Pazmany, 2007b: The structure of tornadoes near Attica, Kansas on 12 May 2004: High-resolution, mobile, Doppler radar observations. Mon. Wea. Rev., 135, 475–506, https://doi.org/10.1175/MWR3295.1.
Bluestein, H. B., M. French, I. PopStefanija, R. Bluth, and J. Knorr, 2010: A mobile, phased-array Doppler radar for the study of severe convective storms. Bull. Amer. Meteor. Soc., 91, 579–600, https://doi.org/10.1175/2009BAMS2914.1.
Dowell, D. C., C. R. Alexander, J. M. Wurman, and L. J. Wicker, 2005: Centrifuging of hydrometeors and debris in tornadoes: Radar-reflectivity patterns and wind-measurement errors. Mon. Wea. Rev., 133, 1501–1524, https://doi.org/10.1175/MWR2934.1.
French, M., H. Bluestein, I. PopStefanija, C. Baldi, and R. Bluth, 2013: Reexamining the vertical development of tornadic vortex signatures in supercells. Mon. Wea. Rev., 141, 4576–4601, https://doi.org/10.1175/MWR-D-12-00315.1.
French, M., H. Bluestein, I. PopStefanija, C. Baldi, and R. Bluth, 2014: Mobile, phased-array, Doppler radar observations of tornadoes at X band. Mon. Wea. Rev., 142, 1010–1036, https://doi.org/10.1175/MWR-D-13-00101.1.
Gall, R. L., 1983: A linear analysis of the multiple vortex phenomenon in simulated tornadoes. J. Atmos. Sci., 40, 2010–2024, https://doi.org/10.1175/1520-0469(1983)040<2010:ALAOTM>2.0.CO;2.
Houser, J., H. Bluestein, and J. Snyder, 2015: Rapid-scan, polarimetric, Doppler radar observations of tornadogenesis and tornado dissipation in a tornadic supercell: The “El Reno, Oklahoma” storm of 24 May 2011. Mon. Wea. Rev., 143, 2685–2710, https://doi.org/10.1175/MWR-D-14-00253.1.
Houser, J. L., H. B. Bluestein, and J. Snyder, 2016: A finescale radar examination of the tornadic debris signature and weak-echo reflectivity band associated with a large, violent tornado. Mon. Wea. Rev., 144, 4101–4130, https://doi.org/10.1175/MWR-D-15-0408.1.
Isom, B., and Coauthors, 2013: The Atmospheric Imaging Radar: Simultaneous volumetric observations using a phased array weather radar. J. Atmos. Oceanic Technol., 30, 655–675, https://doi.org/10.1175/JTECH-D-12-00063.1.
Kosiba, K. A., and J. M. Wurman, 2013: The three-dimensional structure and evolution of a tornado boundary layer. Wea. Forecasting, 28, 1552–1561, https://doi.org/10.1175/WAF-D-13-00070.1.
Kurdzo, J. M., B. L. Cheong, R. D. Palmer, G. Zhang, and J. Meier, 2014: A pulse compression waveform for improved-sensitivity weather radar observations. J. Atmos. Oceanic Technol., 31, 2713–2731, https://doi.org/10.1175/JTECH-D-13-00021.1.
Kurdzo, J. M., D. J. Bodine, B. L. Cheong, and R. D. Palmer, 2015: High-temporal resolution polarimetric X-band Doppler radar observations of the 20 May 2013 Moore, Oklahoma, tornado. Mon. Wea. Rev., 143, 2711–2735, https://doi.org/10.1175/MWR-D-14-00357.1.
Kurdzo, J. M., and Coauthors, 2017: Observations of severe local storms and tornadoes with the Atmospheric Imaging Radar. Bull. Amer. Meteor. Soc., 98, 915–935, https://doi.org/10.1175/BAMS-D-15-00266.1.
Lewellen, D. C., W. S. Lewellen, and J. Xia, 2000: The influence of a local swirl ratio on tornado intensification near the surface. J. Atmos. Sci., 57, 527–544, https://doi.org/10.1175/1520-0469(2000)057<0527:TIOALS>2.0.CO;2.
Mahre, A., J. M. Kurdzo, D. J. Bodine, C. B. Griffin, R. D. Palmer, and T.-Y. Yu, 2018: Analysis of the 16 May 2015 Tipton, Oklahoma, EF-3 tornado at high spatiotemporal resolution using the Atmospheric Imaging Radar. Mon. Wea. Rev., 146, 2103–2124, https://doi.org/10.1175/MWR-D-17-0256.1.
Marquis, J., Y. Richardson, P. Markowski, D. Dowell, and J. Wurman, 2012: Tornado maintenance investigated with high-resolution dual-Doppler and EnKF analysis. Mon. Wea. Rev., 140, 3–27, https://doi.org/10.1175/MWR-D-11-00025.1.
Marquis, J., Y. Richardson, P. Markowski, J. Wurman, K. Kosiba, and P. Robinson, 2016: An investigation of the Goshen County, Wyoming, tornadic supercell of 5 June 2009 using EnKF assimilation of mobile mesonet and radar observations collected during VORTEX2. Part II: Mesocyclone-scale processes affecting tornado formation, mainenance, and decay. Mon. Wea. Rev., 144, 3441–3463, https://doi.org/10.1175/MWR-D-15-0411.1.
Mead, J., G. Hopcraft, S. J. Frasier, B. D. Pollar, C. D. Cherry, D. H. Schaubert, and R. E. McIntosh, 1998: A volume-imaging radar wind profiler for atmospheric boundary layer turbulence studies. J. Atmos. Oceanic Technol., 15, 849–859, https://doi.org/10.1175/1520-0426(1998)015<0849:AVIRWP>2.0.CO;2.
