Comparison of Disdrometer and X-Band Mobile Radar Observations in Convective Precipitation

2414 MONTHLY WEATHER REVIEW VOLUME 142
Comparison of Disdrometer and X-Band Mobile Radar Observations in Convective

Precipitation
EVAN A. KALINA AND KATJA FRIEDRICH

Department of Atmospheric and Oceanic Sciences, University of Colorado Boulder, Boulder, Colorado
SCOTT M. ELLIS
National Center for Atmospheric Research, Boulder, Colorado
DONALD W. BURGESS
Cooperative Institute for Mesoscale Meteorological Studies, University of Oklahoma, and National Severe Storms
Laboratory, Norman, Oklahoma
(Manuscript received 28 January 2014, in final form 11 March 2014)
ABSTRACT
Microphysical data from thunderstorms are sparse, yet they are essential to validate microphysical schemes
in numerical models. Mobile, dual-polarization, X-band radars are capable of providing a wealth of data that
include radar reflectivity, drop shape, and hydrometeor type. However, X-band radars suffer from beam
attenuation in heavy rainfall and hail, which can be partially corrected with attenuation correction schemes. In
this research, the authors compare surface disdrometer observations to results from a differential phase-based
attenuation correction scheme. This scheme is applied to data recorded by the National Oceanic and At-
mospheric Administration (NOAA) X-band dual-polarized (NOXP) mobile radar, which was deployed
during the second Verification of the Origins of Rotation in Tornadoes Experiment (VORTEX2). Results are
presented from five supercell thunderstorms and one squall line (183 min of data). The median disagreement
(radar–disdrometer) in attenuation-corrected reflectivity Z and differential reflectivity ZDR is just 1.0 and
0.19 dB, respectively. However, two data subsets reveal much larger discrepancies in Z (ZDR): 5.8 (1.6) dB in
a hailstorm and 213 (20.61) dB when the radar signal quality index (SQI) is less than 0.8. The discrepancies
are much smaller when disdrometer and S-band Weather Surveillance Radar-1988 Doppler (WSR-88D) Z
are compared, with differences of 21.5 dB (hailstorm) and 20.66 dB (NOXP SQI , 0.8). A comparison of the
hydrometeor type retrieved from disdrometer and NOXP radar data is also presented, in which the same class
is assigned 63% of the time.
1. Introduction low-level cold pool, which changes the near-surface

buoyancy tendency and, as suggested by several recent
The lack of surface microphysical and in situ data is
studies, the tornadogenesis potential (Markowski et al.
a critical obstacle in our attempts to understand and
2002; Shabbott and Markowski 2006; Grzych et al.
model severe thunderstorms accurately. Microphysical
2007). To collect the surface microphysical data re-
processes (e.g., accretion, collision, coalescence, drop
quired to understand and quantify these interactions,
breakup, melting, and evaporation) affect storm be-
Particle Size and Velocity (PARSIVEL) optical dis-
havior and evolution by serving as a crucial link between
drometers were deployed during the second Verification
the storm dynamics and thermodynamics. For example,
of the Origins of Rotation in Tornadoes Experiment
the melting of hail influences the strength and size of the
(VORTEX2) to obtain particle diameter and fall speed
distributions in severe thunderstorms. For the first time,
these deployments were coordinated with X-band, mo-
Corresponding author address: Evan A. Kalina, Department of
Atmospheric and Oceanic Sciences, University of Colorado bile, polarimetric Doppler radars in severe thunder-
Boulder, UCB 311, Boulder, CO 80309. storms, which provided a three-dimensional dataset of
E-mail: [email protected] radar reflectivity Z, differential reflectivity ZDR, and
DOI: 10.1175/MWR-D-14-00039.1
Ó 2014 American Meteorological Society

