Study of Microphysical Signatures Based on Spectral Polarimetry during the RELAMPAGO Field Experiment in Argentina
1. Introduction
Severe atmospheric conditions can give rise to extreme weather events such as severe hailstorms, tornadoes, flash floods, and others that have a significant impact on the socioeconomic well-being of people around the world. Therefore, it is crucial to study and monitor severe weather so as to minimize the impact of natural disasters and save lives. Thunderstorms originating in midlatitude South America, stand out in satellite observations as being stronger than anywhere else on Earth. These storms are accompanied by gargantuan-sized hails, tall convective cores extending above 10-km altitude, and widespread stratiform regions as well as tend to exhibit a higher lightning flash rate (Salio et al. 2007; Anabor et al. 2008; Rasmussen and Houze 2011, 2016). Various studies have examined deep convection environments in South America that lead to the formation of extreme supercell events and mesoscale convective systems (MCS) that cause damage and flash flooding (Salio et al. 2007; Anabor et al. 2008; Rasmussen and Houze 2011, 2016; Mulholland et al. 2019; Trapp et al. 2020; Borque et al. 2020). Satellite observations have also revealed that these thunderstorms predominantly occur in the vicinity of mountain ranges in northern and central Argentina, emphasizing the significant role of complex terrain in precipitation enhancements. All these together emphasize the need for studying various microphysical processes and land–atmosphere interactions that occur within convective precipitation systems. These studies will ultimately facilitate better forecasts and develop situational awareness of these weather events.
A study involving a comprehensive and well-coordinated set of in situ and remote sensing measurements is crucial for a thorough understanding of various microphysical processes that regulate the life cycle of a convective storm. It is also essential that these measurements are performed across a wide range of precipitation events. Thus, the collocated measurements acquired from the deployment of various observation instruments from field experiments are regarded as a proficient method of conducting comprehensive studies. For this reason, the Remote Sensing of Electrification, Lightning, and Mesoscale/Microscale Processes with Adaptive Ground Observations (RELAMPAGO) field campaign was conducted in Cordoba, located in central west Argentina in the vicinity of the Sierras de Cordoba (Nesbitt et al. 2021). The campaign took place from 1 June 2018 until 30 April 2019. Some important research goals of the campaign included a better understanding of microphysical processes related to convective initiation, upscale growth, and the mesoscale organization of convective storms. Therefore, instruments during RELAMPAGO were able to successfully observe and quantify a large number of weather events along with their associated synoptic conditions. In this study, the focus is on the spectral polarimetric analysis performed based on observations by the Colorado State University C-Band Hydrometeorological Instrument for Volumetric Observation (CSU-CHIVO) radar that was deployed during the RELAMPAGO field campaign. The key objective is to study the signatures of microphysical processes and storm dynamics based on spectral polarimetric variables at different locations and times within a storm system. The radar data from range–height indicator (RHI) scans collected during various precipitation events have been used.
Dual-polarization weather radar has already proven to be one of the most valuable remote sensing instruments for accurate observation of various precipitation events. Measurements from dual-polarization weather radars have shown great potential for studying precipitation as polarimetric variables are sensitive to the shape, size, orientation, and composition of hydrometeors (Bringi and Chandrasekar 2001). Polarimetric radar measurements, such as differential reflectivity Zdr, copolar correlation coefficient ρhv, and specific differential phase Kdp have been widely used for precipitation quantification and various microphysical properties retrievals. Some examples include attenuation correction, quantitative precipitation estimation, hydrometeor classification, particle size distribution retrieval, and identification of various ice habits and their growth processes (Kumjian and Ryzhkov 2012; Bechini and Chandrasekar 2015; Andrić et al. 2013; Kedzuf et al. 2021; Biswas et al. 2024; Cifelli and Chandrasekar 2010; Biswas et al. 2020; Bechini et al. 2013; Montopoli et al. 2021). Furthermore, the polarimetric measurements along with Doppler information enable spectral analysis to characterize the dynamics of hydrometeors in a precipitation system as well as the storm kinematics. This type of study requires the polarimetric variables to be expressed as a function of Doppler velocities of individual scatterers in a radar resolution volume (Bringi and Chandrasekar 2001; Doviak et al. 2006; Wang et al. 2019; Yu et al. 2012). This study of spectral properties in conjunction with polarimetric measurements is commonly referred to as spectral polarimetry. A combination of these spectral polarimetric parameters can provide crucial insights into the kinematic characteristics of individual hydrometeor scatterers that are otherwise challenging to obtain through conventional analysis methods as per Keat and Westbrook (2017) and Wang et al. (2019). One such aspect to study is the presence of various particle size distributions resulting from a mixture of hydrometeors in a sheared environment and its implication on spectral polarimetric variables.
In the presence of vertical wind shear, the velocity of hydrometeors observed by the radar is determined by both the gravitational fall velocity and the background shear-induced radial velocity component. Shear-induced size sorting due to vertical wind shear has been investigated by Wang et al. (2019). The authors studied a hailstorm event at a radar scan elevation angle of 0° where they reported similar values of Zh and Zdr for different locations in a storm that were characterized by different spectral differential reflectivity (sZdr) slopes. In addition to determining size sorting, spectral polarimetric variables have also been used to detect ice habits and ice growth processes in deep mixed-phase precipitating clouds. In these precipitation systems pristine ice crystals have a high aspect ratio and fall with their major axis aligned horizontally resulting in high Zdr values. However, mixed-phase precipitation often contains aggregates or irregularly shaped polycrystals alongside pristine crystals, making it challenging and ambiguous to interpret Zdr and accurately detect pristine crystals. In this context, the study by Keat and Westbrook (2017) used Doppler power spectrum measurements from a 35-GHz dual-polarization weather radar to identify the presence of pristine crystals mixed with aggregates in mixed-phase clouds. The researchers observed a bimodal pattern in the Doppler power spectrum at certain altitudes, which was attributed to the coexistence of slow-falling pristine ice crystals and fast-falling aggregates within the same resolution volume. With altitude reduction, they found that the spectral power and corresponding Doppler velocity increased gradually. This behavior was attributed to the increasing mass of hydrometeors due to the depositional growth of pristine crystals and later aggregation. Similar analyses of spectral polarimetric variables at the S band have been performed to categorize particle habits and orientations in mixed-phase precipitation systems by Dufournet and Russchenberg (2011). They have categorized different groups of particle orientations based on the slope of the sZdr as well as the corresponding values. Positive sZdr values have been shown to be due to horizontally oriented ice particles, whereas negative sZdr values are due to vertically oriented particles. On the contrary, sZdr values of the spherical ice particles, such as aggregates, graupel, and hail, are approximately 0 dB. Proceeding further, Pfitzenmaier et al. (2018) have used spectral polarimetric variables to observe the growth of ice particles along fall streaks in mixed-phase clouds. They detected three different growth processes of ice crystals, namely, riming, aggregation, and diffusion growth, while relating them to the presence of supercooled liquid water.
Previous studies, as mentioned earlier, have utilized data obtained from either zenith-pointing radar scans or measurements from elevation angles 0° and 45°. Zenith angle measurements can be very useful in determining particle fall velocities while polarimetric measurements are maximized at near zero elevation whereas measurements from 45° achieve a balance between the two. However, it is necessary to study spectral polarimetric measurements at various altitude levels to accurately characterize the microphysical and dynamical properties of a precipitation system. With this aim, spectral polarimetric variables have been computed at various ranges and elevation angles (corresponding to different altitudes) within a precipitation system, where diverse hydrometeor growth processes and storm dynamics occur. The current study also examines a combination of spectral polarimetric variables, as well as their distributions across different altitudes/regions of the precipitation system. This approach is necessary to accurately depict the vertical distribution of different hydrometeor types within a precipitation system. Additionally, the temporal evolution of a storm can be analyzed based on changes in spectral polarimetric variables over time. Thus, this scientific analysis also presents novel research on the temporal progression of spectral polarimetric variables at regions of interest within a storm.
The paper is organized as follows: Section 2 discusses the RELAMPAGO field campaign along with details of the CSU-CHIVO radar and the study cases selected for analysis. In addition, this section also details the methodology adopted for spectral polarimetric analysis. Analysis results are reported in section 3 and the section is divided into two major parts. Properties of spectral polarimetric variables corresponding to storm kinematics are presented in the first part while the second half discusses the spectral polarimetric signatures for various microphysical processes pertinent to convective and stratiform precipitation. In addition, an analysis of the temporal evolution of spectral polarimetric signatures in a precipitation system is also presented. Section 4 summarizes the paper and provides a discussion of the key findings.
2. C-band radar and RELAMPAGO dataset
The RELAMPAGO field experiment was conducted from 1 June 2018 until 30 April 2019 in Cordoba, in the vicinity of the Sierras de Cordoba located in central western Argentina (Nesbitt et al. 2021). The main goal was the study of different phases of the life cycle of thunderstorms, especially MCSs, which are known to be very frequent and intense in Argentina in comparison with the rest of the world. From a scientific perspective, the campaign was carried out in two phases, namely, an extended hydrometeorology observing period (EHOP) from 1 June 2018 to 30 April 2019 and an intensive observing period (IOP) from 1 November 2018 to 16 December 2018. During the IOP, some of the tallest thunderstorms in the world were observed and these extended up to 20 km in altitude from the mean sea level. These convective systems were associated with very strong updrafts and wind shear and produced hail up to 6 in. (∼15 cm) in diameter (Nesbitt et al. 2021; Schumacher et al. 2021; Trapp et al. 2020). The CSU-CHIVO C-band dual-polarization Doppler weather radar was deployed along with various other ground-based instruments (during this campaign) near the Andes Mountains region in Cordoba, Argentina. The other instruments deployed along with CSU-CHIVO included one C-Band (5-cm wavelength) on Wheels (COW) portable radar, a network of three X-band (3-cm wavelength) Doppler-on Wheels (DOW) mobile radars, six mobile radiosonde systems, three “mobile mesonet” vehicles (MMs) also tasked with the deployment of up to 12 portable surface weather stations (Pods) and up to four disdrometers, and hail pads (Trapp et al. 2020).
a. The CSU-CHIVO radar
The CSU-CHIVO is a research radar maintained and operated by Colorado State University during the course of the campaign. It was located south of Cordoba city at a location 421 m above mean sea level with coordinates of 31.63°S latitude and 64.17°W longitude. Figure 1 shows a picture of the CSU-CHIVO radar along with its deployment location. A brief summary of the technical details of the radar is presented in Table 1. CSU-CHIVO reflectivity measurements have been compared and calibrated with respect to the Dual-Frequency Precipitation Radar on board the Global Precipitation Measurement (GPM) mission core satellite. The comparison and calibration methodology is described in Biswas and Chandrasekar (2018) and is performed as part of the quality control process of the data. A comprehensive study of the CSU-CHIVO radar calibration can be found in Arias and Chandrasekar (2021).
