1. Introduction
Because of the orographic influence at the eastern fringe of the Alps, parts of northern and eastern Austria are among Europe’s most active hail regions (Kaltenböck and Steinheimer 2015; Punge et al. 2017). Each year, large hail causes severe damage to infrastructure, private property, and agriculture in Austria. In the agricultural sector alone, annual loss estimates of hail damage averaged to EUR 45 million in the past five years (H. Starke, Österreichische Hagelversicherung, 2022, personal communication).
While we understand some parts of thunderstorm and hail formation (e.g., Allen et al. 2020; Kumjian et al. 2021), research gaps persist in our understanding of the dynamics and microphysical processes in local thunderstorm evolution, intensification, and large hail formation, and these gaps limit our forecasting capabilities. Dual-polarized weather radars have provided many new insights into microphysical processes since their widespread introduction into operational service in the 2000s and 2010s, but only a few detailed thunderstorm case studies using polarimetric data have been carried out in Europe (e.g., Höller et al. 1994; Figueras i Ventura et al. 2013; Vulpiani et al. 2015; Montopoli et al. 2021). Much of our process understanding is based on studies of thunderstorms in the United States, most notably the Great Plains (e.g., Kumjian and Ryzhkov 2008; Picca and Ryzhkov 2012; Kaltenböck and Ryzhkov 2013; Snyder et al. 2013; Van Den Broeke 2020), where atmospheric characteristics differ significantly from Europe, as convective available potential energy (CAPE) and vertical wind shear are typically much higher, and local orographic influence is negligible (Brooks 2009; Taszarek et al. 2020).
This study focuses on the larger Vienna area in northeast Austria, located in the transition zone between the foothills of the Alps in the west—the Vienna woods—and the Vienna basin to the east. Typical thunderstorm tracks in Vienna are oriented from west or southwest to east or northeast, corresponding to the midtropospheric flow regime ahead of summertime cold fronts and upper-level troughs, which support suitable environments for deep moist convection and promote convective initiation (Kaltenböck 2004). In the boundary layer of the Vienna basin, southeasterly winds are common in this synoptic configuration, as air flows around the northeastern tip of the Alps. However, westerly surface winds of the gust front often arrive ahead of thunderstorms approaching from the west, stabilizing the prestorm environment as cool air descends from the slopes of the Vienna woods. Farther to the east of the city, new cells are then generated at convergence lines between westerly and southeasterly boundary layer winds. Nevertheless, some thunderstorms intensify over Vienna when approaching from the west, particularly if they have not developed a significant cold pool. Using dual-polarized weather radar data, we examine two opposing cases in detail: 1) a multicell storm rapidly intensifying over Vienna, bringing large hail near the city border on 26 June 2020, and 2) a squall line whose westerly outflow preceded the storm on 21 July 2020, not bringing significant hail in the Vienna region.
Our goal is to compare the microphysical and dynamical features of these two cases by using a rare combination of a cost-efficient, low-power solid-state X-band radar and an operational C-band magnetron radar, located 24 km apart. The X-band radar features volume scan observations at an update rate of only 1 min, allowing detailed insights into rapidly evolving microphysical processes; this is the first time that such a high-resolution weather radar is used to observe storm characteristics in Austria. Therefore, we heavily base our polarimetric analysis on the X-band radar. We also want to highlight the potential of this cost-efficient radar type for storm-scale observations and nowcasting, particularly with regard to the well-established feature of ZDR (differential reflectivity) columns, a polarimetric signature indicating the lifting of liquid drops above the environmental 0°C level in the updraft area of severe storms (Kumjian and Ryzhkov 2008; Kumjian et al. 2014; Snyder et al. 2015, 2017; Kuster et al. 2019). In addition, we examine the storm dynamics using 3D winds retrieved by a dual-Doppler algorithm and link these data with observed polarimetric features. The radar data are complemented by the dense network of weather stations in Vienna, the Austrian ground-based lightning detection system, and citizen hail reports in the densely populated research area.
In section 2, we elaborate on specifications of our data sources and outline the processing methods including quality control and bias correction. In section 3, we analyze the prestorm synoptic environment and storm evolution of both selected cases. In section 4, we present our polarimetric analysis and compare ZDR column, dual-Doppler wind, and lightning data between the cases. Last, we synthesize our findings in section 5 and provide an outlook for future work.
2. Data and methods
a. Radar specifications
Weather radar measurements were collected simultaneously at two sites in the Vienna region (Fig. 1). A dual-polarized, low-power solid state X-band radar (hereinafter referred to as BOKURAD) is located on the roof platform of a university building in Vienna, 4 km northwest to the city center. The same radar model has recently been used for quantitative precipitation estimation (QPE) in northern Europe (Hosseini et al. 2020; Schleiss et al. 2020), pyro-convection observations (McCarthy et al. 2018) and a study of coastal deep convection in Australia (Soderholm et al. 2016). BOKURAD uses pulse compression to retrieve meteorological echoes, resulting in a sensitivity of approximately 15 dBZ at 20-km range and 19 dBZ at 30-km range. Because of the long pulse duration of the compressed pulse, a blind ring emerges around the radar until 8-km range. This area is recovered by a significantly shorter and uncompressed second pulse, resulting in a sensitivity jump at 8 km. The second radar, a dual-polarized C-band radar (hereinafter referred to as RAU) operated by the Austrian weather service Austro Control, is located 5 km south of Vienna airport and 24 km to the southeast of BOKURAD (Kaltenböck 2012). Operating with a magnetron transmitter, this radar is significantly more sensitive to weak echoes with a sensitivity of −11 dBZ at 20-km range. Detailed system specifications of both radars are given in Table 1.
Locations of RAU (green dot) and BOKURAD (red dot), including BOKURAD domain boundary (dashed red line) and dual-Doppler lobes (purple and yellow), as well as the radiosonde launch point Vienna Hohe Warte (“TEMP”; blue triangle).
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
System characteristics of BOKURAD and RAU.
Aside from wavelength-specific differences in attenuation and resonance scattering, another major difference between the systems is their temporal and spatial resolution. While BOKURAD employs a rapid 1-min volumetric scan at a range resolution of 50 m, RAU performs a volumetric half-scan (interleave scan scheme) every 2.5 min at 250-m range resolution. While RAU has a maximum range of 224 km, the maximum range of BOKURAD was limited to only 30 km in 2020. We note BOKURAD’s high half-power beamwidth of 2.7°, resulting in a cross-sectional beamwidth of approximately 1.4 at 30-km range. The sharper beam of RAU reaches the same cross-sectional width only at 91-km range.