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC
JULY 2014 KALINA ET AL. 2415
differential phase CDP that is needed to characterize Bringi et al. 2001; Anagnostou et al. 2006; Steiner et al.
microphysical processes throughout thunderstorms. 2009). The quantity KDP is the range derivative of FDP,
The VORTEX2 measurements of supercell thunder- which must be calculated from the radar-measured total
storm microphysics with disdrometers and mobile differential phase CDP. The CDP is the sum of FDP and
X-band radars are unprecedented, since both sets of the backscatter differential phase d. The terms CDP,
instruments were deployed close to the storm and yiel- FDP, d, and KDP are related via Eq. (1):
ded high-resolution information near and at the sur- ðr
face. This dataset provides researchers with a unique 2 KDP (r0 ) dr0 5 CDP 2 d 5 FDP , (1)
opportunity to compare disdrometer data to output 0
from hydrometeor classification schemes that are based
on dual-polarization radar observations. However, the where r is the radar range. The quantity d is only sig-
measurement accuracy of both instruments is strongly nificant in the Mie scattering regime, which applies to
affected by the severe nature of the storms, which contain rain drops with d . 2.3 mm at a temperature of 208C at X
hail and strong winds. To combine in situ microphysical band (Ryzhkov et al. 2011). Therefore, d must be esti-
data at the surface with three-dimensional radar imagery, mated before FDP and KDP can be used to correct the
microphysical data need to be quality controlled and rain attenuation. Each attenuation correction scheme differs
and hail particles must be discriminated. In addition, at- in the method used to calculate d. Anagnostou et al.
tenuation of the X-band radar signal must be corrected (2006) uses the differential reflectivity ZDR in an itera-
using algorithms that may be error prone, particularly tive approach to estimate d, while Steiner et al. (2009)
when the radar samples mixed-phase precipitation. A applies the iterative, finite impulse response filter from
proven algorithm to correct attenuation in hail does not Hubbert and Bringi (1995) to the measured CDP field.
yet exist (Borowska et al. 2011; Ryzhkov et al. 2013a), Once corrected Z, ZDR, and KDP are obtained, a fuzzy
although recent efforts to develop a scheme valid in logic hydrometeor classification scheme [section 3b(2)]
melting hail are presented in Ryzhkov et al. (2013a,b). can be applied to the radar data to determine the
Because supercell thunderstorms often contain large dominant hydrometeor type observed in each range
amounts of hail, attenuation correction schemes designed gate (Vivekanandan et al. 1999; Liu and Chandrasekar
for rain will not always yield accurate results. 2000; Iwanami et al. 2007; Park et al. 2009; Dolan and
In this paper, we compare attenuation-corrected radar Rutledge 2009; Snyder et al. 2010).
data and hydrometeor classifications to surface dis- To our knowledge, this study is one of the first to use
drometer measurements in supercell thunderstorms. disdrometer observations to analyze the performance of
Can disdrometer data be used to provide guidance on attenuation correction and hydrometeor classification
the performance of radar attenuation correction schemes for X-band radar measurements in severe
schemes, and therefore to provide a measure of radar thunderstorms. Most previous studies that compared
data quality? To investigate, we first apply a quality- radar and disdrometer data developed empirical re-
control algorithm and a hydrometeor classification lationships between reflectivity and rainfall rate (e.g.,
scheme for in situ disdrometer data that uses the particle Schuur et al. 2001; Ulbrich and Miller 2001; Bringi et al.
size and fall speed distributions from the disdrometer to 2003; Kanofsky and Chilson 2008; Huang et al. 2010). In
classify particles as rain, small hail (2 , d , 5 mm), and addition, such studies have primarily been conducted in
large hail (d . 5 mm; note that in this study, ‘‘large’’ is stratiform precipitation (e.g., Geotis 1978; Goddard
simply relative to the small hail class and is not meant to et al. 1982; Thomson and List 1996; Zhang et al. 2011),
be an argument against the typical definition of large hail while only a few comparisons have been performed in
of d . 20 mm). We then assess the performance of the severe thunderstorms (Schuur et al. 2001; Thurai et al.
attenuation correction scheme by comparing disdrometer- 2010, 2011). X-band radar attenuation correction
derived Z and ZDR to X-band radar Z and ZDR and to schemes have mostly been evaluated through compari-
S-band radar Z. Comparisons between the disdrometer son with S-band radar data, which are less attenuated
hydrometeor classification scheme and an existing scheme (e.g., Anagnostou et al. 2006; Steiner et al. 2009; Snyder
for X-band radar data are also provided. et al. 2010). Here, we propose an alternative method to
A brief review of the different techniques that can be evaluate attenuation correction schemes with surface
used to correct attenuation is now given. Several at- disdrometer data. While we do not argue that this ap-
tenuation correction schemes use the propagation dif- proach outperforms comparisons with S-band radar, the
ferential phase FDP and specific differential phase KDP question remains whether disdrometers could be used
to estimate the total and specific attenuation, re- if no nearby S-band radars are present. In addition, be-
spectively (e.g., Carey et al. 2000; Testud et al. 2000; cause the S-band Weather Surveillance Radar-1988