(a) The CSU-CHIVO C-band dual-polarization weather radar. (b) The CSU-CHIVO radar deployment location during the RELAMPAGO field experiment. The black circles correspond to radar range radii of 50, 100, and 150 km, going outward.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
Technical specifications of the CSU-CHIVO radar.
The radar adopted several scan strategies such as volume plan position indicator (PPI) 360° surveillance scans, sector scans, and RHI scans. Different scan strategies named HYDRO360, PFAR360, PNL360A, and RHI45 (V. Chandrasekar et al. 2019, unpublished manuscript) were used to sample different phenomena from convective initiation to upscale growths. HYDRO360 is a low-level 360° surveillance scan at elevation angles of 0.5° and 1.5°. PFAR360 is a full-volume 360° PPI scan at 16 elevation angles between 0.5° and 16.0° whereas PNL360A is a full-volume 360° PPI scan performed at 19 elevation angles between 0.5° and 28.7°. The RHI45 scans were performed at azimuth angles 0° to 360° and elevation angles 0° to 45°. The number of RHI elevation scan angles and the azimuth angles were selected based on the storm cells and their location with respect to the radar. In this study, RHI scans have been chosen over PPI scans so as to observe the full extent of precipitation vertical structure without any observational discontinuity.
b. Radar observations of precipitation events
1) Case study I: 30 November 2018
A precipitation event from 30 November 2018 is chosen as the first case study. This event started on 29 November 2018 as an upper-level trough that crossed Argentina providing favorable conditions for a deep convective initiation over the high terrains of the Sierras de Cordoba. An MCS formed overnight from merging convective cells to the west that gradually moved over the Cordoba region giving rise to a weaker widespread stratiform precipitation event at around 0300 UTC 30 November 2018. The CSU-CHIVO radar was able to capture this event and dual-polarization radar observations from an RHI scan at 240° azimuth at 0337 UTC (30 November 2018) are presented in Fig. 2 where radar Zh and Zdr are corrected for attenuation in the liquid precipitation region. During this precipitation event, the environmental freezing level was at an altitude of approximately 3 km where the radar Zh shows a typical brightband signature with a corresponding drop in Zdr and ρhv. The corresponding measured Doppler velocity dealiasing is performed based on a region-based algorithm using the Python ARM Radar Toolkit (Py-ART) (Helmus and Collis 2016). Note that negative velocities correspond to moving toward the radar while positive values indicate outbound velocities.
Observations of (top left) radar reflectivity, (top right) differential reflectivity, (bottom left) copolar correlation coefficient, and (bottom right) radial velocity from the CSU-CHIVO radar 240° RHI scan at 0337 UTC 30 Nov 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
In the uniform rain region below the melting layer, one can observe values of Zh above 35–40 dBZ corresponding to raindrops. For the same region, Zdr values are nearly 1–2 dB, indicating predominantly large particles dominating the sampling volume, whereas ρhv shows values around 1 because of the high homogeneity. Above the melting layer, there exists deep ice/mixed-phase precipitation at a distance of 10–50 km from the radar and the corresponding reflectivity and differential reflectivity values are in the range from −20 to 20 dBZ and from 0 to 2.5 dB, respectively. In this mixed-phase precipitation ρhv shows high values of 0.95–1.
Furthermore, as depicted in Fig. 2, anvil clouds at a distance of 72 km and at altitudes of 5–8 km from the radar exhibit Zh values in the range of 5–20 dBZ and Zdr values of approximately 2–2.5 dB. A detailed characterization/analysis of the anvil clouds is given in section 3b. The above radar polarimetric variables are used for the hydrometeor classification algorithm, that is, DROPS2 (Chen et al. 2017), and the results are shown in Fig. 3. This precipitation is observed to have vertical wind shear, various microphysical and ice growth processes with unique polarimetric signatures, as well as anvil clouds with enhanced Zdr at altitudes of 5–8 km in the trailing edge of stratiform precipitating clouds. However, the polarimetric variables change significantly over time due to various dynamical and microphysical processes in mixed-phased precipitation. These dynamical and microphysical processes at 0300–0430 UTC have been discussed and analyzed using spectral analysis in section 3.
Hydrometeor classification results for the same event as in Fig. 2.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
2) Case study II: 14 December 2018
A severe convection was initiated in the east of the Cordoba region that moved into the field of view of the radar at around 0026 UTC 14 December 2018 during the IOP17 period. The CSU-CHIVO radar captured some intense convective cells that later formed into supercells. These supercells were associated with hail, updrafts, and strong wind shear with cloud tops reaching up to 15 km above the ground. Observations from radar RHI scan at approximately 0206 UTC and at an azimuth of 160° is presented in Fig. 4. The core of the storm can be seen to extend up to 8 km in height with Zh values reaching up to 70 dBZ along with high Zdr values in the range of 6–8 dB and a corresponding drop in ρhv. This is indicative of hail formation aloft and is corroborated by the hydrometeor classification results presented in Fig. 5. The melting layer boundary can be observed at around 5 km where hail and graupel melted into rain–hail mixtures producing heavy rain and large drops. The environmental freezing level height was also verified using radiosonde data at the Atmospheric Radiation Measurement (ARM) program site in Cordoba.
As in Fig. 2, but for the 160° RHI scan of a convective storm event at 0206 UTC 14 Dec 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
As in Fig. 3, but for the same event as in Fig. 4.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
c. Spectral polarimetry
3. Characterization of microphysical and dynamical properties of hydrometeors in storms
a. Spectral signatures of hydrometeor dynamics
The dynamical characteristics of hydrometeors are influenced by their size and weight, as well as gravity and background wind velocity. These dynamical characteristics can result in a range of unique spectral polarimetric features that can be used to study and characterize the underlying processes. In this section, shear-induced size sorting and associated spectral characterization in a column of high inbound velocity along the depth of precipitation are discussed.
1) Shear-induced size sorting of hydrometeors
The mass and density of hydrometeors are related to particle size as well as diameter, which in turn influence its terminal fall velocity. However, the trajectories of hydrometeors in a dynamic storm environment are not only influenced by the terminal fall velocity but also by the kinematics of the precipitation system (Marshall 1953; Spek et al. 2008; Dawson et al. 2015; Kumjian and Ryzhkov 2012). It has been shown by Pinsky and Khain (1994) that the approximate trajectory of a hydrometeor can be related to wind shear. Vertical wind shear occurs when the horizontal wind velocity changes with altitude and the resulting vertical gradient in the horizontal wind velocity cause hydrometeors of different sizes to be sorted in space (Wang et al. 2019; Dawson et al. 2015; Kumjian and Ryzhkov 2012; Biswas et al. 2022). This study uses radar observations at elevation angles 0°–5° to detect the occurrence of shear-induced size sorting in different types of precipitation systems. From a microphysics perspective, two different regions of interest are selected—a region of graupel in a stratiform event and a region of raindrops consisting of different size distributions in a convective event.
(i) Case I—Stratiform precipitation
The stratiform precipitation event discussed in section 2 is considered first. A region at a distance of 45 km from the radar and at an altitude range of 2.9–3.3 km is chosen, which has graupel according to the hydrometeor classification shown in Fig. 3. The presence of vertical wind shear in this region is determined based on the range profiles of radial velocities at elevation angles of 3.7°, 4.0°, and 4.2° as shown in Fig. 6. The magnitude of the shear is given as the ratio of the difference in velocity estimates to the difference in altitude. The vertical wind shear is determined to be approximately 0.02 s−1 with a direction toward the radar, which by convention is taken as negative. Spectral polarimetric variables are calculated in this region following the methodology discussed in section 2, and the results are shown in Fig. 7. The sZh and sZdr are presented as a function of radial velocities where the maximum unambiguous velocity of the radar is 13.625 m s−1. It can be observed that the maximum values of sZh are in the range of 10–25 dB for altitudes 2.7–3.3 km, whereas the values of sZdr are in the range from −2 to 2 dB.