For the selected cases, BOKURAD covered a fixed set of eight elevation angles between 2° and 20° each scan, whereas the interleave scan scheme of RAU alternated between a dual-PRF scan covering eight elevations between 1° and 65° and a single-PRF scan covering eight elevations between 0.5° and 33°.
b. Additional observations
Automated weather station data from the national weather service [Zentralanstalt für Meteorologie und Geodynamik (ZAMG)] and Austro Control are used to determine surface impacts of the analyzed storms, e.g., wind gusts and precipitation. In addition, we use routine radiosonde data from launch point Vienna Hohe Warte (ZAMG Headquarters, site marked by “TEMP” in Fig. 1), located 2 km to the northeast of BOKURAD, to determine freezing level height and environmental wind. Furthermore, ZAMG hail reports including location, hail size diameter and report time serve as ground truth to verify polarimetric hail signatures. These manually compiled reports are based on eyewitness and news reports and include entries of the European Severe Weather Database (ESWD). Last, lightning data from the Austrian Lightning Detection and Information System (ALDIS) encompass all intracloud and cloud-to-ground flashes (Schulz et al. 2016).
c. Data processing
Data analyses and visualizations were performed using Python 3.9 with the modules Py-ART (Helmus and Collis 2016), PyDDA (Jackson et al. 2020; Potvin et al. 2012; Shapiro et al. 2009), and CSU RadarTools (Lang et al. 2019). BOKURAD and RAU ZH (reflectivity) and ZDR were attenuation-corrected using Py-ART’s implementation of the Z-PHI method by Testud et al. (2000), which accounts for attenuation in rain but is insufficient to reliably estimate attenuation in hail. Coefficients used for the Z-PHI method are given in Table 2. The ZDR was frequently calibrated based on ZDR values in dry snow and vertical scans in stratiform precipitation (Ryzhkov 2019).
Z-PHI method parameters for attenuation correction of BOKURAD and RAU.
We performed the dual-Doppler analysis with PyDDA, which minimizes a cost function using the three-dimensional variational (3DVAR) method. For the cost function, some of the weights for the input variables must be selected, which we did based on manual comparisons between the retrieved winds in lower levels and surface station data. We increased the weight for the radial velocity data constraint (Co = 100) while retaining the predefined weight for the mass continuity equation (Cm = 1500). Before the retrieval, BOKURAD and full-scan RAU radial velocity data were de-aliased using the 4DD algorithm (James and Houze 2001) and subsequently projected onto a Cartesian grid using the “Barnes2” scheme implemented in Py-ART (Pauley and Wu 1990). Grid spacing was 400 m horizontally and 330 m vertically on 30 levels between 600 and 10 600 m altitude. For the radius of influence (ROI), we used the “dist_beam” option of Py-ART’s interpolation algorithm, which scales ROI based on distance from the radar, and we selected a minimum ROI of 500 m. The final analysis is available at a time step of 5 min, the scan interval of a full scan of RAU.
We primarily used BOKURAD data for the qualitative characterization of polarimetric features and storm development. Therefore, use of RAU variables apart from dual-Doppler retrievals will be mentioned explicitly. For all analyses exclusively with BOKURAD, we used a finer Cartesian grid to take advantage of the high spatial resolution. The grid spacing was 50 m horizontally and 320 m vertically, all other interpolation settings remaining equal.
We also employed our ZDR column tracking algorithm on this gridded dataset. For the tracking, we defined ZDR columns as contiguous volumes of grid points meeting the following criteria: 1) grid point elevation above the environmental freezing level defined by the 0°C isotherm height from the Vienna sounding, 2) ZDR exceeding 1.5 dB and 3) column volume exceeding 10 km3. We note that the ZDR threshold is slightly higher than in previous studies that commonly used 1 dB (e.g., Snyder et al. 2015; Kuster et al. 2019), because we found this to yield more robust ZDR column detections that were less affected by sporadic overcorrections of differential attenuation when the BOKURAD site was hit by strong precipitation. We then analyzed the evolution of multiple variables related to the ZDR column: the 90th percentile of ZDR, ZH, and updraft speed within the column as well as the total ZDR column volume. Instead of the 90th percentile, we also evaluated the median and the 95th percentile of ZDR/ZH/updraft speed but found the 90th percentile to be a good compromise between sometimes excessive scan-to-scan fluctuations of the 95th percentile and less pronounced cell-to-cell differences when using the median. We also tested the 98th percentile of column height above freezing level, but the limited coverage at altitudes > 5.5 km above ground level (AGL) did not allow conclusive results. For the sake of comparison, we also applied the same detection algorithm and analyses to 5-min RAU data on a grid with 200-m horizontal and 330-m vertical spacing.
In addition to the ZDR column metrics, we calculated the size of the updraft area >10 m s−1 and tracked the lightning rate for the targeted cell in each case study. For calculation of these variables, we manually assessed the extent of the respective cell to include flashes in the periphery as well.
We also applied a hydrometeor identification algorithm (HID) to gridded BOKURAD and RAU data. We used a fuzzy-logic classification scheme developed by Colorado State University and implemented in their RadarTools Python module (Dolan and Rutledge 2009; Dolan et al. 2013). It provides a classification for a total of 10 hydrometeor types: drizzle, rain, ice crystals, aggregates, wet snow, vertical ice, low-density graupel, high-density graupel, hail and big drops (>5 mm). Recently, this algorithm was also successfully tested for X-band observations of a hailstorm during the RELAMPAGO campaign in Argentina (Bechis et al. 2022).
3. Case study selection, synoptic-scale environment, and storm evolution
Both hailstorms selected for a detailed examination approached Vienna from the west but developed differently on the transition to the Vienna basin. On 26 June 2020, a rapidly evolving multicell storm brought large hail with a maximum diameter of 5 cm in the south of Vienna, while on 21 July 2020, a fast-propagating squall line brought small hail < 2 cm only for brief periods while within range of BOKURAD.
a. Synoptic environment
At 1200 UTC 26 June 2020, an upper-level low was centered over the Czech Republic, north of the radar domain. This configuration resulted in a westerly to southwesterly upper-level flow over northeastern Austria and a deep moist unstable layer above 800 hPa due to the presence of cold air aloft. Near the center of the upper-level low, moisture was abundant in mid and upper-levels, but only moderate at the surface with afternoon dewpoints in Vienna between 15° and 16°C and maximum air temperatures of 27°–28°C. Weak southeasterly winds prevailed in the Vienna basin below 1 km AGL, promoting warm air advection from the Pannonian basin. As evidenced by the Vienna sounding at 1200 UTC (Fig. 2a), the boundary layer was topped by a stable layer between 800 and 880 hPa, providing CIN of 25 J kg−1. Surface-based virtual temperature-corrected CAPE reached 1050 J kg−1, 0–6-km bulk shear amounted to 26 m s−1, and 0–3-km storm-relative helicity (SRH), calculated from observed storm motions, was 116 m2 s−2.
Radiosonde observations (skew T–logp diagrams) at Vienna Hohe Warte at 1200 UTC (a) 26 Jun and (b) 21 Jul 2020. Colored thick lines show ambient temperature (red) and dewpoint temperature (green). Red shading indicates CAPE. Wind barbs plotted with half and full barbs indicate 5 and 10 kt (1 kt ≈ 0.51 m s−1), respectively. The hodograph is color coded, ranging from surface (dark blue) to 300 hPa (yellow).