TABLE 1. Deployment details for the cases included in this analysis. All of the cases listed are supercell thunderstorms, except for the
squall line of 12 Jun 2010.
NOXP distance (km) WSR-88D distance (km)

Date Times (UTC) Location Disdrometers (beam height, m) (beam height, km)
17 May 2218–2333 Artesia, NM CU01, UF04–07 16–45 (280–920) KFDX 200–207 (4.1–4.3)
19 May 2059–2122 Kingfisher, OK CU01, UF01, UF05, UF07 29–38 (550–740) KTLX 168–173 (3.1–3.3)
2 Jun 2320–2326 Benkelman, NE UF03 23 (430) KGLD 67 (0.85)
7 Jun 0005–0143 Mitchell, NE CU01, UF01, UF05, UF07 30–37 (580–720) KCYS 123–138 (2.0–2.3)
9 Jun 0118–0139 Scottsbluff, NE CU01, UF01, UF03, UF05–06 24–27 (460–510) KCYS 103–107 (1.5–1.6)
12 Jun 2100–2203 Gruver, TX CU01, UF01, UF05–06 10–21 (190–390) KAMA 111–113 (1.7)
Doppler (WSR-88D) dual-polarization upgrade was not VORTEX2, two types of disdrometers (Fig. 2) were
complete in 2010, mobile radar ZDR measurements deployed: articulating disdrometers (denoted as UF01
corrected for differential attenuation and results from and UF03), with a measurement volume that was ori-
mobile radar hydrometeor classification algorithms ented continuously perpendicular to the 10-s running
cannot be validated with traditional S-band radar com- average of the particle trajectory of a 1.2-mm raindrop
parisons. Therefore, the goal of this study is to determine (Friedrich et al. 2013b), and stationary disdrometers
the relative quality of the VORTEX2 disdrometer and (denoted as CU01, UF04, UF05, UF06, and UF07), with
X-band radar data with the intention of improving the a measurement volume that remained fixed and parallel
results from future observational analyses and numerical to the ground. For supercell thunderstorms, the dis-
modeling studies that use these datasets for microphysical drometers were deployed in advance of the southern
retrievals. side of the forward-flank downdraft (Fig. 3), with the
mobile weather radars deployed to the southeast of the
thunderstorm. The distance between the disdrometers
2. Cases, instruments, and data collection
and the radar ranged from 10 to 45 km, with a median
a. Case selection distance of 20 km. Further details on the deployment
strategy are provided in Friedrich et al. (2013b).
Coordinated radar and disdrometer data were obtained
from ;36 severe thunderstorms during the second year of c. Radar measurements
VORTEX2, a 12-week field campaign conducted in the
Great Plains of the United States during May and June of Radar data were obtained from the National Oceanic
2009 and 2010 (Wurman et al. 2012). For this study, three and Atmospheric Administration (NOAA) X-band dual-
criteria were used for case selection: 1) radar data above polarized (NOXP) mobile radar (Palmer et al. 2009;
the disdrometer sites were available for at least 5 min, Burgess et al. 2010). Table 2 provides a summary of the
2) the disdrometers observed Z . 20 dBZ (approximate radar configuration during VORTEX2. For the cases
threshold between drizzle and light rain; Rinehart 2004) considered in this research, the size of the radar reso-
for at least 5 min, and 3) the distance between the radar lution volume above the disdrometer sites ranged from
and the disdrometers was less than 45 km. Based on these 74 m 3 87 m 3 175 m at 10-km range to 74 m 3 393 m 3
criteria, we consider data from five supercell thunder- 785 m at 45-km range in range, azimuth, and height,
storms and one squall line (total of ;183 min of data). respectively. The radar was electronically leveled, and
Table 1 provides details on the cases, and Fig. 1 shows a digital compass was used to record its heading. With
examples of radar reflectivity from each case at an eleva- respect to the location of the disdrometers, the height of
tion angle of 18. the lowest radar beam ranged from 0.2 to 1 km AGL, but
was mainly below 0.6 km AGL (Table 1).
b. Disdrometer measurements
The OTT PARSIVEL optical disdrometer (L€ offler-
3. Data processing
Mang and Joss 2000) is an integrated laser transmitter–
receiver that uses a 180-mm-long, 30-mm-wide, and a. Disdrometer
1-mm-thick light sheet to detect the diameter and fall
1) QUALITY CONTROL AND HYDROMETEOR
speed of precipitation particles. More information about
CLASSIFICATION SCHEME
the measurement principle can be found in L€ offler-
Mang and Joss (2000), L€ offler-Mang and Blahak (2001), A quality-control procedure (Fig. 4) was applied to the
Yuter et al. (2006), and references within. During stationary and articulating disdrometer data to address