Range profile of radial velocity at three nearby elevation angles of 3.7°, 4.0°, and 4.2°. The observations correspond to CSU-CHIVO’s RHI scan at 240° azimuth angle at 0337 UTC 30 Nov 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
(left) Spectral reflectivity and (right) spectral differential reflectivity at (bottom) 2.7-, (bottom middle) 2.9-, (top middle) 3.1-, and (top) 3.3-km altitudes and at a distance of 45 km from the radar. The observations correspond to CSU-CHIVO RHI scan at 240° azimuth angle at 0337 UTC 30 Nov 2018. The sZdr slope is indicated by fitted dashed lines in the sZdr plots, and the value of the slope [dB (m s−1) −1] is also shown.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
From the results, it is clear that sZh shows characteristics of a non-Gaussian distribution with a fairly broad and flat-topped spectrum at all altitudes. This is because turbulence or wind shear causes random motion of hydrometeors in a radar resolution volume, resulting in a range of radial velocities and thereby increasing the spectrum width. Similar results have been reported by Yu et al. (2009) and Wang et al. (2019) in strongly sheared environments. Furthermore, sZdr distribution has a negative slope that is confirmed by a linear fit. The gradient increases from −0.03 dB (m s−1)−1 at 2.7 km to −0.35 dB (m s−1)−1 at 3.3 km. These negative sZdr slopes indicate hydrometeor size sorting, that is, hydrometeors of different sizes and/or shapes have different radial velocities, with particles of sZdr values above 0 dB are moving with higher radial velocity, while particles of sZdr values below 0 dB moving with lower radial velocity. This size sorting of hydrometeors is due to the presence of vertical wind shear as the horizontal wind has a significant contribution toward the radial velocity at elevation angles less than 4°. Size sorting results in a reduction in the number of smaller particles from top to bottom of a shear-induced region, with smaller particles being almost absent at the bottom even though ice particles both small and large in size can be found at the top of the sheared region. This type of particle size distribution can result in a negative sZdr slope similar to that seen at altitudes 3.3, 3.1, and 2.9 km. The particles with sZdr values below 0 dB are conical graupel and those with sZdr values of around 0 dB are graupel.
(ii) Case II—Convective precipitation
In this study, the size sorting due to vertical wind shear has been analyzed for a convective storm event using RHI scan observations at 0206 UTC 14 December 2018. The particular observations at an azimuth angle of 160° are shown in Fig. 4 and the presence of vertical wind shear can be inferred from a tilt in the Zdr column located at a distance of 35–40 km from the radar. Moreover, the vertical wind shear has been confirmed based on radial velocities measurements by the radar.
The sZh and sZdr have been calculated at altitudes of 1.5, 2, and 2.5 km (using elevation angles of 2.5°, 3.3°, and 4°, respectively) and at a distance of 35 km from the radar and are shown in Fig. 8. The sZh maximum values are in the range of 22–30 dB, and sZh spectrum is non-Gaussian and broad at 2–2.5 km, but it becomes narrower as the altitude decreases to 1.5 km. The values of sZdr range from 2 to 8 dB at altitudes of 1.5–2.5 km where there is liquid-phase precipitation (as per hydrometeor classification). Large drops from fully or partially melted hail have resulted in sZdr values of around 6 dB at 2 km. In this analysis, sZdr spectrum shows a positive slope of 0.455 and 0.57 dB (m s−1)−1 at 2.5- and 2-km altitudes, respectively, whereas the slope at 1.5 km is 0.18 dB (m s−1)−1. It can be observed that the slope changes from around 0.5 to 0.2 dB (m s−1)−1 with decreasing altitude. The different slopes of the sZdr spectrum at nearby altitudes indicate that the particle shape and size distributions are different in the respective radar resolution volume.
As in Fig. 7, but at (bottom) 1.5-, (middle) 2-, and (top) 2.5-km altitudes and at a distance of 35 km from the radar, corresponding to an RHI scan at 160° azimuth angle at 0206 UTC 14 Dec 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
These sZdr slopes also indicate the size sorting of hydrometeors (raindrops) in the radar resolution volume because of the presence of vertical wind shear at these locations. Size sorting results in a decrease in the number of small drops from the top to the bottom of a resolution volume with significant removal of small raindrops at the bottom (Kumjian and Ryzhkov 2012). This size-sorting and sZdr slope relationship is also corroborated based on theoretical simulations as well as experimental results in Wang et al. (2019). The relative size of the particles can be distinguished in the sZdr spectrum with respect to the radial velocities. The relatively small-sized drops are at higher radial velocities than larger and more oblate drops with higher sZdr values. This is because of the lower elevation angles of observation where the horizontal wind speed has a significant contribution to the radial velocity and the smaller drops are easily advected by the wind. The range of sZdr values and slope changes with decreasing altitude indicating the change in drop size distribution possibly due to processes like coalescence and breakup as well as evaporation in liquid-phase precipitation. It has been proven by Wang et al. (2019) that change in sZdr slopes could also be used to classify the types of particles and determine the change in the particle size distribution.
The impact of vertical wind shear profile on the trajectories of raindrops of different sizes have been analyzed by previous studies and have shown that bigger raindrops have steep trajectories while smaller drops are carried a long distance in the wind before they reach the ground. Similarly, the trajectories of different ice particle types (of different sizes in terms of their maximum dimension) in a sheared wind profile have been analyzed and presented in Fig. 9. As per the results the small and light platelike pristine ice crystals (1 mm) are easily advected by the wind whereas bigger particles like aggregates (5 mm) and graupel (8 mm) have steeper trajectories in a sheared wind profile. Thus, vertical wind shear can cause particles of different sizes in the same radar resolution volume to move with different velocities and is responsible for linear sZdr slopes for the observations at low elevation angle.
(a) Wind profile (showing vertical wind shear). (b) Trajectories of ice particles of different sizes for the wind profile shown in (a).
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
2) Characterization of ice crystals near melting layer
This analysis of the hydrometeor dynamics deals with spectral characterization in a column of high inbound velocity. Crystals were observed close to the melting layer (as per hydrometeor classification), at 0357 UTC 30 November 2018, in the RHI scan at an azimuth angle of 240°. This RHI scan polarimetric variables observed by the CSU-CHIVO radar are shown in Fig. 10 and the corresponding hydrometeor classification is depicted in Fig. 11. In the RHI scan, the region at a distance of 33 km from the radar and just above the melting layer is characterized by a pocket of low reflectivity values (in comparison with the surrounding reflectivity values) along with low ρhv. Moreover, an inversion in the reflectivity profile can be noticed at this specific location, as depicted in Fig. 12. Specifically, close to the melting layer Zh exhibits values in the range from −5 to 20 dBZ, whereas the Zdr values are in the range from −2 to 3 dB and ρhv is between 0.7 and 1. The Zh values are as low as −1 dBZ—that is, 20–25 dB lower than that of the neighboring range gates—resulting in a slope of approximately 23 dBZ km−1. This significant decrease in reflectivity within a 1-km increase in altitude has been analyzed. Similar to the sharp change in the gradient of Zh values, there is a sudden drop in ρhv to 0.85. This region of low Zh with crystals is anomalous as crystals are not formed near the melting layer because temperatures close to 0°C are not conducive to crystal nucleation. This region of crystals was also observed at the edge of an X-band radar (collocated with the C-band radar) hydro classification at approximately 25 km from the radar.
RHI plots of (top left) reflectivity, (top right) differential reflectivity, (bottom left) copolar correlation coefficient, and (bottom right) radial velocity at 240° azimuth angle at 0357 UTC 30 Nov 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
As in Fig. 3, but for the same event as in Fig. 10.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
Vertical profile of Zh, temperature, relative humidity, copolar correlation coefficient ρhv, and Doppler velocity Vd at 33-km distance from the CSU-CHIVO radar, with the background colors depicting the hydrometeor classes. The environmental temperature and relative humidity profiles are obtained from the ARM radiosonde station in Cordoba.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
Therefore, spectral polarimetric variables associated with this phenomenon have been computed and analyzed using observations from 5.2°, 6.2°, and 7.76° elevation angles of the RHI scan, and the results are presented in Fig. 13. It can be observed that maximum sZh values are in the range from −10 to 7 dB, whereas the sZdr values are primarily in the range of 1–2 dB, and the sρhv values are in the range of 0.84–0.99 for all three altitudes. These spectral variables indicate that the particles present are of very small size and are primarily horizontally oriented. Furthermore, it can be observed that the peak sZh has a significant gradient of −6.5 and 17 dB km−1 as the altitude decreases from 4.5 to 3 km. This decrease in spectral power with a decrease in altitude rules out the possibility of ice crystal growth by aggregation or riming; instead, it indicates the presence of an ice crystal population. However, there is no possibility of ice nucleation resulting in small ice particles due to the warm temperatures near the melting layer (at 2.5–3 km). Therefore, the presence of crystals is due to two mechanisms, that is, the transport of small particles into this region from upper levels and the advection of rime splinters from neighboring riming zones. These crystals have been transported to this region by wind flow that can be inferred because of the increase in the Doppler velocity from −4 m s−1 at 4.5 km to −12 m s−1 at 3.6-km altitude. A rearward wind flow in this region is also confirmed from Fig. 10.
(left) Spectral reflectivity, (center) spectral differential reflectivity, and (right) copolar coherency spectrum estimated at altitudes of (bottom) 3, (middle) 3.6, and (top) 4.2 km at a distance of 33 km from the radar. The observations correspond to CSU-CHIVO’s RHI scan at 240° azimuth angle at 0357 UTC 30 Nov 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
Here, reduced sZh and enhanced sZdr at high radial velocities at low elevation angles indicate ice crystals as the dominant ice habit type in the radar resolution volume. Along with that, the lowering of the copolar coherency spectrum together with sZdr slope and broad power spectrum indicates the presence of a combination of vertical wind shear and turbulence. This is confirmed by the velocity of horizontal wind in this region, which has a very high gradient of 10 m s−1 in 1 kilometer in the vertical direction. This turbulence results in the mixed orientation of ice crystals in the radar resolution volume. The hydrometeor classification (Fig. 11) shows the presence of a fall streak of crystals in this region that is devoid of dry snow or graupel indicating the absence of ice growth processes like aggregation and riming. This is again confirmed by the vertical profile of relative humidity (Fig. 12) that shows a drop to 40% at altitudes near and above the melting layer. This is associated with a dry environment as determined by the atmospheric temperature profile shown in Fig. 12 at this altitude. The high velocity and low number concentration (due to low sZh) of crystals in the fall streak together with a dry environment do not favor ice growth processes by aggregation or riming. The dry environment might have resulted from cooling by sublimation (at temperature < 0°C above the melting layer) and melting (at temperature > 0°C below the melting layer). The presence of a dry environment can result in the sinking of air where a narrow column of high inbound velocity (approximately −20 m s−1) exists in the entire depth of the precipitation. Vertical wind shear and turbulence can act to support this column of inbound velocity. These dynamical and microphysical features are usually observed at the edge of a low-level inflow jet associated with stratiform precipitation as documented in Smith et al. (2009) and Grim et al. (2009). The presence of a rearward inflow jet is well depicted in the RHI plots of radial velocities in Fig. 10. At the edge of this low-level inflow jet, a narrow column of high inbound velocity (approximately −20 m s−1) can also be observed.