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
On 21 July 2020, the Vienna basin was located downstream of a 500-hPa trough, leading to westerly flow over Austria. The surface cold front extended from the surface low in Finland to Belarus, western Ukraine, Slovakia, and the Czech Republic. In northeast Austria, south of the almost stationary western tip of the cold front, warm and moist air masses were present in the boundary layer. Surface winds were weak and – ahead of a prefrontal convergence line in the afternoon – mostly blowing from northeast in the Vienna basin. Winds subsequently veered clockwise to west with increasing altitude, resulting in 0–6-km bulk shear of 20 m s−1 and 0–3-km SRH of 101 m2 s−2. In contrast to 26 June, there was a distinct midlevel dry layer present and surface-based CAPE was slightly higher with 1570 J kg−1 at 1200 UTC (Fig. 2b).
b. Storm initiation and evolution
On 26 June 2020, initial cells developed between 1230 and 1310 UTC over the Austrian northeastern Alpine region, 60–70 km southwest of Vienna (Fig. 3) and therefore outside of the BOKURAD domain, but within range of RAU. Surface wind measurements indicate mesoscale flow directed from the plains toward Alpine terrain as part of a diurnal wind system known as “Alpine pumping” (Weissmann et al. 2005; Bica et al. 2007), promoting convective initiation and locally increased wind shear where the inflow converges in mountainous areas. Subsequently, the cells propagated northeast, with the northernmost cell heading toward Vienna. This cell significantly intensified between 1320 and 1340 UTC in an area of surface wind convergence. Despite not exhibiting the radar appearance of a supercell, the cell split around 1345 UTC, shortly before reaching the western city border of Vienna and already within range of BOKURAD. The right mover further intensified upon reaching the Vienna basin shortly after 1400 UTC, and veered along the southern city border of Vienna, now propagating toward the east-southeast (Fig. 3, left panel 1415–1515 UTC). Hailstones with diameters up to 5 cm were reported between 1440 and 1500 UTC directly south of Vienna. After 1500 UTC, the storm passed Vienna airport and quickly dissipated upon crossing the Danube River after 1530 UTC.
Comparison of storm evolution with RAU 1.8° PPIs of radar reflectivity (left) from 1345 to 1545 UTC 26 Jun 2020 and (right) from 1330 to 1530 UTC 21 Jul 2020.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
On 21 July 2020, thunderstorms that later moved into the Vienna area developed as cluster of small convective cells between 1000 and 1120 UTC over the southwestern Czech Republic, close to the Austrian border. The storms were triggered downstream of the surface cold front at a convergence line, which progressively moved eastward toward northern Austria together with the thunderstorm cells. The cells transitioned into a more line-shaped appearance after 1300 UTC. At 1340 UTC, eyewitnesses reported 6-cm hailstones in a city 55 km west-northwest of Vienna. Wind gusts of 15 m s−1 were recorded at a ZAMG weather station in the same city. Thereafter, the line weakened and split up into two separate cells, without significant hail for the next hour. The southern cell veered to the right and hit Vienna at 1430 UTC approaching from the Vienna Woods area, and the northern cell propagated on a quasi-straight path over the Weinviertel region north of Vienna, further weakening as it moved eastward. BOKURAD was hit by the southern cell between 1440 and 1452 UTC. The outflow of this cell intensified over Vienna, reaching gusts of 21 m s−1 from northwest at 1450 UTC in the city center. Close to the city center, there was a single report of small hail < 2 cm. As the outflow pushed ahead of the storm over Vienna, new cells were triggered southeast of the city border after 1445 UTC. They quickly intensified at the southern flank, hitting RAU at 1502 UTC. Between 1500 and 1510 UTC, a maximum gust of 22 m s−1 was recorded at Vienna airport. The system was now moving into a higher-CAPE environment, as insolation had been ongoing throughout most of the day in this area south of the thunderstorm anvil. The storms proceeded southeastward, producing small hail with 2–3-cm diameter at multiple locations southeast of RAU and outside of the BOKURAD domain. The system persisted for several more hours after crossing the Hungarian border.
For both case study days, animations of 1-min BOKURAD ZH including ALDIS lightning observations (squares and diamonds for intracloud and cloud-to-ground lightning, respectively) are provided in the online supplemental material (file “SUPP1.mp4” showing 26 June 2020, and file “SUPP2.mp4” showing 21 July 2020).
4. Results
a. 26 June 2020
Our analysis of the 26 June 2020 event starts with the approach of the cells toward the Vienna basin around 1400 UTC, shortly after the storm split described in section 3b. Initially, three distinct updraft areas are present as indicated by ZDR columns (see Fig. 4a). Of these, the southernmost cell is of prime interest, as this is the only one rapidly intensifying during the transition from the Vienna Woods to the basin, encountering southeasterly surface winds and orographic lifting. Conversely, the central updraft quickly dissipates after reaching the basin (unfortunately obscured by the cone of silence as it approaches BOKURAD), and the northern updraft also weakens, albeit slower than the central updraft (see Fig. 4b with a ZDR column still present at the northern cell). A key factor for the intensification of the southernmost cell at the transition to the basin is the poorly developed cold pool, as cooling may have been limited by the high moisture content throughout the troposphere. Therefore, low-level southeasterly winds are sustained in the Vienna basin during thunderstorm approach and provide orographic lifting.
Evolution of BOKURAD ZDR on CAPPI level 3.7 km above mean sea level (MSL), approximately 700 m above the environmental freezing level, between (a) 1410 and (b) 1420 UTC 26 Jun 2020, and zoomed-in illustrations for the map region indicated by the red-outlined rectangle in (b) between (c) 1420 and (d) 1430 UTC. These also show an overlay of dual-Doppler wind retrieval with horizontal wind (arrows) and updraft speed contours (5, 10, 15, and 20 m s−1).
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
The zoomed-in area in Fig. 4c illustrates that the ZDR column at the rear (upshear) side of the cell overlaps with the main updraft area derived by dual-Doppler analysis. Initially, ZH within the column is remarkably low (15–30 dBZ), indicative of sparse large drops or wet hail. Between 1420 and 1430 UTC, the magnitude of ZDR within the ZDR column further increases, as the top of the column reaches an altitude of almost 6 km AGL (cf. with Fig. 5e, which is slightly displaced to the east of the maximum ZDR column altitude), close to 3 km above the environmental 0°C level. During the same time span, ZH increases as more hydrometeors grow in the updraft (not shown). The areal overlap between ZDR column and main updraft persists, although the updraft area is slightly larger at this height and extends farther to the east (see Fig. 4d). A pattern of convergence at the bottom of the updraft and divergence at its top—at least in the plane of the cross section—emerges and is sustained during the entire lifetime of the updraft (cf. Fig. 5). Within the volume of strongest updrafts itself, between 3 and 7 km AGL, horizontal wind speeds are significantly lower than in the surroundings, especially toward the southeast (∼5 vs >15 m s−1). This increases the residence time of hydrometeors within the updraft, an important factor for hail growth as found in modeling studies (Kumjian and Lombardo 2020; Kumjian et al. 2021).