FIG. 1. Plan position indicator displays of attenuation-corrected radar reflectivity measured by NOXP at 18 ele-
vation angle at (a) 2212 UTC 17 May 2010, (b) 2118 UTC 19 May 2010, (c) 2320 UTC 2 Jun 2010, (d) 0002 UTC 8 Jun
2010, (e) 0130 UTC 10 Jun 2010, and (f) 2136 UTC 12 Jun 2010. Disdrometer and radar locations are denoted by open
circles and filled squares, respectively. The arrow shows the storm motion direction.
three documented error sources: strong winds, particles PARSIVEL disdrometer data collected in a tropical
that only partially transect the sample volume (i.e., margin cyclone and two supercell thunderstorms with wind
fallers), and splashing (e.g., Sevruk 1982; Illingworth and speeds up to 30 m s21, using six stationary and two ar-
Stevens 1987; Ne^spor et al. 2000; Schuur et al. 2001; Kruger ticulating disdrometers. Misclassified particles, with d .
and Krajewski 2002; Barthazy et al. 2004; Thurai and 5 mm and unphysically slow fall speeds (y , 1 m s21),
Bringi 2005; Yuter et al. 2006; Friedrich et al. 2013a). were identified in the stationary (but not the articulat-
Friedrich et al. (2013a) analyzed wind-induced errors in ing) disdrometer data at wind speeds as slow as 10 m s21

FIG. 3. An idealized schematic that shows the disdrometer and

radar deployment strategy for supercell thunderstorms. The dis-
drometers were deployed in a line that was perpendicular to the
storm motion vector with an instrument spacing of 0.2–1 km. The
radar was deployed ahead of the forward-flank downdraft of
the thunderstorm and was always within 45 km of the disdrometer
deployments (ideally within 15 km). The location of the 40-dBZ
isoecho is shown in black.
Chandrasekar 2001). Therefore, small and large hail