Therefore, the decrease in sZh with altitude, sZdr slope, broadened spectrum, and lowering of copolar coherency spectrum are the spectral signatures associated with a column of high inbound velocity at the edge of a low-level inflow jet in stratiform precipitation. This is associated with a dry environment resulting from the cooling by melting and sublimation as well as the presence of vertical wind shear and turbulence. In addition to that, ice splinters from neighboring riming regions might be advected to this region in strong negative velocity due to the rearward flow. Small ice splinters are created during graupel formation, and this secondary ice production can occur by different mechanisms; the common among them is rime splintering (Hallett and Mossop 1974; Kalesse et al. 2016; Oue et al. 2018; Pfitzenmaier et al. 2018; Hogan et al. 2002). This advection of ice crystals increases its number concentration and as it descends to the warmer temperatures near the melting layer results in enhanced sZh due to melting.
b. Spectral signature of particle microphysical growth process
Depending on the state of the atmosphere and storm dynamics, various microphysical growth processes exist in different regions of precipitation, resulting in unique spectral polarimetric signatures. Accordingly, the spectral signatures of two different hydrometeor growth processes, namely, aggregation and riming, have been studied here.
1) Aggregation growth process in anvil clouds
This study deals with the aggregation growth process in anvil clouds, which are high-topped clouds extending outward from the raining cores of MCSs. The cloud base is at least higher than 3 km and a lower precipitating cloud may or may not exist. The analysis of the internal structures of these long-lasting and nonprecipitating anvil clouds is important for understanding the microphysical processes within the clouds. These anvil clouds at the trailing end of stratiform precipitation have some unique characteristics such as depositional growth and aggregation of ice crystals along with charge production by collision (Cetrone and Houze 2011; Smith et al. 2009; Dye and Bansemer 2019). This charge production can cause the vertical orientation of the ice particles resulting in negative values of Zdr. The polarimetric variables in anvil clouds are shown in Fig. 2 and the vertical profiles of Zh and Zdr (at a distance of 72 km from the radar) in the anvil cloud region are shown in Fig. 14. The Zh values are in the range of 7–15 dBZ and are inversely related to the altitude range of interest, that is, 4–8 km. The Zdr values are in the range of 0–3.5 dB, and the values reduce from 3.5 dB at 8 km to 0.5 dB at 6 km, indicating the growth of ice crystals. As can be observed in Fig. 14 the hydrometeor types are pristine ice crystals from 6 to 8 km and dry snow for altitudes of 4.3–6 km.
Vertical profile of Zh, Zdr, temperature, and Vd at 72-km distance from radar, with the background colors depicting the hydrometeor classes. These profiles correspond to the CSU-CHIVO radar RHI scan at 240° azimuth angle at 0337 UTC 30 Nov 2018. The environmental temperature profile is obtained from the ARM radiosonde station in Cordoba at 0337 UTC 30 Nov 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
Thus, the aggregation growth process of ice particles in the anvil cloud is investigated using sZh, sZdr, and sρhv calculated at altitudes of 5.7–7 km (at a distance of 72 km from the radar location) and is shown in Fig. 15. For the altitude range of interest, the sZh values are in the range from −15 to 5 dB, whereas the sZdr spectrum values are in the range of 0–2.5 dB, and the sρhv spectrum values are between 0.9 and 0.99. At 7-km altitude, sZh less than 0 dB and sZdr greater than 2 dB are observed, indicating the presence of pristine ice crystals characterized by small size and preferentially horizontal orientation. It can also be observed that as the altitude reduces to 6.5 km, the sZh increases and becomes skewed with corresponding low sZdr values indicating particles of bigger size in the radar resolution volume. With the further reduction in altitude to 5.7 km, sZh shows two clear dominant modes with the mode having a higher sZh peak having sZdr values of 0–1 dB. This sZh peak coincides with the higher radial velocity from −8 to −12 m s−1, indicating the presence of bigger aggregates (Cetrone and Houze 2011). Thus, as the altitude decreases from 7 to 5.7 km, the peak sZh increases from around −5.8 to 1 dB with a corresponding increase in the radial velocity. This increase in spectral power indicates the growth of ice crystals to a larger size by depositional growth and aggregation to form dry snow. Similar findings have been reported in Pfitzenmaier et al. (2018) and Cetrone and Houze (2011), where dry snow is formed from the aggregation of crystals present at 6–8-km altitudes. The presence of aggregates is further confirmed because the onset of aggregation leads to the formation of low-density particles with an axial ratio close to one resulting in a significant reduction in the sZdr values. This is in contrast to the growth of particles by vapor diffusion that results in particles retaining their pristine crystal shape and having sZdr values higher than 1.5 dB (Pfitzenmaier et al. 2018). Therefore, signatures of the ice growth process, which is aggregation in this case, can be inferred from the sZh and sZdr values. Because of aggregation, ice crystals and bigger aggregates may be present in the same resolution volume as verified by the simulation results presented later in this section.
As in Fig. 13, but for (bottom) 5.7-, (middle) 6.5-, and (top) 7-km altitudes at a distance of 72 km from the radar at 0337 UTC 30 Nov 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
It has also been observed here that the spectral position of sZh peak and sZdr shifts with altitude. The shift in the spectral position of sZh peak with altitude happens at low elevation angles of observations due to the presence of vertical wind shear. This is because, at low elevation angles (<10°), the background horizontal wind affects the observed radial velocities. Note that the sZh peak shifts by approximately −3 m s−1 within a 500-m (from 7 to 6.5 km) decrease in altitude indicating the presence of vertical wind shear here. Thereafter at 5.7-km altitude, a bimodal spectrum is observed with a high sZh peak (of 1 dB) associated with low sZdr values (of 0.5 dB) at high Doppler velocities (−9 m s−1) and a low sZh peak (−5 dB) coinciding with high sZdr values (1.8 dB) at slow radial velocities (−7 m s−1). Small and horizontally oriented ice crystals are thereby separated (in velocity) from bigger and spherical snowflakes present in the same resolution volume. Therefore, a distinction of ice crystals from snowflakes is possible using spectral polarimetry in the anvil clouds at low elevation angles of observations because of the presence of vertical wind shear (as low as 0.006 s−1).
Next, the scenario corresponding to crystals and aggregates being present in the same radar resolution volume is assumed, and the spectral variables are simulated at a low elevation angle of 4.6°. Therefore, sZh and sZdr have been simulated for a mixture of plates and aggregates present in a radar resolution volume (Spek et al. 2008) and have been compared with the observations discussed above. For this simulation, the backscattering cross section of plates and aggregates have been modeled based on Russchenberg (1994). The normalized intercept parameter Nw for plates is assumed to be constant at 4000 mm−1 m−3 while D0 for plates and aggregates are assumed to be constant at 0.35 and 1.5 mm, respectively. In addition to that, the authors also want to understand the effect (on spectral polarimetric variables) of increasing the number density of aggregates in a resolution volume in the presence of crystals. Therefore, the Nw for aggregates is varied over a range from 2000 to 8000 mm−1 m−3, and the impact of varying the Nw for aggregates on sZh and sZdr is shown in Fig. 16. The range of sZh (from −15 to −5 dB) and sZdr (0–3 dB) values of the simulation output is similar to that of the observations in the anvil cloud discussed above. The simulation results show that with a continuous increase in the concentration of aggregates in the radar resolution volume, the contribution of aggregates increases, and there is a substantial increase in sZh and a decrease in sZdr. Also, an increase in the concentration of aggregates results in a skewed sZh spectrum showing a clear separation in the radial velocities of aggregates and plates. The effect of ambient wind velocity is to shift the radial velocities left/right at low elevation angles but has not been considered in the simulations. Therefore, the simulation results are in good agreement with the observations in the anvil cloud discussed earlier.
Simulation of (left) sZh and (right) sZdr spectra for varying Nw of aggregates in case of a mixture of plates and aggregates.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
2) Spectral polarimetric signatures in graupel formation
Graupel is usually formed by the riming of ice crystals or dendrites, when supercooled liquid water droplets collide and freeze onto bigger ice crystals, resulting in large, dense, and almost spherical particles (Li et al. 2021; Giangrande et al. 2016; Vogel and Fabry 2018). However, supercooled liquid water cannot be identified at the C-band wavelength, making it difficult to differentiate between riming and aggregation processes in bulk radar observations (Spek et al. 2008). In this study, spectral polarimetry provides a framework for studying the spectral signature of graupel formation. Consequently, regions of graupel from the 30 November 2018 precipitation event (discussed in section 2) have been used in this study. Graupel formation regions have been identified at a distance of 10 and 20 km from the radar based on hydrometeor classification. The corresponding vertical profiles of Zh, Zdr, and Doppler velocity are shown in Figs. 17 and 18. For graupel at 10 km from the radar, Zh increases from 7 to 28 dBZ as the altitude decreases from 7 to 4 km, indicating the growth of ice crystals by riming. The corresponding values of Zdr are around 0–1 dB at these altitudes, indicating particles with an axial ratio close to 1 or horizontal orientation. The second chosen region at 20 km from the radar shows an increase in Zh values from 29 to 37 dBZ as the altitude decreases from 4 to 2.5 km and has Zdr values around 0–1 dB, and here graupel is formed by riming of dendrites. For both the graupel regions, negative Zdr values can also be observed between 3- and 4-km altitude, which possibly indicates the presence of conical graupel (Oue et al. 2015). The region of graupel formation requires supercooled liquid water droplets that in this scenario have been transported to the higher altitudes because of a weak updraft (Majewski and French 2020; Korolev and Field 2008). The presence of weak vertical updraft can be inferred from the radial velocity profiles shown in Figs. 17 and 18 where the Doppler velocity changes from negative to positive radial direction at altitudes of 3–4 km.