Beginning of hail formation based on BOKURAD data at 1430 UTC 26 Jun 2020 as seen in cross sections through the storm inflow: (a) ZH CAPPI (2.3 km MSL) with dual-Doppler horizontal wind and updraft contours as in Fig. 4, with the red arrow indicating the cross section shown in (b)–(f), which show (b) HID, (c) ZH, (d) ρhv, (e) ZDR, and (f) KDP. For the wind arrows in (b)–(f), the horizontal component along the cross section is paired with the vertical wind component.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
The inflow of the storm, southeast of the main updraft region, follows an increasingly cyclonic path in the lowest levels, which is first seen around 1430 UTC and intensifies within the next minutes. On the northwest side, a weaker anticyclonic vortex is visible (see Fig. 5a). The cross section illustrated in Fig. 5 is placed at the eastern fringe of the ZDR column, parallel to the direction of the low-level inflow (southeast to northwest) and roughly orthogonal to the cell propagation direction (west to east). At this stage, the HID indicates the presence of hailstones along the southern periphery of the main updraft at >6 km AGL with a broad high-density graupel zone beneath (Fig. 5b). A circulation-like pattern governed by the southeasterly low-level inflow, the strong updraft downstream and the reverse flow (northwest) at upper levels has developed almost orthogonal to the cell propagation. Hydrometeors grow in the updraft and are then largely displaced toward the overhanging echo in upper levels. Simultaneously, some hydrometeors falling into the low-level inflow seem to be recirculated (Fig. 5c). Figure 5d provides an overview of ρhv (cross correlation coefficient), which is very low (0.75–0.85) in the main updraft, likely indicating the mixed phase growth region, and is of modest values > 0.9 in the initial hail area of the overhanging echo, suggesting smaller-sized hailstones (Picca and Ryzhkov 2012). Focusing on ZDR (Fig. 5e) we see the separation between liquid and solid particles in the inflow along the environmental 0°C level and continued higher values extending above the environmental 0°C isotherm in the updraft regime, constituting the eastern part of the ZDR column. In the transition zone between liquid and solid hydrometeors along the inflow, negative KDP is found (see Fig. 5f), which is not confirmed by RAU data (not shown). We suspect that KDP estimation in this area is erroneous, for which multiple explanations are possible. Difficulties in filtering the backscatter differential phase δ from the propagation phase shift in the melting layer, for example, could be a major factor. Gradients in δ can then cause local positive and negative errors of KDP, resulting in a KDP “dipole” (similar to the pattern seen in Fig. 6). Another possibility is non-Rayleigh resonance scattering of large hydrometeors: Large melting hail has been shown to cause negative KDP in the X-band in scattering simulations (e.g., Ryzhkov et al. 2013). Further options include conditions of nonuniform beam filling (Ryzhkov 2007) and statistical uncertainties (e.g., Reimel and Kumjian 2021).
The KDP dipole pattern in an RHI plot (azimuth angle 180°) of BOKURAD (a) differential phase shift (ΦDP), (b) KDP, (c) ρhv, and (d) ZDR at 1430 UTC 26 Jun 2020.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
Either way, these negative KDP values influence HID results and contribute to the “wet snow” area found in the melting layer, as the fuzzy logic classification strongly responds to negative KDP.
To illustrate the same area of the storm and account for storm propagation, we displace the cross section slightly eastward for the next time step of the dual-Doppler analysis at 1435 UTC (Fig. 7a). Here, based on ZH, a significant intensification is obvious relative to 1430 UTC (Fig. 5a), and a bounded weak echo region (BWER) above the inflow is visible (Fig. 7c). Hail is more abundant in mid and upper levels (Fig. 7b), and in a CAPPI illustration of the HID classification at 3.8 km AGL (Fig. 8b), some hundred meters above the environmental freezing layer, the hail category is found along an arc to the north, east and southeast of the updraft core. While the retrieval of hydrometeor trajectories is outside the scope of this work, the dual-Doppler data can provide indications for hailstone pathways despite the wind streamlines not matching the hydrometeor trajectories exactly. The dual-Doppler data suggest that the arc-shaped hail distribution may be caused by the divergence in the mid and upper portions of the updraft, where hydrometeors are ejected into these areas (Fig. 8c). Closer to the surface, however, the hail stones falling along the southeastern periphery of the updraft are captured by the strong low-level inflow and advected back in northerly direction (Fig. 8a). This transport and potential recirculation are clearly visible in Figs. 7b and 7c between 3.5 < X, Y < 5.5 km and Z = 2.5 km. At this time ρhv has decreased in the western tip of the cell with a local minimum of 0.7 close to the main updraft (Fig. 9b), resembling the “ρhv hole” described by Kaltenböck and Ryzhkov (2013); ρhv has also decreased along the cross section (Fig. 7d), indicative of the wet growth of hydrometeors, some of them now entering the X-band resonance scattering regime. This is further corroborated by the ZDR cross section (Fig. 7e); ZDR has dropped below 0 in some regions of the HID hail category, and in regions of lowest ρhv. This arises either directly from the resonance scattering properties of large hail or indirectly from under-corrected differential attenuation. Meanwhile, the ZDR column is still present at the western part of the main updraft, although ZDR has dropped relative to previous time steps (Fig. 9c).
As in Fig. 5, but for 1435 UTC 26 Jun 2020.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
BOKURAD HID classification for CAPPI levels (left) 1.3, (center) 4.1, and (right) 6.1 km MSL at (a)–(c) 1435 and (d)–(f)1440 UTC 26 Jun 2020 . Dual-Doppler horizontal wind and updraft contours are presented as in Fig. 4.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
BOKURAD CAPPI (4.1 km MSL) of (a) ZH, (b) ρhv, (c) ZDR, and (d) KDP at 1435 UTC 26 Jun 2020. Dual-Doppler horizontal wind and updraft contours [in (a)] are presented as in Fig. 4.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
Within the next minutes, there is an ongoing increase of ZH, driven by fallout of large hydrometeors—particularly at lower levels and, at first, on the inflow side in the southern segment (not shown). Starting from 1438 UTC, the HID begins to show larger areas of hailfall near the surface, shifting slowly to the northern flank of the cell within the next 2–4 min. This shift is caused by the southerly inflow in lower levels, displacing hailstones toward the north (Fig. 8d). At 1442 UTC, the polarimetric signatures of BOKURAD (Fig. 10) are indicative of large hail in the lowest levels: In the hail swath, ρhv < 0.9 is collocated with ZH > 60 dBZ, ZDR values are positive between 2 and 5 dB and KDP is close to zero. Polarimetric variables of RAU hint to hailfall in the same region, which is reflected by the HID hail category largely overlapping between both radars (Fig. 10e). Citizen hail reports broadly confirm the HID hail classification, although the small area with the highest density of reports at X = 5 km and Y = −15 km was not entirely in the hail category area of BOKURAD due to strong attenuation. Attenuation occurs down-radial of the hail region both for BOKURAD and RAU, seen by a total loss of signal south and southeast of the hail region in case of BOKURAD and as a significant reduction of ZH at the northwestern edge of the cell for RAU.