need to be assigned different fractional water contents
when the transition (T-) matrix method is used to
compute the scattering properties of the particle size
distribution [PSD; see section 3a(2) and Table 3]. Ap-
pendix A demonstrates that for the PSDs in this study,
FIG. 2. A photograph of an articulating disdrometer (fore-
ground) and a stationary disdrometer (background) deployed in the sensitivity of disdrometer Z and ZDR to the small
Artesia, NM, on 17 May 2010. hail fractional water content is less than 0.1 dB, because
the small sample area of the PARSIVEL disdrometer
captures relatively few hailstones. Second, while rain-
and were observed consistently at wind speeds larger
drop fall speed is a well-defined function of diameter
than 20 m s21. To avoid misclassified particles in this
(Gunn and Kinzer 1949; Atlas et al. 1973), the various
analysis, we remove entire time steps in which the sta-
shapes, densities, and water loadings of hail make it
tionary disdrometers observed particles with d . 5 mm
and slow fall speeds (y , 1 m s21). Second, margin fallers
and splashing raindrops are removed from both types of
TABLE 2. NOXP radar characteristics for the 2010 VORTEX2
disdrometer data by eliminating raindrops with a fall field campaign.
speed more than 60% faster or slower than the fall
speed–diameter relationship for rain (Gunn and Kinzer Wavelength 3.21 cm (X band)
Transmission mode Simultaneous transmission and
1949; Atlas et al. 1973), based on a study of PARSIVEL reception (SHV)
disdrometer accuracy by Jaffrain and Berne (2011). Beamwidth 18
The decision to classify a particle as rain, small hail, or Range resolution 74 m
large hail is based on its diameter and fall speed (Fig. 4). Azimuthal resolution 0.58
While the reader is referred to Friedrich et al. (2013b) Elevation angles scanned 18 to # 158 (18 increments)
Maximum unambiguous range 60.0 km
for a detailed description of the classification scheme, Nyquist velocity 19.9 m s21
two notes are merited here. First, we distinguish be- Pulse repetition frequency 2500 Hz
tween small and large hail because small ice particles Moments and variables Reflectivity, Doppler velocity,
in thunderstorms are partially melted hailstones that spectrum width, differential
consist of a torus of liquid water that surrounds an ice reflectivity, differential phase,
and correlation coefficient
core (Rasmussen and Heymsfield 1987; Bringi and

FIG. 4. Fall speed vs diameter plot depicting the quality-control procedures and the hydro-
meteor classification scheme applied to the disdrometer data [adapted from Fig. 5 in Friedrich
et al. (2013b)]. The white solid lines in the rain, small hail, and large hail regions are the em-
pirical diameter–fall speed relationships for rain (Atlas et al. 1973), graupel (Locatelli and
Hobbs 1974), and hail (Knight and Heymsfield 1983), respectively.
possible for an ice particle to have a range of fall speeds (this issue is partially addressed by averaging the radar
for a given diameter, raising the possibility that a hail data; see section 3c). Next, the T-matrix method
particle could be erroneously classified as rain. In addi- (Vivekanandan et al. 1991; Bringi and Chandrasekar
tion, the disdrometer hydrometeor classification scheme 2001) was used to compute Z and ZDR for each 60-s time
(Fig. 4) uses the fall speed curves for dry graupel to
define the small hail region, following Friedrich et al.
(2013b). However, because the bulk density of small hail TABLE 3. Parameters used in the T-matrix program [i.e., canting
in thunderstorms exceeds that of graupel, the graupel angle (CA), axis ratio (AR), hydrometeor density (HD), fractional
water content (FWC), and temperature (T)]. The mean and stan-
fall speed relations are likely less than the fall speeds dard deviation are denoted by m and s, respectively. For compar-
observed here. Therefore, some of the small hail may be ison to NOXP radar (WSR-88D) data, calculations were
detected in the unclassified region between the small performed at a radar frequency of 9.41 GHz (2.895 GHz) and
hail and rain classes in Fig. 4. Because there is un- a radar elevation angle of 18 (0.58).
certainty over whether particles in this region are rain or Hydrometeor Parameters References
small hail, these particles are left unclassified. Appendix
Rain CA: m 5 08, s 5 7.58 Huang et al. (2008)
B explores the sensitivity of the results to whether par-
T: 158C
ticles in the unclassified region are included in the Small hail CA: m 5 08, s 5 508 Snyder et al. (2010);
analysis, revealing a mean sensitivity of ;0.2 dB in Z Ryzhkov et al. (2011)
and 0.01 dB in ZDR (relative to Z and ZDR obtained by AR: 0.8 Huang et al. (2005)
excluding the particles). HD: 0.9 g cm23 Vivekanandan et al.
(1993)
2) COMPUTATION OF METEOROLOGICAL FWC: 0.5 Huang et al. (2005)
VARIABLES FROM DISDROMETER DATA T: 08C
Large hail CA: m 5 08, s 5 508 Snyder et al. (2010);
Before calculating Z and ZDR from the disdrometer Ryzhkov et al. (2011)
data, number distributions were accumulated over pe- AR: 0.8 Knight (1986);
Balakrishnan and
riods of 60 s to obtain sufficiently large particle samples.
Zrnic (1990)
The 60-s accumulation time causes higher-frequency HD: 0.9 g cm23 Vivekanandan et al.
variations in the PSD to be lost. These variations would (1993); Solheim et al.
likely be retained in the radar data due to the excellent (1999)
range resolution (74 m), possibly causing the radar and FWC: 0.2 Aydin et al. (1998)
T: 08C
disdrometer measurements to represent different PSDs