As in Fig. 14, but at 10-km distance from radar.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
As in Fig. 14, but at a location of 20 km from the radar.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
For the spectral signature analysis of graupel formation region (based on riming), observations at elevation angles of 22°, 29°, and 31° are used to determine radar spectral polarimetric variables at an altitude range of 4–6 km (at the distance of 10 km). The calculated sZh and sZdr are shown in decreasing order of altitudes (6–4 km) in Fig. 19 where the sZh spectra values are between 0 and 18 dB, while sZdr values are between 0 and 1.3 dB at these altitudes. The spectral power can be observed to increase with a decrease in altitude, indicating ice particles growing to larger sizes and the sZh values correspond to that of graupel and the sZdr spectral patterns are similar to that of riming as per Pfitzenmaier et al. (2018). The sZh changes from being Gaussian at 6-km altitude to being skewed at 5.5 km and then bimodal spectrum with peaks centered at two different velocities (at 4-km altitude). The sZdr values corresponding to both peaks (at 4 km) are around 1 dB, indicating the presence of compact and isotropic particles for slow and fast radial velocities. These slightly enhanced sZdr values (1–1.2 dB) correspond to needle or columnar crystals in a temperature range from −8° to −3°C formed by secondary ice production through the Hallet–Mossop ice multiplication mechanism (Oue et al. 2018). This kind of secondary ice multiplication has been reported in previous studies (Sullivan et al. 2018; Hallett and Mossop 1974; Kalesse et al. 2016) and occurs due to different mechanisms such as rime splintering, in which ice hydrometeors collide with and freeze supercooled droplets to form rime, which then splinters off as the hydrometeor continues to fall. These ice splinters formed in the riming regions are small in size and horizontally oriented, and a large concentration of these secondary ice particles in the riming zones can result in positive values of sZdr. However, the radar resolution volume with only graupel results in sZdr values near 0 dB (rimed particles are near spherical in shape). A particle size distribution with a mixture of graupel and secondary ice particles will result in positive sZdr values (Oue et al. 2015, 2018) as observed here.
(left) Spectral reflectivity and (right) spectral differential reflectivity estimated for (bottom) 4-, (middle) 5.5-, and (top) 6-km altitudes at a distance of 10 km from the radar. The observations correspond to the CSU-CHIVO RHI scan at 240° azimuth angle at 0337 UTC 30 Nov 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
Next, an analysis of graupel formation from dendrites has also been carried out using spectral polarimetric variables. The spectral analysis has been performed using observations at elevation angles of 11°, 14°, and 15° and the spectral variables are calculated at altitudes 4, 5, and 5.5 km at a distance of 20 km from the radar. Dendrites are low-density particles and are bigger than plates but smaller than graupel in size (Spek et al. 2008) and they exist with their major axis horizontally oriented resulting in high Zdr values. The results of spectral variables in Fig. 20 shows that peak sZh is between 7 and 20 dB, while sZdr is between 0 and 1.5 dB at these altitudes where dendrites undergo riming to form graupel. An increase in spectral power can be observed with a decrease in altitude, indicating that the ice particles grow to a larger size. The values of sZdr are observed to reduce from a range from 1–1.5 to 0–1 dB as the altitude decreases from 5.5 to 4 km, showing that the axial ratio of the particles becomes close to one on riming. Here, the spectral power is higher than that observed at a distance of 10 km from radar, where crystals undergo riming to form graupel indicating that bigger particles are being formed.
As in Fig. 19, but for (top) 5.5-, (middle) 5-, and (bottom) 4-km altitudes at a distance of 20 km from the radar.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
Figure 20 shows that sZh changes from the spectrum with two subpeaks at 5.5 km to bimodal skewed left at 5 km and skewed right at 4-km altitude. It can also be observed that the spectral sZh at 5-km altitude has a main peak at higher radial velocity than the subpeak, indicating the presence of two different hydrometeor populations (dendrites and graupel) in the radar resolution volume. Additionally, it has been observed that the sZh main peak is at approximately 12 dB with corresponding sZdr values higher than 1 dB, indicating slightly horizontally oriented hydrometeors. On the other hand, the sZh lower peak is at around 7 dB with corresponding sZdr values at approximately 0.5 dB indicating a nearly spherical orientation of particles. Lower sZh values may be attributed to the low number concentration of graupel within the radar resolution volume while higher sZh values may be attributed to the high number concentration of dendrites. This is because the size range of dendrites is 0.03–0.4 cm and that of graupel is 0.2–0.8 cm (Spek et al. 2008). This shows that the size of dendrites can be as high as 0.4 cm and the size of graupel can be as low as 0.2 cm. Thus, a higher concentration of large dendrites results in the main peak of the sZh spectrum while the relatively low concentration of small graupel (than that of dendrites) at these altitudes constitutes the subpeak of the sZh spectrum. This is therefore the altitude corresponding to the onset of the riming of dendrites. Furthermore, at 4-km altitude, the sZh spectrum broadens and becomes skewed right indicating continued ice particle growth by riming of dendrites resulting in graupel formation. At the same altitude, a negative slope of sZdr can be observed, indicating the possible size sorting of hydrometeors because of wind shear. This is supported by the vertical profile of radial velocity at a distance of 20 km from radar shown in Fig. 18. The sZdr values at approximately 1 dB as observed at these altitudes indicate the presence of horizontally oriented particles, that is, rime splinters resulting from secondary ice production in the riming zones as discussed in the previous case. Here, note that the polarimetric Zdr values for dendrites at elevation angle measurements are comparable to spectral Zdr. The Zdr values for dendrites decrease with an increase in elevation angle, and therefore the range of Zdr values for dendrites is from 1 to 2 dB at a higher elevation angle at C band as discussed in Thompson et al. (2014). Similarly, the sZdr for dendrites is estimated with a maximum value of 1.5 dB at an elevation angle of around 14° in the current study and is in the expected range. From the above discussions, it is clear that the hydrometeor types and their particle size distributions significantly affect the Doppler power spectrum and sZdr spectrum (Spek et al. 2008). The skewed and bimodal skewed sZh provide inferences regarding the mixture of different hydrometeor types in the radar resolution volume.
c. Study of the temporal evolution of a storm
The temporal evolution of a stratiform precipitation event has been studied using spectral polarimetric variables. These spectral variables have been determined for the precipitation event on 30 November 2018 that evolved from a convective storm accompanied by heavy rainfall. As part of this analysis, the temporal changes in radar reflectivity for RHI scans at 200° azimuth have been presented in Fig. 21. The reflectivity images at 0325 and 0337 UTC correspond to the development of the stratiform precipitation event from a convective storm, whereas those at 0347 to 0417 UTC correspond to the intensification of the precipitation system. The precipitation-decaying convective core (characterized by reflectivity values of up to 50 dBZ) can be observed at 0325 UTC at a distance of 20 km from the radar and rising up to an altitude of 2 km. The corresponding hydrometeor classification for 0325 UTC (Fig. 22) confirms the presence of rain and graupel extending from the ground up to 6-km altitude. Later on at 0347 and 0357 UTC, the bright band is seen to appear at an altitude of 3.5 km while the hydrometeor classification shows the presence of graupel, wet snow, and crystals at these altitudes. These hydrometeors type evolution has been studied using spectral analyses at a distance of 40 km from radar and at altitudes of 5 and 6.5 km to infer the temporal evolution signatures.
RHI plots of reflectivity showing the temporal evolution of stratiform precipitation from a decaying convective storm and its weakening stage. The observations correspond to CSU-CHIVO RHI scan at 200° azimuth angle on 30 Nov 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
Hydrometeor classification showing the temporal evolution of hydrometeor types in stratiform precipitation, where the observations correspond to CSU-CHIVO RHI scan at 200° azimuth angle on 30 Nov 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
The spectral polarimetric variables at an altitude of 6.5 km and at a distance of 40 km from the radar are computed (Fig. 23). This region corresponds to graupel and/or ice crystals as observed from the hydrometeor classification. For this region at 0325–0347 UTC, sZh spectra are either broad or bimodal with a maximum value of 0 dB while sZdr values are in the range of 0.5–1.5 dB, demonstrating the properties of a mix of ice crystals and graupel. However, for the time span of 0357–0417 UTC, sZh distribution transforms to a narrow Gaussian shape, and the peak sZdr increases above 1.5 dB and close to 2 dB, indicating the presence of an elevated number of crystals. This shows that each hydrometeor class has a specific spectral signature.