Pazmany, A., J. Mead, H. Bluestein, J. Snyder, and J. Houser, 2013: A mobile rapid-scanning X-band polarimetric (RaXPol) Doppler radar system. J. Atmos. Oceanic Technol., 30, 1398–1413, https://doi.org/10.1175/JTECH-D-12-00166.1.
Rasmussen, E., and J. Straka, 2007: Evolution of low-level angular momentum in the 2 June 1995 Dimmitt, Texas, tornado cyclone. J. Atmos. Sci., 64, 1365–1378, https://doi.org/10.1175/JAS3829.1.
Rotunno, R., 1978: A note on the stability of a cylindrical vortex sheet. J. Fluid Mech., 87, 761–771, https://doi.org/10.1017/S0022112078001871.
Salazar-Cerreño, J. L., and Coauthors, 2017: Development of a mobile C-band polarimetric atmospheric imaging radar (PAIR). Special Symp. on Meteorological Observations and Instrumentation, Seattle, WA, Amer. Meteor. Soc., 1B.1, https://ams.confex.com/ams/97Annual/webprogram/Paper308655.html.
Snyder, J., and H. Bluestein, 2014: Some considerations for the use of high-resolution mobile radar data in tornado intensity determination. Wea. Forecasting, 29, 799–827, https://doi.org/10.1175/WAF-D-14-00026.1.
Tanamachi, R. L., H. B. Bluestein, W. C. Lee, M. Bell, and A. L. Pazmany, 2007: Ground-based velocity track display (GBVTD) analysis of W-band radar data in a tornado near Stockton, Kansas, on 15 May 1999. Mon. Wea. Rev., 135, 783–800, https://doi.org/10.1175/MWR3325.1.
Tanamachi, R. L., H. B. Bluestein, J. B. Houser, S. J. Frasier, and K. M. Hardwick, 2012: Mobile X-band, polarimetric Doppler radar observations of the 4 May 2007 Greensburg, Kansas, tornadic supercell. Mon. Wea. Rev., 140, 2103–2125, https://doi.org/10.1175/MWR-D-11-00142.1.
Tanamachi, R. L., H. B. Bluestein, M. Xue, W. C. Lee, K. Orzel, S. Frasier, and R. M. Wakimoto, 2013: Near-surface vortex structure in a tornado and in a sub-tornado-strength convective-storm vortex observed by a mobile, W-band radar during VORTEX2. Mon. Wea. Rev., 141, 3661–3690, https://doi.org/10.1175/MWR-D-12-00331.1.
Wakimoto, R. M., and Coauthors, 2016: Aerial damage survey of the 2013 El Reno tornado combined with mobile radar data. Mon. Wea. Rev., 144, 1749–1776, https://doi.org/10.1175/MWR-D-15-0367.1.
Weiss, C. C., T. Cermak, R. Metzger, A. Reinhart, and P. Skinner, 2014: Insights into tornado structure afforded by high-frequency mobile radar. 27th Conf. on Severe Local Storms, Madison, WI, Amer. Meteor. Soc., P9.4, https://ams.confex.com/ams/27SLS/webprogram/Paper255350.html.
Wurman, J., 2002: The multiple-vortex structure of a tornado. Wea. Forecasting, 17, 473–505, https://doi.org/10.1175/1520-0434(2002)017<0473:TMVSOA>2.0.CO;2.
Wurman, J., and S. Gill, 2000: Fine-scale radar observations of the Dimmitt, Texas (2 June 1995), tornado. Mon. Wea. Rev., 128, 2135–2164, https://doi.org/10.1175/1520-0493(2000)128<2135:FROOTD>2.0.CO;2.
Wurman, J., and M. Randall, 2001: An inexpensive, mobile, rapid-scan radar. Preprints, 30th Int. Conf. on Radar Meteorology, Munich, Germany, Amer. Meteor. Soc., P3.4, https://ams.confex.com/ams/30radar/techprogram/paper_21577.htm.
Wurman, J., and K. Kosiba, 2013: Finescale radar observations of tornadoes and mesocyclone structures. Wea. Forecasting, 28, 1157–1174, https://doi.org/10.1175/WAF-D-12-00127.1.
Wurman, J., J. Straka, and E. Rasmussen, 1996: Fine-scale Doppler radar observations of tornadoes. Science, 272, 1774–1777, https://doi.org/10.1126/science.272.5269.1774.
Wurman, J., Y. Richardson, C. Alexander, S. Weygandt, and P. F. Zhang, 2007: Dual-Doppler and single-Doppler analysis of a tornadic storm undergoing mergers and repeated tornadogenesis. Mon. Wea. Rev., 135, 736–758, https://doi.org/10.1175/MWR3276.1.
Wurman, J., K. Kosiba, and P. Robinson, 2013: In situ, Doppler radar, and video observations of the interior structure of a tornado and the wind–damage relationship. Bull. Amer. Meteor. Soc., 94, 835–846, https://doi.org/10.1175/BAMS-D-12-00114.1.
Here and elsewhere in the paper, the tornado diameter is defined as the distance between the maximum outbound and maximum inbound Doppler velocities within the tornado.
Note that the slowest decay occurring in the most tilted layer means that the downward decay may have been slower than what is presented because of biases owing to tornado geometry.
In this case, a negative lag would mean that changes near the surface preceded changes aloft (i.e., upward strengthening/weakening) and a positive lag means that changes aloft preceded changes near the surface (downward strengthening/weakening).