ðr
step. Because the scattering properties of rain and ice
PIAH 5 AH (r) dr 5 g 3 [FDP (r) 2 FDP (0)] , (2)
particles differ, separate rain, small hail, and large hail 0
distributions were input to the T-matrix program for
0
each time step. For rain particles, the drop shape model ZH 5 ZH 1 PIAH , (3)
from Beard and Chuang (1987) was chosen, which re-
mains accurate at the large drop diameters present in where PIAH is the path-integrated attenuation at hori-
convective weather (i.e., within 64% of the measured zontal polarization (dB), r is the radar range (km), AH is
axis ratios at d . 5 mm; Thurai et al. 2009). The rain- the specific attenuation at horizontal polarization
drops were assumed to have a temperature of 158C, (dB km21), g is an empirical constant equal to 0.3006 in
a mean canting angle of 08, and a canting angle standard rain [average of g values in Table 3 of Anagnostou et al.
deviation of 7.58 (Huang et al. 2008). Of these settings, Z 0
(2006)], and ZH is the corrected reflectivity at horizontal
and ZDR display the greatest sensitivity to the canting polarization. Another modification to the Anagnostou
angle standard deviation; however, the change in ZDR is et al. (2006) technique was to relate the differential at-
generally less than 0.1 dB across the range of physically tenuation directly to PIAH following Park et al. (2005)
reasonable values (18–108; Kwiatkowski et al. 1995). and Gorgucci et al. (2006). Once an initial estimate of
Hail particles were assumed to have a canting angle PIAH was made, the path-integrated differential atten-
mean and standard deviation of 08 and 508, respectively uation PIADP was calculated as a function of PIAH [Eq.
(Snyder et al. 2010; Ryzhkov et al. 2011), and were as- (4)], and ZDR was corrected with Eq. (5):
signed a temperature of 08C. The hailstones were mod-
eled as a uniform mixture of ice and liquid water using the PIADP 5 « 3 PIAH , (4)
Maxwell–Garnett mixing formula, with ice (liquid water)
0
as the matrix and liquid water (ice) as the inclusions in ZDR 5 ZDR 1 PIADP , (5)
large (small) hailstones (Bringi and Chandrasekar 2001;
0
Ryzhkov et al. 2011). Table 3 summarizes the parameters where ZDR is the corrected ZDR and « 5 0:173 in rain.
used in the T-matrix program for rain, small hail, and 2) RADAR HYDROMETEOR CLASSIFICATION
large hail. SCHEME
b. Radar data processing Once radar data were corrected for attenuation, the
hydrometeor classification scheme from Snyder et al.
1) RADAR ATTENUATION CORRECTION SCHEME (2010) was applied. This fuzzy logic scheme was originally
The attenuation correction scheme from Steiner et al. devised at S band by Park et al. (2009) and adapted to X
(2009) is applied to the radar data. Steiner et al. (2009) band by Snyder et al. (2010). The algorithm uses the
evaluated the performance of this scheme with data that following dual-polarization radar variables as inputs: ZH,
were collected in convective storms during the 2002 ZDR, 10(logKDP), copolar cross-correlation coefficient at
International H2O Project (IHOP) by the National lag zero (rHV), reflectivity texture, and total differential
Observatory of Athens (NOA) X-band radar. For phase texture. The membership functions used by the
X-band radars with simultaneous horizontal and vertical scheme are trapezoidal in shape and are derived from
polarization (SHV) transmit, attenuation correction in T-matrix simulations of observed and idealized PSDs of
heavy rain was found to be most accurate when ZDR was rain and hail. The membership functions are defined for
not used to estimate the attenuation or differential at- six hydrometeor classes: ground clutter/anomalous pro-
tenuation (Steiner et al. 2009). Potential biases in ZDR pagation (GC/AP), biological scatterers (BS), big drops
caused by antenna and depolarization errors as CDP (BD), rain (RA), heavy rain (HR), and rain–hail mixture
increases (Ryzhkov and Zrnic 2007; Hubbert et al. (RH). Only the latter four classes are retained for
2010a,b; Zrnic et al. 2010) are thereby avoided. comparison with the PARSIVEL disdrometer, since
The Steiner et al. (2009) attenuation correction scheme the disdrometer does not detect GC/AP or BS. The
is modified from the differential phase-based algorithm output values of the membership functions are weighted
presented in Anagnostou et al. (2006). The first modifi- according to Park et al. (2009) and then summed for each
cation is that d was removed from CDP with five iterations hydrometeor class. The class with the largest sum is then
of the finite impulse response filter from Hubbert and assigned to the radar range gate.
Bringi (1995), rather than using ZDR to estimate d. The
c. Radar–disdrometer comparison method
smoothed FDP range profile was then used to estimate
the path-integrated attenuation at horizontal polarization To select data for comparison, radar Z and ZDR were
[Eq. (2)] and to correct ZH [Eq. (3)]: averaged over a 3 3 3 array of range gates, centered on