(left) Spectral reflectivity and (right) spectral differential reflectivity estimated at an altitude of 6.5 km at a distance of 40 km from the radar at (from top to bottom) 0325, 0337, 0347, 0357, 0407, and 0417 UTC. The observations correspond to CSU-CHIVO RHI scan at 200° azimuth angle on 30 Nov 2018.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
Next, spectral variables at 5-km altitude and 40-km distance from the radar are computed for studying the temporal evolution of ice particles during the precipitation event (Fig. 24). At 0325 UTC, the sZh spectrum at this altitude is found to be Gaussian and has a peak value of around 17 dB and sZdr values of around 1 dB, indicating the presence of graupel (as also confirmed by hydrometeor classification). The presence of graupel is confirmed because it has a higher sZh and lower sZdr than crystals. As time advances to 0337 UTC, the sZh spectrum becomes skewed, and the peak value is reduced to 10 dB, while the corresponding sZdr values are around 1.8 dB. Meanwhile, sZdr is 0 dB for the sZh range of 0–5 dB, indicating the presence of a mixture of dry snow, crystals, and graupel of varying number concentrations. Later on at 0347 UTC, the sZh peak value increases to 15 dB with corresponding sZdr of around 1.5 dB, showing the presence of a mixture of graupel and crystals in the radar resolution volume. Furthermore, sZh at 0357 UTC is found to have a peak value of 7 dB and sZdr values of around 1 dB. Based on the values of sZh and sZdr, the particles are smaller than graupel but still horizontally oriented so the hydrometeors are a mixture of dry snow and crystals (as determined based on hydrometeor classification). With further passing of time (at 0407 and 0417 UTC), sZh values have reduced (by 3 dB), indicating the decrease in the size of particles, while sZdr values have increased, showing the presence of horizontally oriented particles like crystals in the radar resolution volume. At 0417 UTC, it can be observed that the Doppler velocities are positive (0–3 m s−1), corresponding to the region of crystals in the spectrum, showing the crystals being carried by an outbound horizontal wind (since the observations are at a low elevation angle of 7°).
As in Fig. 23, but at an altitude of 5 km.
Citation: Journal of Atmospheric and Oceanic Technology 41, 3; 10.1175/JTECH-D-22-0113.1
4. Summary and discussion
In this study, spectral polarimetric variables are calculated using the observations from the CSU-CHIVO radar deployed during the RELAMPAGO field campaign. Subsequently, radar polarimetric variables and their spectral distribution, along with hydrometeor classification, are used to perform an in-depth microphysical properties analysis of different precipitation events during the RELAMPAGO field campaign. To enhance the analysis, the study makes use of temperature and humidity profiles obtained from radiosonde measurements during the experiment, as well as reanalysis data. Here, the analysis can be broadly categorized into two parts. In the first part, the dynamical properties of hydrometeors in a turbulent environment are studied using RHI scan observations at low elevation angles (less than 10°) while in the second part, the microphysical processes related to hydrometeor growth are studied at elevation angles in the range of 0°–25°. The details of the microphysical and dynamical characteristics along with spectral signatures are summarized in Table 2.
Summary of spectral polarimetric characterization of precipitation systems.
The initial storm dynamics study is on the spectral polarimetric analysis of a region of ice-phase hydrometeors (in stratiform precipitation) that is impacted by vertical wind shear. Similarly, the impact of vertical wind shear in a region of large drops near the core of a convective storm is also studied. This study is important because vertical wind shear causes variation in the trajectory of particles of different sizes resulting in the size sorting of hydrometeors in a resolution volume. For the stratiform precipitation affected by vertical wind shear, the sZh values are in the range of 10–20 dB, and sZdr values are in the range from −2 to 2 dB. For the convective storm, sZh values are in the range of 20–30 dB, and sZdr values are in the range of 2–8 dB. Spectral variables at low elevation angles (less than 5°) of observation for sheared and turbulent precipitation are characterized by non-Gaussian flat-topped sZh spectrum along with sZdr spectrum that has a clear linear fit slope. The sZdr linear fit slope for ice hydrometeors in stratiform precipitation is found to be negative, while it is positive for raindrops in the convective storm. The characteristic broadening of flat-topped sZh is due to hydrometeors of different sizes moving with different radial velocities in the same radar resolution volume. Therefore, patterns and values of sZh and sZdr at low elevation angles in the region of vertical wind shear and turbulence could be used to infer the nature of hydrometeor size distributions in the radar resolution volume.
The next part of the storm dynamics study involves the characterization of ice crystals (found) near the melting layer of the stratiform rain event with low-reflectivity region (and significant inversion of Zh profile) at the edge of a low-level inflow jet. This is significant because ice crystals do not form at temperatures near 0°C. Spectral variables have been estimated at elevation angles less than 10°, and it can be observed that the sZh and sZdr values range from −10 to 0 dB and from −2 to 2 dB, respectively, at this altitude. The spectral variable values correspond to that of pristine ice crystals. The sZh and sZdr show trends similar to that of reflectivity and differential reflectivity with altitude; that is, sZh and Zh show a decrease and then an increase in values. Additionally, this region has enhanced Zdr and reduced ρhv and is identified to have pristine ice crystals descending from higher altitudes and ice splinters advecting from surrounding riming zones. This low-reflectivity region is a result of the presence of ice crystals in a column of high inbound velocity associated with the sinking of air in a dry environment. This dry environment has resulted from the cooling by melting (below the melting layer) and sublimation (above the melting layer) and is reinforced by vertical wind shear and turbulence.
The subsequent investigation involves analyzing the spectral polarimetric signatures associated with microphysical growth processes of hydrometeors, that is, the aggregation growth process of crystals in anvil clouds and graupel formation (from riming) from crystals and dendrites. The first growth process studied here is the aggregation of crystals in the anvil cloud region at altitudes 6–7 km. Radar observations at low elevation angles (less than 6°) have been used for the study. In this region, low values of sZh ranging from −10 to 0 dB for Doppler velocities from −12 to 0 m s−1 were observed, accompanied by high sZdr values around 2 dB. These observations indicate the presence of pristine ice crystals at the cloud top as confirmed by hydrometeor classification and a significant horizontal component of background wind velocity at that altitude. Furthermore, a gradual increase in spectral power and decrease in sZdr with decreasing altitude suggest the growth of ice crystals via aggregation (Pfitzenmaier et al. 2018). During this growth process, some noteworthy features observed are the skewed and bimodal spectra of sZh as altitude reduces to the cloud bottom. This bimodal spectrum exhibits two clear peaks: one with higher sZh values at fast radial velocities, associated with low sZdr values, and another peak with lower sZh values at slow radial velocities, linked to high sZdr values. This happens due to the coexistence of aggregates with pristine ice crystals that are separated by radial velocity components. Therefore, a distinction of ice crystals from dry snow in the anvil clouds is possible using spectral polarimetry. In addition, the presence of aggregates leads to collisions, generating electrical charge that can alter the orientation of the ice particles, resulting in Zdr values below 0 dB. Thus, the combination of near-spherical aggregates and horizontally oriented pristine ice crystals, along with variable orientations induced by electrical charge within the anvil cloud, contributes to reduced sρhv values.
The second part of the microphysical growth process study examines the characteristics of spectral polarimetric variables (at elevation angles greater than 20°) for graupel formation by the riming of crystals. The corresponding sZh peak values are higher than 10 dB and exhibit an ascending trend with decreasing altitude from 6 to 4 km showing that the crystals are growing in size. The sZdr shows a decreasing trend with the reduction in altitude with values at approximately 0.5–1 dB. Normally, sZdr for rimed ice particles like graupel would be near zero, but the observed sZdr indicates the presence of secondary ice particles in the form of needle or columnar shapes. These ice particles are formed through the Hallet–Mossop ice multiplication processes within the riming region. Consequently, the spectral analysis revealed that graupel along with needles exhibited two sZh peaks with values ranging from 10 to 20 dB, while sZdr values range between 0.5 and 1.5 dB. This bimodal spectrum has peaks centered at two different velocities. In a related context, this study examines the spectral polarimetric variables (at elevation angles greater than 10°) for graupel forming from dendrites. Here graupel is formed through the riming process and sZh peak values at approximately 10 dB are observed and sZh increases with a reduction in altitude demonstrating the growth process. Conversely, the values of sZdr are observed to reduce from a range from 1–1.5 dB to 0–1 dB as the altitude decreases from 5.5 to 4 km, suggesting that the axial ratio of the particles becomes close to one on riming. At 5-km altitude, particles with sZh values of approximately 12 dB and sZdr values above 1 dB and having faster radial velocities (of 12 m s−1) indicate horizontally oriented hydrometeors. These spectral values suggest the presence of larger as well as horizontally oriented particles. In contrast, particles with sZh values of nearly 7 dB and sZdr values around 0.5 dB and having slower radial velocities indicate a nearly spherical orientation. These spectral values suggest the presence of smaller and spherically oriented particles. These findings show that there is a higher concentration of large dendrites resulting in the main peak and lower concentration (than that of dendrites) of small graupel constituting the subpeak of the sZh spectrum. This is therefore the altitude corresponding to the onset of riming of dendrites.
In the final phase of the study, the focus of the analysis is on the temporal variation of the spectral polarimetric variables at two altitudes of 5 and 6.5 km for evolving storm dynamics and particle types associated with a stratiform precipitation event. At 6.5 km, for the time frame from 0325 to 0347 UTC, the particles are characterized by broad or bimodal sZh distribution with maximum values below 0 dB while sZdr values are around 0.5–1 dB. This is significant because low sZh values correspond to ice crystals whereas low sZdr values correspond to graupel. Both sZh and sZdr being low indicates a mixture of crystals and graupel. However, during the time window from 0357 to 0417 UTC, the sZh distribution becomes narrower and Gaussian shaped, while the sZdr value increases above 1.5 dB, approaching 2 dB. This sZdr increase indicated the presence of crystals, which was confirmed through hydrometeor classification. The spectral reflectivity at 5 km also shows Gaussian or broad spectrum, but the sZh peak values are higher than 5 dB throughout most of the time window under consideration. The sZh at 17 dB and sZdr values at approximately 1 dB correspond to dry snow or graupel, as verified by the hydroclassification. Furthermore, sZh at 5–10 dB and sZdr at 0–2 dB correspond to a mixture of dry snow, graupel, and crystals. The analysis reveals that the types and sizes of particles vary over time, corresponding to changes in storm type. These temporal variations exhibit distinct spectral signatures, which can be employed to analyze the evolving microphysical processes at specific altitudes and distances from the radar.