(top) BOKURAD and (bottom) RAU CAPPI (1.0 km MSL) of (a) ZH, (b) ρhv, (c) ZDR, (d) KDP, and (e) HID category with citizen hail reports (triangles are colored on the basis of reported hail diameter) at 1442 UTC 26 Jun 2020.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
After the severe initial pulse of hail, the updraft remains strong, with maximum vertical wind speeds consistently greater than 20 m s−1 in midlevels. Subsequently, multiple pulses of smaller hail can be traced by their polarimetric signature as they descend toward the surface (see file “SUPP3.mp4” in the online supplemental material). They reach the surface approximately at 1453, 1457, 1503, and 1507 UTC. The ZH in the cell core remains greater than 60 dBZ until 1530 UTC; afterward the system quickly dissipates.
b. 21 July 2020
The storm event on 21 July 2020 differed in various ways, particularly with regard to the well-developed cold pool of the approaching thunderstorms, the faster propagation, the triggering of new cells along the outflow boundaries in the east of Vienna and the lack of large hail observations. Below, we discuss the spatiotemporal evolution in detail, again beginning with the transition of the storms into the Vienna basin and the development of a new updraft within the dual-Doppler lobes.
The storm hitting Vienna is preceded by its gust front, causing a wind shift from northeast to northwest at Viennese weather stations around 1420 UTC, more than 10 min prior to the thunderstorm passage over the city. This corresponds to a regularly observed behavior of squall lines approaching Vienna from the west (see section 1). In this case, the cell persists over the city area without much change in reflectivity, but based on the dual-Doppler retrieval, updraft speeds decrease over the city area.
The outflow boundaries of the cell under discussion and another cell north of Vienna are clearly seen in RAU ZH and dual-Doppler data at 1450 UTC (Fig. 11). The outflow boundary is not detected by BOKURAD, because its ZH sensitivity is too low to trace the low reflectivity associated with convergence lines. The outflow leads to strong convergence between northwesterly and southeasterly winds southeast of Vienna. On the southern end of the main system, close to the convergence line, a strong new updraft develops, indicated by the dual-Doppler data, increasing reflectivity above 3 km AGL and a confined ZDR column, which grows considerably within a few minutes (Fig. 12). We note that in contrast to the case on 26 June 2020, the ZDR column is located at the front of the cell and elongated perpendicular to the direction of movement, illustrating the lift along the outflow boundary near the surface. This is corroborated by the dual-Doppler vertical wind speeds at 1455 UTC seen in the cross sections in Fig. 13. The wind field also shows the linear nature of this storm as northwesterly winds prevail in all levels.
RAU ZH on PPI 1.8° with dual-Doppler horizontal wind at 0.9 km MSL compared with ground observations at 1450 UTC 21 Jul 2020. The dashed black lines indicate the positions of outflow boundaries visualized by locally enhanced ZH.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
As in Fig. 4, evolution of BOKURAD ZDR on CAPPI level 3.7 km MSL with dual-Doppler wind overlay at (a) 1450, (b) 1455, and (c) 1500 UTC 21 Jul 2020 . Red arrows in (b) and (c) indicate the position of cross sections in Fig. 13, below.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
Cross sections of BOKURAD (a),(b) ZH; (c),(d) ZDR; (e),(f) (ρhv); (g),(h) HID category at (left) 1455 and (right) 1500 UTC 21 Jul 2020 along the transects indicated by the red arrows in Figs. 12b and 12c. Wind arrows depict the horizontal wind component along the cross section and the vertical component.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
Between 1450 and 1455 UTC, maximum BOKURAD ZH in midlevels increases from 35 to >50 dBZ, and rain begins to reach the surface (Figs. 13a,b). The ZDR > 5 dB in low levels north of the cross sections (not shown) indicates very large drops, characteristic for differential sedimentation at the onset of precipitation in a developing thunderstorm (Loeffler and Kumjian 2020). At the leading edge of the storm the overhanging echoes, consisting mainly of low-density graupel as indicated by the HID (Fig. 13g), are found significantly ahead of surface precipitation. These echoes are, due to the strong northwesterly flow at upper levels, present in all longer-lived cells on this day. Between 1455 and 1500 UTC, the ZDR column almost dissipates while ZH further increases. The ρhv drops substantially, suggesting presence of larger graupel or hail within the cell in alignment with the HID classification at 1500 UTC (Fig. 13h). Alongside the decaying ZDR column, the main updraft shifts backward from the leading edge and toward higher altitude. As the updraft moves into the area of high hydrometeor load, the upward motion is undercut by latent cooling and downward momentum transfer and is now found above the cold pool, which itself accelerated due to strong latent cooling. Subsequently, the weakened updraft cannot sustain substantial hail growth, resulting overall in smaller graupel and hail than in the 26 June 2020 event, despite higher CAPE on 21 July. This is confirmed by the HID classification that mostly shows the “rain” and “big drops” categories in lower levels. However, the system continues to develop new updrafts along its southern edge, illustrated by an emerging ZDR column at 1502 UTC (not shown), which, due to the fast propagation of the storm, leaves the radar domain of BOKURAD soon thereafter.
In closing, we briefly highlight a remarkable upper-level feature of this event. Doppler data of BOKURAD and RAU reveal convergence within the overhanging precipitation region prior to the squall line, seen between 1415 and 1450 UTC to the north of Vienna. On closer inspection using a cross section, a wave pattern between 4 and 8 km AGL is visible in ZH and dual-Doppler winds (Fig. 14). Based on HID, the precipitation is composed of low-density graupel, aggregates and vertical ice. While the radiosonde ascent at 1200 UTC (Fig. 2b) does not show significant changes of wind speed or direction in this layer, the dual-Doppler analysis implies strong vertical wind shear in the area of wave observations, potentially providing a suitable environment for trapped gravity waves (Trier and Sharman 2018).
Upper-level gravity waves in the overhanging precipitation region at 1425 UTC 21 Jul 2020: (a) BOKURAD ZH at CAPPI level 6.8 km MSL with dual-Doppler horizontal wind and updraft speed contours and with the red arrow indicating the location of the cross sections of (b) ZH and (c) HID.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
c. ZDR column, updraft, and lightning evolution
Tracking of ZDR column properties such as maximum height or volume has been proposed as potential nowcasting technique for thunderstorm evolution with lead times of 10–20 min (Snyder et al. 2015; Kuster et al. 2019). Here, we compare the temporal evolution of multiple ZDR column, dual-Doppler updraft, and lightning metrics for the main cells of 26 June 2020 (hereinafter “event A” and “column A”) and 21 July 2020 (“event B” and “column B”) to relate them to microphysical processes and to evaluate their potential nowcasting application. We will first focus on BOKURAD ZDR column data, but also point out the major differences to tracking metrics from RAU at the end of this section.
We note that one of the most prominent differences between the events is the lifetime of their ZDR columns. While BOKURAD tracked column A for almost one hour between 1412 and 1505 UTC, ZDR columns in event B lasted between 5 and 15 min only. Therefore, we combine the two short-lived ZDR columns of the southern cell described in the previous section and treat them as single event, given that they belong to the same cell. These columns last from 1450 to 1458 UTC and from 1505 to 1512 UTC, respectively. Despite their different lifetimes, the onset of columns A and B is similar, and they rapidly gain size within the first minutes of their lifetime (Fig. 15c). After the initial expansion, BOKURAD tracking indicates a phase of stagnating volumes with increasing ZH in both columns, with the latter hinting to hydrometeor growth in the updraft (Fig. 15b). At this stage, the 90th percentile of ZDR strongly differs between the columns: Column A reaches a peak 90th percentile of ZDR > 5 dB at 1424 UTC, as compared with approximately 3 dB in column B (Fig. 15a). We hypothesize that two processes are responsible for higher ZDR values during event A: 1) more humid midtropospheric conditions and the longer-lived updraft, well separated from the precipitation core, supporting formation and lifting of larger hydrometeors within the ZDR column, and 2) size sorting in the updraft. We note that the 90th percentile of ZDR of column A is significantly higher than in all other detected columns of the same day, which do not exceed a 90th percentile of 4 dB throughout their lifetime when tracked with BOKURAD (not shown).