the gate that contained the disdrometer. Time steps with Of the 183 uncorrected radar Z values (97%), 177 are
sharp horizontal reflectivity gradients (.35 dB km21) weaker than the corresponding disdrometer observa-
and deployments with ground clutter (Z . 0 dBZ and tions, with a median difference (radar Z 2 disdrometer
near-zero Doppler velocity) near the disdrometer sites Z) of 217 dB (all data in Fig. 5a). When the attenuation
were excluded from the analysis. Radar and disdrometer correction scheme is applied, the distribution shifts to-
data were then paired so that the time difference be- ward larger radar Z (all data in Fig. 5b). The median
tween the observations did not exceed 30 s. difference (radar Z 2 disdrometer Z) after attenuation
Despite the exclusion of reflectivity gradients larger correction is 1.0 dB, and 48% of the radar Z values are
than 35 dB km21 and the averaging of radar range gates, weaker than the disdrometer values. These statistics
precipitation particle advection and the height differ- indicate that the attenuation correction scheme has re-
ence between the radar beam and the surface are po- moved the overall negative bias in the radar observa-
tentially large error sources in this analysis. As an tions. Twenty-five percent of the radar Z values are
example, consider a raindrop of d 5 1 mm, which has larger than the PARSIVEL disdrometer sampling un-
a terminal fall speed of y 5 4 m s21. If this drop is at certainty, 30% are smaller, and 45% are within the
a height of 1 km when the radar observes it, the drop will sampling uncertainty (gray shading in Fig. 5), quantified
not reach the ground until 250 s later. Assuming a mean by Jaffrain and Berne (2011). In the Jaffrain and Berne
horizontal wind speed of 10 m s21, the drop will be ad- study, two PARSIVEL disdrometers were collocated
vected 2.5 km downstream from the point in space and sampled ;990 h of light-to-moderate rainfall. The
where it was observed by the radar, a distance of nearly standard deviation of the difference in the moments
34 radar range gates (in the worst case scenario). Thus, (i.e., the sampling uncertainty) that were derived from
the PSDs observed by the radar and disdrometer may be each drop size distribution was then calculated (Tables
different due to the strong low-level winds that often B4 and B5 in Jaffrain and Berne 2011). Since the mea-
accompany supercell thunderstorms. A sensitivity test surements in this study were taken in severe thunder-
was performed in which the radar data were averaged storms that often contained heavier rainfall rates, hail,
over various windows, ranging from 3 3 3 to 11 3 11 and strong winds, the sampling uncertainty from Jaffrain
range gates (not shown). Aside from a ;0.25-dB im- and Berne (2011) should be considered a lower bound
provement in the ZDR agreement when a 3 3 3 aver- on the uncertainty in this research.
aging window was used versus no averaging, the results Scatterplots of ZDR before and after attenuation
are not affected by the size of the window. In addition, correction are presented in Figs. 5c and 5d, respectively.
no correlation was found between the disagreement in Differential attenuation is clearly evident in the un-
the radar and disdrometer data and the radar beam corrected radar ZDR, with 50% of the radar observations
height (not shown). Growth and evaporation of rain- having ZDR , 0 dB (Fig. 5c). After attenuation correc-
drops between the radar beam and the surface, however, tion was applied (Fig. 5d), the median difference (radar
are nevertheless error sources. ZDR 2 disdrometer ZDR) improves from 22.7 to
0.19 dB. Of the corrected observations, 38% are within
the sampling uncertainty of the PARSIVEL dis-
4. Results and discussion drometer (gray shaded region in Fig. 5d), compared to
a. Radar and disdrometer comparison of Z and ZDR 20% prior to correction. Similar to the reflectivity,
a large number of points lie outside the sampling un-
1) X-BAND RADAR Z AND ZDR certainty (gray shaded region in Fig. 5d), with 37%
Figure 5 presents scatterplots of disdrometer and ra- (25%) of the radar ZDR values larger (smaller) than the
dar Z for all cases in the analysis (Table 1; Fig. 1) before disdrometer sampling uncertainty.
(Fig. 5a) and after (Fig. 5b) the attenuation correction 2) S-BAND RADAR Z
scheme was applied. Two data subsets are identified,
To provide a benchmark for comparison, disdrometer
which will be discussed in sections 4b and 4c: the 17 May
Z is also compared to Z from the nearest WSR-88D at
2010 hailstorm (red plus signs) and data with radar signal
the 0.58 elevation angle (Fig. 6) for all cases in the
quality index (SQI) , 0.8 (blue plus signs). SQI
analysis. In comparing Fig. 5b to Fig. 6, we note that the
(SIGMET 2009) is related to signal-to-noise ratio (SNR)
median disagreement in radar and disdrometer Z is
and spectrum width (W):
changed from 1.0 to 21.9 dB (all plus signs) when the
S-band WSR-88D data are compared to the disdrometer-
SNR p2 W 2 derived data. However, the median disagreement in Z for
SQI 5 exp 2 . (6)
SNR 1 1 2 two of the data subsets (i.e., the 17 May 2010 hailstorm,