In this study, RHI scan measurements at multiple elevation angles have been utilized to estimate the profiles of spectral polarimetric variables at specific distances from radar. Profiles at low elevation angles (1°–10°) have demonstrated their suitability for studying storm dynamics, enabling the estimation of polarimetric variables across a range of radial Doppler velocities. Consequently, essential information such as background wind and its impact on particle motion, such as size sorting, can be easily determined using spectral variables. The spectral sZdr signatures exhibit distinctive characteristics like a specific linear fit slope, enabling straightforward extraction of such signatures. On the other hand, measurements in the range of elevation angles 10°–30° have proved useful for studying the microphysical growth processes. Particularly, the spectral variables profiles facilitate the identification of altitude ranges associated with specific ice growth processes, and secondary ice formation. Furthermore, spectral polarimetry helps in the separation of different particle types within the radar resolution volume.
This study is different from analyses using spectral polarimetric variables from zenith-pointing radars that use size-dependent terminal fall velocity to distinguish between hydrometeor types of different sizes in the radar resolution volume. On the contrary, this analysis uses measurements at elevation angles 0°–10° extensively. This is because the spectral polarimetric variables for rain and ice particles are effectively determined since the hydrometeor primary (horizontal) axis is parallel to the radar beam. Along with that, the retrieved radial velocities have no significant contribution from the terminal fall velocity and are primarily impacted by the horizontal wind component (Dufournet and Russchenberg 2011). The variables have been used to analyze the size sorting of particles, growth of crystals in anvil clouds, and temporal evolution of stratiform storm events. Conversely, the polarimetric variables are known to change slightly between 10° and 20° and significantly for elevation angles above 20°. At elevation angles higher than 20°, the radar beam is not oriented along the primary (horizontal) particle axis, resulting in degraded measurements of Zdr (Ryzhkov et al. 2005a,b), that is, lower values of Zdr for oblate ice crystals (Evans and Vivekanandan 1990; Dolan and Rutledge 2009). This is because elevation angle dependence shows that Zdr for dendrites could change from 4 dB at 1° to 2.7 dB at 30° elevation angle. Furthermore, Zdr for plates could change from 5.2 dB at 1° to 3.5° dB at a 30° elevation angle. Thus, the measured Zdr values for plates is 0.8 dB greater than that of dendrites (at 30°) resulting in the reduction of Zdr difference between plates and dendrites. Therefore, the different ranges of measured values of Zdr at elevation angles lower than 10° are more useful for hydrometeor classification than those at 30°. Considering the above details, spectral polarimetric variables estimated at elevation angles 10°–20° and 20°–30° have been used for studying the microphysical processes in graupel formation using relative variation in spectral Zdr. This is because, even though the sZdr difference between hydrometeor types might be low at the elevation angles of 10°–20° and 20°–30°, the relative variation of Zdr (between particles of different types) along with their radial velocities has been used for analyzing and deriving various useful inferences about the microphysical properties and growth processes.
Thus, to characterize precipitation, it is necessary to study the ongoing microphysical and kinematic processes in the vertical structure of a storm. Spectral polarimetry aids in this process by relating polarimetric signatures with storm dynamics. Furthermore, spectral polarimetric analyses can be used to determine parameters such as ice crystal number density to help reduce critical assumptions made in cloud modeling. It is also useful for determining behaviors of hydrometeors in different storms and precipitation systems, which ultimately helps in accurate estimation of precipitation type and rate. Consequently, the utilization of spectral polarimetric measurements can contribute to the enhanced characterization of ice particle types and their number density. This, in turn, enables more accurate simulation of the relevant microphysical processes associated with distinct particle types, leading to significant improvements in the modeling of convective and stratiform precipitation events (Dufournet and Russchenberg 2011). Overall, the information obtained from spectral polarimetry can be used to improve the representation of various precipitation processes in weather models, which in turn can help to improve the accuracy of precipitation forecasts.
Acknowledgments.
RELAMPAGO field experiment data have been downloaded from the Earth Observing Laboratory (EOL) data web page. The authors thank Ivan Arias Hernandez for his helpful discussions related to the CSU-CHIVO radar data processing.
Data availability statement.
Data are available from the authors and will be shared upon request.
REFERENCES
Anabor, V., D. J. Stensrud, and O. L. L. de Moraes, 2008: Serial upstream-propagating mesoscale convective system events over southeastern South America. Mon. Wea. Rev., 136, 3087–3105, https://doi.org/10.1175/2007MWR2334.1.
Andrić, J., M. R. Kumjian, D. S. Zrnić, J. M. Straka, and V. M. Melnikov, 2013: Polarimetric signatures above the melting layer in winter storms: An observational and modeling study. J. Appl. Meteor. Climatol., 52, 682–700, https://doi.org/10.1175/JAMC-D-12-028.1.
Arias, I., and V. Chandrasekar, 2021: Cross validation of the network of ground-based radar with GPM during the Remote Sensing of Electrification, Lightning, and Mesoscale/Microscale Processes with Adaptive Ground Observations (RELAMPAGO) field campaign. J. Meteor. Soc. Japan, 99, 1423–1438, https://doi.org/10.2151/jmsj.2021-069.
Bechini, R., and V. Chandrasekar, 2015: A semisupervised robust hydrometeor classification method for dual-polarization radar applications. J. Atmos. Oceanic Technol., 32, 22–47, https://doi.org/10.1175/JTECH-D-14-00097.1.
Bechini, R., L. Baldini, and V. Chandrasekar, 2013: Polarimetric radar observations in the ice region of precipitating clouds at C-band and X-band radar frequencies. J. Appl. Meteor. Climatol., 52, 1147–1169, https://doi.org/10.1175/JAMC-D-12-055.1.
Biswas, S. K., and V. Chandrasekar, 2018: Cross-validation of observations between the GPM Dual-Frequency Precipitation Radar and ground based dual-polarization radars. Remote Sens., 10, 1773, https://doi.org/10.3390/rs10111773.
Biswas, S. K., R. Cifelli, and V. Chandrasekar, 2020: Improving quantitative precipitation estimation by X-band dual-polarization radars in complex terrain over the bay area in California, USA. 2020 IEEE Int. Geoscience and Remote Sensing Symp., Waikoloa, HI, IEEE, 5411–5414, https://doi.org/10.1109/IGARSS39084.2020.9323239.
Biswas, S. K., V. Chandrasekar, S. Sahoo, and A. K. Lakshmi, 2022: Study of a convective event during the RELAMPAGO field experiment using spectral polarimetry. 2022 IEEE Int. Geoscience and Remote Sensing Symp., Kuala Lumpur, Malaysia, IEEE, 6534–6537, https://doi.org/10.1109/IGARSS46834.2022.9884392.
Biswas, S. K., R. Cifelli, and V. Chandrasekar, 2024: Quantitative precipitation estimation using X-band radar for orographic rainfall in the San Francisco Bay Area. IEEE Trans. Geosci. Remote Sens., https://doi.org/10.1109/TGRS.2022.3207829, in press.
Borque, P., L. Vidal, M. Rugna, T. J. Lang, M. G. Nicora, and S. W. Nesbitt, 2020: Distinctive signals in 1-min observations of overshooting tops and lightning activity in a severe supercell thunderstorm. J. Geophys. Res. Atmos., 125, e2020JD032856, https://doi.org/10.1029/2020JD032856.
Bringi, V. N., and V. Chandrasekar, 2001: Polarimetric Doppler Weather Radar: Principles and Applications. Cambridge University Press, 636 pp.
Cetrone, J., and R. A. Houze Jr., 2011: Leading and trailing anvil clouds of West African squall lines. J. Atmos. Sci., 68, 1114–1123, https://doi.org/10.1175/2011JAS3580.1.
Chen, H., V. Chandrasekar, and R. Bechini, 2017: An improved dual-polarization radar rainfall algorithm (DROPS2.0): Application in NASA IFloodS field campaign. J. Hydrometeor., 18, 917–937, https://doi.org/10.1175/JHM-D-16-0124.1.
Cifelli, R., and V. Chandrasekar, 2010: Dual-polarization radar rainfall estimation. Rainfall: State of the Science, Geophys. Monogr., Vol. 191, Amer. Geophys. Union, 105–125, https://doi.org/10.1029/2010GM000930.
Dawson, D. T., II, E. R. Mansell, and M. R. Kumjian, 2015: Does wind shear cause hydrometeor size sorting? J. Atmos. Sci., 72, 340–348, https://doi.org/10.1175/JAS-D-14-0084.1.
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.
Doviak, R. J., and Coauthors, 2006: Doppler Radar and Weather Observations. Courier Corporation, 562 pp.
Dufournet, Y., and H. Russchenberg, 2011: Towards the improvement of cloud microphysical retrievals using simultaneous Doppler and polarimetric radar measurements. Atmos. Meas. Tech., 4, 2163–2178, https://doi.org/10.5194/amt-4-2163-2011.
Dye, J. E., and A. Bansemer, 2019: Electrification in mesoscale updrafts of deep stratiform and anvil clouds in Florida. J. Geophys. Res. Atmos., 124, 1021–1049, https://doi.org/10.1029/2018JD029130.
Edde, B., 1993: Radar: Principles, Technology, Applications. Prentice Hall, 717 pp.
Evans, K. F., and J. Vivekanandan, 1990: Multiparameter radar and microwave radiative transfer modeling of nonspherical atmospheric ice particles. IEEE Trans. Geosci. Remote Sens., 28, 423–437, https://doi.org/10.1109/TGRS.1990.572908.
Giangrande, S. E., T. Toto, A. Bansemer, M. R. Kumjian, S. Mishra, and A. V. Ryzhkov, 2016: Insights into riming and aggregation processes as revealed by aircraft, radar, and disdrometer observations for a 27 April 2011 widespread precipitation event. J. Geophys. Res. Atmos., 121, 5846–5863, https://doi.org/10.1002/2015JD024537.