Temporal evolution of (a)–(c) ZDR column metrics (separate for BOKURAD and RAU); (d),(e) updraft metrics; and (f) the number of flashes per minute (separate for intracloud and cloud-to-ground lightning) for the analyzed cells on 26 Jun and 21 Jul 2020, with 0 being based on the timing of the first ZDR column detection. The red-shaded time intervals indicate timing of hailfall at the surface on 26 Jun 2020.
Citation: Monthly Weather Review 151, 4; 10.1175/MWR-D-22-0185.1
Column B reaches a maximum 90th percentile of ZH of 57 dBZ at 1457 UTC, 7 min after its initial detection. At this time, its 90th percentile of ZDR and volume have already decreased. This corresponds to the shift of the updraft location from the leading edge into the cell core detailed in section 4b. Within 2 min, column B collapses due to hydrometeor load and does not fulfil the volume criterion of >10 km3 anymore. Updraft speeds and size lag behind (Figs. 15d,e): 90th percentile of w increases until 1505 UTC, before sharply declining afterward. The new ZDR column emerging at 1505 UTC is located closer to the reflectivity core, as initial 90th percentile of ZH values within the column are already >50 dBZ, while 90th percentile of ZDR is still around 3 dB (Figs. 15a,b). Within the following minutes up to the exit of the radar domain, the ZDR column metrics remain relatively constant.
In column A, the initial increase of 90th percentile of ZH is slower, but continuous for more than 20 min until 1435 UTC, the onset time of large hail formation in upper levels. This evolution of 90th percentile of ZH is consistent with the increase of updraft size and 90th percentile of updraft speeds (Figs. 15d,e), which show a lag of 10–15 min relative to peak 90th percentile of ZDR. A similar effect was described in a modeling study by Snyder et al. (2015), who found that changes in ZDR column height preceded changes in updraft intensity.
In contrast to ZH, 90th percentile of ZDR already drops after 1422 UTC, presumably due to the lower ZDR of graupel or hail that is incorporated into the ZDR column. The column volume oscillates between 1420 and 1435 UTC but remains considerably higher than the imposed threshold. At 1435 UTC, 90th percentile of ZH begins to drop significantly, followed by a sharp decline of column volume 4 min later. The ZDR column does not overlap with high ZH anymore as high ZH regions are now filled with hail, having low ZDR. After the initial pulse of large hail, reaching the ground at approximately 1442 UTC, 90th percentile of ZH in the column recovers until 1448 UTC and remains close to 57 dBZ until 1451 UTC. The column volume, however, does not recover to previous levels. At 1453 UTC, there is another oscillation with a sharp decline and subsequent increase of ZH, corresponding to the next pulse of hailfall. The column finally dissipates during the last hail pulse shortly after 1500 UTC, more than 50 min after its first detection. The pulses of hailfall can also be traced in successive ρhv cross sections (see file “SUPP3.mp4” in the online supplemental material).
Lightning data reveal significantly higher 1-min flash numbers for event B and a high amount of intracloud lightning in comparison with cloud-to-ground lightning in both events (Fig. 15f). While lightning activity decreases following the dissipation of the first ZDR column of event B, a sharp increase emerges after development of the new column a few minutes later. Total flash numbers of event A slightly increase toward the first large hail occurrence, but remain well below 10 flashes per minute, a threshold commonly associated with severe weather (Schultz et al. 2009). There are no indications for a link between flash rate to ZDR column or updraft metrics in this case.
RAU tracking (dotted lines in Figs. 15a–c) did not differentiate between the emerging ZDR columns detected by the BOKURAD tracking on the one hand and columns of preexisting updrafts on the other. Therefore, they were treated as single continuously existing ZDR columns, which can be considered erroneous based on the 1-min BOKURAD ZDR data. This example indicates the potential added value of rapid-update scans for tracking ZDR columns and detecting new updrafts. Also, the ZDR column volume with RAU tracking was significantly higher in the beginning of the targeted periods, which could be caused by multiple factors: 1) less influence of the cone of silence due to higher elevation angles; 2) coarser horizontal grid resolution and therefore less separation between small-scale features (i.e., RAU tracks one object while BOKURAD sees two columns); or 3) the offset between elevation angles during the 5-min scan period artificially increases the volume. The timing of peak ZDR magnitudes was captured in a similar way by RAU and BOKURAD, with the ZDR difference between the two columns being less pronounced for RAU. Also, in contrast to BOKURAD, the tracking of column A stopped before the first hail pulse due to the strong decrease of volume, therefore the ZH fluctuations around the timings of hailfall could not be seen in RAU data.
5. Summary and conclusions
We analyzed the evolution of two hail-bearing storms in the Vienna basin using a combination of a cost-efficient polarimetric X-band weather radar (BOKURAD) and an operational C-band weather radar (RAU). The main dynamic difference of the storms was the well-developed cold pool in the small hail case on 21 July 2020 and the missing (or at least much weaker) cold pool in the large hail case on 26 June 2020. The development of the cold pool on 21 July 2020 was likely favored by dry midlevel air. The cold pool and the lifting along its leading edge favored the storm propagation in form of a squall line. The outflow with westerly winds preceded the storm in Vienna and triggered a new updraft southeast of the city. Large hail formation was inhibited, as the updraft core weakened when it shifted from the leading edge to the center of the storm within minutes after initiation. In contrast, the complex wind field on 26 June 2020 sustained a strong and wide updraft for almost 1 h. It was located on the rear (i.e., upshear) side of the storm, displaced from the reflectivity core, and overlapped with an intense ZDR column. Based on dual-Doppler winds and a hydrometeor classification algorithm, the motion of hail stones could be deduced, and polarimetric data revealed multiple rapid pulses of hail formation and fallout. For both radars, we found an overall good agreement between the hail category of the hydrometeor classification and the location of citizen hail reports.
We compared metrics of ZDR column development, updraft speed, and lightning activity for both cases. The high temporal resolution of the polarimetric radar observations (1 min update rate) was found to be very valuable, as observed time scales of ZDR column evolution were extremely short. Some of the metrics in use, particularly the ZDR column volume, varied significantly between both radars, highlighting uncertainties related to the different measurement characteristics and resolutions. The most promising forecasting parameter and discriminator between the large hail and non-severe case was the 90th percentile of ZDR within the column, which was highest on 26 June 2020 approximately 20 min before the first large hail occurrence at the surface and preceded highest updraft speeds by more than 10 min. In contrast to ZDR column height, which was used as forecasting metric in other studies, 90th percentile of ZDR is not as sensitive to the vertical resolution of the volume scan, potentially opening a wider area of application. Future studies addressing nowcasting of severe storms should consider testing the potential of this ZDR column metric on a larger dataset. Furthermore, a systematic comparison between dual-Doppler and ZDR column data could further improve our understanding of this polarimetric feature and provide a more complete picture of the forecasting applicability under different synoptic conditions.
Acknowledgments.
We acknowledge the support of Georg Pistotnik (ZAMG) for providing data of citizen hail reports and valuable input in discussions. Furthermore, we thank Wolfgang Schulz (ALDIS) for providing lightning data. We also express our gratitude to the anonymous reviewers whose comments helped to significantly improve this article.
Data availability statement.
BOKURAD X-band data of 26 June 2020 and 21 July 2020 are available online (https://doi.org/10.5281/zenodo.6780038). Austro Control radar data are proprietary but can be obtained upon request from the company for research purposes. The same applies to ALDIS lightning data.
REFERENCES
Allen, J. T., I. M. Giammanco, M. R. Kumjian, H. Jurgen Punge, Q. Zhang, P. Groenemeijer, M. Kunz, and K. Ortega, 2020: Understanding hail in the Earth system. Rev. Geophys., 58, e2019RG000665, https://doi.org/10.1029/2019RG000665.
Bechis, H., and Coauthors, 2022: A case study of a severe hailstorm in Mendoza, Argentina, during the RELAMPAGO-CACTI field campaign. Atmos. Res., 271, 106127, https://doi.org/10.1016/j.atmosres.2022.106127.
Bica, B., and Coauthors, 2007: Thermally and dynamically induced pressure features over complex terrain from high-resolution analyses. J. Appl. Meteor. Climatol., 46, 50–65, https://doi.org/10.1175/JAM2418.1.
Brooks, H. E., 2009: Proximity soundings for severe convection for Europe and the United States from reanalysis data. Atmos. Res., 93, 546–553, https://doi.org/10.1016/j.atmosres.2008.10.005.
Dolan, B., and S. A. Rutledge, 2009: A theory-based hydrometeor identification algorithm for X-band polarimetric radars. J. Atmos. Oceanic Technol., 26, 2071–2088, https://doi.org/10.1175/2009JTECHA1208.1.
Dolan, B., S. A. Rutledge, S. Lim, V. Chandrasekar, and M. Thurai, 2013: A robust C-band hydrometeor identification algorithm and application to a long-term polarimetric radar dataset. J. Appl. Meteor. Climatol., 52, 2162–2186, https://doi.org/10.1175/JAMC-D-12-0275.1.
Figueras i Ventura, J., F. Honoré, and P. Tabary, 2013: X-band polarimetric weather radar observations of a hailstorm. J. Atmos. Oceanic Technol., 30, 2143–2151, https://doi.org/10.1175/JTECH-D-12-00243.1.
Helmus, J. J., and S. M. Collis, 2016: The Python ARM radar toolkit (Py-ART), a library for working with weather radar data in the Python programming language. J. Open Res. Software, 4, e25, https://doi.org/10.5334/jors.119.
Höller, H., M. Hagen, P. F. Meischner, V. N. Bringi, and J. Hubbert, 1994: Life cycle and precipitation formation in a hybrid-type hailstorm revealed by polarimetric and Doppler radar measurements. J. Atmos. Sci., 51, 2500–2522, https://doi.org/10.1175/1520-0469(1994)051<2500:LCAPFI>2.0.CO;2.
Hosseini, S. H., and Coauthors, 2020: Evaluation of a new X-band weather radar for operational use in south Sweden. Water Sci. Technol., 81, 1623–1635, https://doi.org/10.2166/wst.2020.066.
Jackson, R., S. Collis, T. Lang, C. Potvin, and T. Munson, 2020: PyDDA: A Pythonic direct data assimilation framework for wind retrievals. J. Open Res. Software, 8, 20, https://doi.org/10.5334/jors.264.
James, C. N., and R. A. Houze Jr., 2001: A real-time four-dimensional Doppler dealiasing scheme. J. Atmos. Oceanic Technol., 18, 1674–1683, https://doi.org/10.1175/1520-0426(2001)018<1674:ARTFDD>2.0.CO;2.
Kaltenböck, R., 2004: The outbreak of severe storms along convergence lines northeast of the Alps: Case study of the 3 August 2001 mesoscale convective system with a pronounced bow echo. Atmos. Res., 70, 55–75, https://doi.org/10.1016/j.atmosres.2003.11.003.
Kaltenböck, R., 2012: New generation of dual polarized weather radars in Austria. Extended Abstracts, Seventh European Conf. on Radar in Meteorology and Hydrology, Toulouse, France, ERAD, 6 pp., https://www.meteo.fr/cic/meetings/2012/ERAD/extended_abs/NET_166_ext_abs.pdf.
Kaltenböck, R., and A. V. Ryzhkov, 2013: Comparison of polarimetric signatures of hail at S and C bands for different hail sizes. Atmos. Res., 123, 323–336, https://doi.org/10.1016/j.atmosres.2012.05.013.
Kaltenböck, R., and M. Steinheimer, 2015: Radar-based severe storm climatology for Austrian complex orography related to vertical wind shear and atmospheric instability. Atmos. Res., 158–159, 216–230, https://doi.org/10.1016/j.atmosres.2014.08.006.
Kumjian, M. R., and A. V. Ryzhkov, 2008: Polarimetric signatures in supercell thunderstorms. J. Appl. Meteor. Climatol., 47, 1940–1961, https://doi.org/10.1175/2007JAMC1874.1.
Kumjian, M. R., and K. Lombardo, 2020: A hail growth trajectory model for exploring the environmental controls on hail size: Model physics and idealized tests. J. Atmos. Sci., 77, 2765–2791, https://doi.org/10.1175/JAS-D-20-0016.1.
Kumjian, M. R., A. P. Khain, N. Benmoshe, E. Ilotoviz, A. V. Ryzhkov, and V. T. J. Phillips, 2014: The anatomy and physics of ZDR columns: Investigating a polarimetric radar signature with a spectral bin microphysical model. J. Appl. Meteor. Climatol., 53, 1820–1843, https://doi.org/10.1175/JAMC-D-13-0354.1.
Kumjian, M. R., K. Lombardo, and S. Loeffler, 2021: The evolution of hail production in simulated supercell storms. J. Atmos. Sci., 78, 3417–3440, https://doi.org/10.1175/JAS-D-21-0034.1.
Kuster, C. M., J. C. Snyder, T. J. Schuur, T. T. Lindley, P. L. Heinselman, J. C. Furtado, J. W. Brogden, and R. Toomey, 2019: Rapid-update radar observations of ZDR column depth and its use in the warning decision process. Wea. Forecasting, 34, 1173–1188, https://doi.org/10.1175/WAF-D-19-0024.1.
Lang, T., B. Dolan, N. Guy, C. A. M. Gerlach, and J. Hardin, 2019: CSU-Radarmet/CSU_RadarTools: CSU_RadarTools v1.3. Zenodo, accessed 30 May 2022, https://doi.org/10.5281/zenodo.2562063.
Loeffler, S. D., and M. R. Kumjian, 2020: Idealized model simulations to determine impacts of storm‐relative winds on differential reflectivity and specific differential phase fields. J. Geophys. Res. Atmos., 125, e2020JD033870, https://doi.org/10.1029/2020JD033870.