Grim, J. A., G. M. McFarquhar, R. M. Rauber, A. M. Smith, and B. F. Jewett, 2009: Microphysical and thermodynamic structure and evolution of the trailing stratiform regions of mesoscale convective systems during BAMEX. Part II: Column model simulations. Mon. Wea. Rev., 137, 1186–1205, https://doi.org/10.1175/2008MWR2505.1.
Hallett, J., and S. C. Mossop, 1974: Production of secondary ice particles during the riming process. Nature, 249, 26–28, https://doi.org/10.1038/249026a0.
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.
Hogan, R. J., P. R. Field, A. J. Illingworth, R. J. Cotton, and T. W. Choularton, 2002: Properties of embedded convection in warm-frontal mixed-phase cloud from aircraft and polarimetric radar. Quart. J. Roy. Meteor. Soc., 128, 451–476, https://doi.org/10.1256/003590002321042054.
Kalesse, H., W. Szyrmer, S. Kneifel, P. Kollias, and E. Luke, 2016: Fingerprints of a riming event on cloud radar Doppler spectra: Observations and modeling. Atmos. Chem. Phys., 16, 2997–3012, https://doi.org/10.5194/acp-16-2997-2016.
Keat, W. J., and C. D. Westbrook, 2017: Revealing layers of pristine oriented crystals embedded within deep ice clouds using differential reflectivity and the copolar correlation coefficient. J. Geophys. Res. Atmos., 122, 11 737–11 759, https://doi.org/10.1002/2017JD026754.
Kedzuf, N. J., and Coauthors, 2021: Retrieving microphysical properties of concurrent pristine ice and snow using polarimetric radar observations. Atmos. Meas. Tech., 14, 6885–6904, https://doi.org/10.5194/amt-14-6885-2021.
Korolev, A., and P. R. Field, 2008: The effect of dynamics on mixed-phase clouds: Theoretical considerations. J. Atmos. Sci., 65, 66–86, https://doi.org/10.1175/2007JAS2355.1.
Kumjian, M. R., and A. V. Ryzhkov, 2012: The impact of size sorting on the polarimetric radar variables. J. Atmos. Sci., 69, 2042–2060, https://doi.org/10.1175/JAS-D-11-0125.1.
Li, H., A. Korolev, and D. Moisseev, 2021: Supercooled liquid water and secondary ice production in Kelvin–Helmholtz instability as revealed by radar Doppler spectra observations. Atmos. Chem. Phys., 21, 13 593–13 608, https://doi.org/10.5194/acp-21-13593-2021.
Majewski, A., and J. R. French, 2020: Supercooled drizzle development in response to semi-coherent vertical velocity fluctuations within an orographic-layer cloud. Atmos. Chem. Phys., 20, 5035–5054, https://doi.org/10.5194/acp-20-5035-2020.
Marshall, J. S., 1953: Precipitation trajectories and patterns. J. Meteor., 10, 25–29, https://doi.org/10.1175/1520-0469(1953)010%3C0025:PTAP%3E2.0.CO;2.
Moisseev, D. N., and V. Chandrasekar, 2009: Polarimetric spectral filter for adaptive clutter and noise suppression. J. Atmos. Oceanic Technol., 26, 215–228, https://doi.org/10.1175/2008JTECHA1119.1.
Montopoli, M., E. Picciotti, L. Baldini, S. Di Fabio, F. 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.
Mulholland, J. P., S. W. Nesbitt, and R. J. Trapp, 2019: A case study of terrain influences on upscale convective growth of a supercell. Mon. Wea. Rev., 147, 4305–4324, https://doi.org/10.1175/MWR-D-19-0099.1.
Nesbitt, S. W., and Coauthors, 2021: A storm safari in subtropical South America: Proyecto RELAMPAGO. Bull. Amer. Meteor. Soc., 102, E1621–E1644, https://doi.org/10.1175/BAMS-D-20-0029.1.
Oue, M., M. R. Kumjian, Y. Lu, Z. Jiang, E. E. Clothiaux, J. Verlinde, and K. Aydin, 2015: X-band polarimetric and Ka-band Doppler spectral radar observations of a graupel-producing Arctic mixed-phase cloud. J. Appl. Meteor. Climatol., 54, 1335–1351, https://doi.org/10.1175/JAMC-D-14-0315.1.
Oue, M., P. Kollias, A. Ryzhkov, and E. P. Luke, 2018: Toward exploring the synergy between cloud radar polarimetry and Doppler spectral analysis in deep cold precipitating systems in the Arctic. J. Geophys. Res. Atmos., 123, 2797–2815, https://doi.org/10.1002/2017JD027717.
Pfitzenmaier, L., C. M. H. Unal, Y. Dufournet, and H. W. J. Russchenberg, 2018: Observing ice particle growth along fall streaks in mixed-phase clouds using spectral polarimetric radar data. Atmos. Chem. Phys., 18, 7843–7862, https://doi.org/10.5194/acp-18-7843-2018.
Pinsky, M. B., and A. P. Khain, 1994: Initiation of drop velocity oscillations during their fall in a vertically sheared flow with embedded velocity fluctuation. Geophys. Astrophys. Fluid Dyn., 78, 169–192, https://doi.org/10.1080/03091929408226577.
Rasmussen, K. L., and R. A. Houze Jr., 2011: Orogenic convection in subtropical South America as seen by the TRMM satellite. Mon. Wea. Rev., 139, 2399–2420, https://doi.org/10.1175/MWR-D-10-05006.1.
Rasmussen, K. L., and R. A. Houze Jr., 2016: Convective initiation near the Andes in subtropical South America. Mon. Wea. Rev., 144, 2351–2374, https://doi.org/10.1175/MWR-D-15-0058.1.
Russchenberg, H. W. J., 1994: Ground-Based Remote Sensing of Precipitation Using a Multi-Polarized FM-CW Doppler Radar. Delft University Press, 207 pp.
Ryzhkov, A. V., S. E. Giangrande, V. M. Melnikov, and T. J. Schuur, 2005a: Calibration issues of dual-polarization radar measurements. J. Atmos. Oceanic Technol., 22, 1138–1155, https://doi.org/10.1175/JTECH1772.1.
Ryzhkov, A. V., T. J. Schuur, D. W. Burgess, and D. S. Zrnić, 2005b: Polarimetric tornado detection. J. Appl. Meteor. Climatol., 44, 557–570, https://doi.org/10.1175/JAM2235.1.
Salio, P., M. Nicolini, and E. J. Zipser, 2007: Mesoscale convective systems over southeastern South America and their relationship with the South American low-level jet. Mon. Wea. Rev., 135, 1290–1309, https://doi.org/10.1175/MWR3305.1.
Schumacher, R. S., and Coauthors, 2021: Convective-storm environments in subtropical South America from high-frequency soundings during RELAMPAGO-CACTI. Mon. Wea. Rev., 149, 1439–1458, https://doi.org/10.1175/MWR-D-20-0293.1.
Smith, A. M., G. M. McFarquhar, R. M. Rauber, J. A. Grim, M. S. Timlin, B. F. Jewett, and D. P. Jorgensen, 2009: Microphysical and thermodynamic structure and evolution of the trailing stratiform regions of mesoscale convective systems during BAMEX. Part I: Observations. Mon. Wea. Rev., 137, 1165–1185, https://doi.org/10.1175/2008MWR2504.1.
Spek, A. L. J., C. M. H. Unal, D. N. Moisseev, H. W. J. Russchenberg, V. Chandrasekar, and Y. Dufournet, 2008: A new technique to categorize and retrieve the microphysical properties of ice particles above the melting layer using radar dual-polarization spectral analysis. J. Atmos. Oceanic Technol., 25, 482–497, https://doi.org/10.1175/2007JTECHA944.1.
Sullivan, S. C., C. Hoose, A. Kiselev, T. Leisner, and A. Nenes, 2018: Initiation of secondary ice production in clouds. Atmos. Chem. Phys., 18, 1593–1610, https://doi.org/10.5194/acp-18-1593-2018.
Thompson, E. J., S. A. Rutledge, B. Dolan, V. Chandrasekar, and B. L. Cheong, 2014: A dual-polarization radar hydrometeor classification algorithm for winter precipitation. J. Atmos. Oceanic Technol., 31, 1457–1481, https://doi.org/10.1175/JTECH-D-13-00119.1.
Trapp, R. J., and Coauthors, 2020: Multiple-platform and multiple-Doppler radar observations of a supercell thunderstorm in South America during RELAMPAGO. Mon. Wea. Rev., 148, 3225–3241, https://doi.org/10.1175/MWR-D-20-0125.1.
Vogel, J. M., and F. Fabry, 2018: Contrasting polarimetric observations of stratiform riming and nonriming events. J. Appl. Meteor. Climatol., 57, 457–476, https://doi.org/10.1175/JAMC-D-16-0370.1.
Wang, Y., T.-Y. Yu, A. V. Ryzhkov, and M. R. Kumjian, 2019: Application of spectral polarimetry to a hailstorm at low elevation angle. J. Atmos. Oceanic Technol., 36, 567–583, https://doi.org/10.1175/JTECH-D-18-0115.1.
Yu, T.-Y., R. R. Rondinel, and R. D. Palmer, 2009: Investigation of non-Gaussian Doppler spectra observed by weather radar in a tornadic supercell. J. Atmos. Oceanic Technol., 26, 444–461, https://doi.org/10.1175/2008JTECHA1124.1.
Yu, T.-Y., X. Xiao, and Y. Wang, 2012: Statistical quality of spectral polarimetric variables for weather radar. J. Atmos. Oceanic Technol., 29, 1221–1235, https://doi.org/10.1175/JTECH-D-11-00090.1.