McCarthy, N., H. McGowan, A. Guyot, and A. Dowdy, 2018: Mobile X-pol radar: A new tool for investigating pyroconvection and associated wildfire meteorology. Bull. Amer. Meteor. Soc., 99, 1177–1195, https://doi.org/10.1175/BAMS-D-16-0118.1.
Montopoli, M., E. Picciotti, L. Baldini, S. Di Fabio, F. S. Marzano, and G. Vulpiani, 2021: Gazing inside a giant-hail-bearing Mediterranean supercell by dual-polarization Doppler weather radar. Atmos. Res., 264, 105852, https://doi.org/10.1016/j.atmosres.2021.105852.
Pauley, P. M., and X. Wu, 1990: The theoretical, discrete, and actual response of the Barnes objective analysis scheme for one- and two-dimensional fields. Mon. Wea. Rev., 118, 1145–1164, https://doi.org/10.1175/1520-0493(1990)118<1145:TTDAAR>2.0.CO;2.
Picca, J., and A. Ryzhkov, 2012: A dual-wavelength polarimetric analysis of the 16 May 2010 Oklahoma City extreme hailstorm. Mon. Wea. Rev., 140, 1385–1403, https://doi.org/10.1175/MWR-D-11-00112.1.
Potvin, C. K., A. Shapiro, and M. Xue, 2012: Impact of a vertical vorticity constraint in variational dual-Doppler wind analysis: Tests with real and simulated supercell data. J. Atmos. Oceanic Technol., 29, 32–49, https://doi.org/10.1175/JTECH-D-11-00019.1.
Punge, H. J., K. M. Bedka, M. Kunz, and A. Reinbold, 2017: Hail frequency estimation across Europe based on a combination of overshooting top detections and the ERA-INTERIM reanalysis. Atmos. Res., 198, 34–43, https://doi.org/10.1016/j.atmosres.2017.07.025.
Reimel, K. J., and M. Kumjian, 2021: Evaluation of KDP estimation algorithm performance in rain using a known-truth framework. J. Atmos. Oceanic Technol., 38, 587–605, https://doi.org/10.1175/JTECH-D-20-0060.1.
Ryzhkov, A. V., 2007: The impact of beam broadening on the quality of radar polarimetric data. J. Atmos. Oceanic Technol., 24, 729–744, https://doi.org/10.1175/JTECH2003.1.
Ryzhkov, A. V., and D. S. Zrnić, Eds., 2019: Data quality and measurement errors. Radar Polarimetry for Weather Observations, Springer, 153–156.
Ryzhkov, A. V., M. R. Kumjian, S. M. Ganson, and A. P. Khain, 2013: Polarimetric radar characteristics of melting hail. Part I: Theoretical simulations using spectral microphysical modeling. J. Appl. Meteor. Climatol., 52, 2849–2870, https://doi.org/10.1175/JAMC-D-13-073.1.
Schleiss, M., and Coauthors, 2020: The accuracy of weather radar in heavy rain: A comparative study for Denmark, the Netherlands, Finland and Sweden. Hydrol. Earth Syst. Sci., 24, 3157–3188, https://doi.org/10.5194/hess-24-3157-2020.
Schultz, C. J., W. A. Petersen, and L. D. Carey, 2009: Preliminary development and evaluation of lightning jump algorithms for the real-time detection of severe weather. J. Appl. Meteor. Climatol., 48, 2543–2563, https://doi.org/10.1175/2009JAMC2237.1.
Schulz, W., G. Diendorfer, S. Pedeboy, and D. R. Poelman, 2016: The European lightning location system EUCLID—Part 1: Performance analysis and validation. Nat. Hazards Earth Syst. Sci., 16, 595–605, https://doi.org/10.5194/nhess-16-595-2016.
Shapiro, A., C. K. Potvin, and J. Gao, 2009: Use of a vertical vorticity equation in variational dual-Doppler wind analysis. J. Atmos. Oceanic Technol., 26, 2089–2106, https://doi.org/10.1175/2009JTECHA1256.1.
Snyder, J. C., H. B. Bluestein, V. Venkatesh, and S. J. Frasier, 2013: Observations of polarimetric signatures in supercells by an X-Band mobile Doppler radar. Mon. Wea. Rev., 141, 3–29, https://doi.org/10.1175/MWR-D-12-00068.1.
Snyder, J. C., A. V. Ryzhkov, M. R. Kumjian, A. P. Khain, and J. Picca, 2015: A ZDR column detection algorithm to examine convective storm updrafts. Wea. Forecasting, 30, 1819–1844, https://doi.org/10.1175/WAF-D-15-0068.1.
Snyder, J. C., H. B. Bluestein, D. T. Dawson II, and Y. Jung, 2017: Simulations of polarimetric, X-Band radar signatures in supercells. Part II: ZDR columns and rings and KDP columns. J. Appl. Meteor. Climatol., 56, 2001–2026, https://doi.org/10.1175/JAMC-D-16-0139.1.
Soderholm, J., H. McGowan, H. Richter, K. Walsh, T. Weckwerth, and M. Coleman, 2016: The Coastal Convective Interactions Experiment (CCIE): Understanding the role of sea breezes for hailstorm hotspots in eastern Australia. Bull. Amer. Meteor. Soc., 97, 1687–1698, https://doi.org/10.1175/BAMS-D-14-00212.1.
Taszarek, M., J. T. Allen, T. Púčik, K. A. Hoogewind, and H. E. Brooks, 2020: Severe convective storms across Europe and the United States. Part II: ERA5 environments associated with lightning, large hail, severe wind, and tornadoes. J. Climate, 33, 10 263–10 286, https://doi.org/10.1175/JCLI-D-20-0346.1.
Testud, J., E. Le Bouar, E. Obligis, and M. Ali-Mehenni, 2000: The rain profiling algorithm applied to polarimetric weather radar. J. Atmos. Oceanic Technol., 17, 332–356, https://doi.org/10.1175/1520-0426(2000)017<0332:TRPAAT>2.0.CO;2.
Trier, S. B., and R. D. Sharman, 2018: Trapped gravity waves and their association with turbulence in a large thunderstorm anvil during PECAN. Mon. Wea. Rev., 146, 3031–3052, https://doi.org/10.1175/MWR-D-18-0152.1.
Van Den Broeke, M. S., 2020: A Preliminary polarimetric radar comparison of pretornadic and nontornadic supercell storms. Mon. Wea. Rev., 148, 1567–1584, https://doi.org/10.1175/MWR-D-19-0296.1.
Vulpiani, G., L. Baldini, and N. Roberto, 2015: Characterization of Mediterranean hail-bearing storms using an operational polarimetric X-band radar. Atmos. Meas. Tech., 8, 4681–4698, https://doi.org/10.5194/amt-8-4681-2015.
Weissmann, M., F. J. Braun, L. Gantner, G. J. Mayr, S. Rahm, and O. Reitebuch, 2005: The Alpine Mountain–Plain circulation: Airborne Doppler lidar measurements and numerical simulations. Mon. Wea. Rev., 133, 3095–3109, https://doi.org/10.1175/MWR3012.1.