FIG. 5. Comparison of radar and disdrometer observations (left) before and (right) after attenuation correction
for (a),(b) Z and (c),(d) ZDR. The gray shaded region is the sampling uncertainty of the PARSIVEL disdrometer
taken from Jaffrain and Berne (2011). Uncertainties for Z . 50 dBZ and ZDR . 3 dB are outlined by the bold green
lines and were obtained via linear extrapolation. Observations from the hailstorm on 17 May 2010 are plotted in
red, while observations with radar SQI , 0.8 are plotted in blue. All other observations are plotted in black. Note
that 4 of the 51 observations from the hailstorm have SQI , 0.8 and are included in the hailstorm subset. The
median disagreement (radar 2 disdrometer) for all data is shown in the upper left, while the bottom right shows the
median disagreement for each subset. The number of observations in each plot is 183, consisting of cases described
in section 2a and Table 1.
red plus signs, and SQI , 0.8, blue plus signs) decreases see Table 1). These details will be explored further in
when WSR-88D data are used, from 5.8 to 21.5 dB sections 4b–d, in which three case studies are examined:
(hailstorm; red plus signs) and 213 to 20.66 dB (SQI , a supercell with large hail on 17 May 2010, a supercell on
0.8; blue plus signs). These results imply that the X-band 9 June 2010, and a squall line on 12 June 2010.
radar attenuation correction scheme partially contrib-
b. 17 May 2010: Supercell with radar Z and ZDR
utes to the disagreement in these subsets, since S-band
larger than disdrometer values
radar data compare more favorably with the dis-
drometer data (this inference, however, is complicated On 17 May 2010, a high precipitation supercell
by differences in the X- and S-band radar beam heights; thunderstorm was observed near Artesia, New Mexico.

the presence of hail (e.g., Steiner et al. 2009; Tabary

et al. 2009; Snyder et al. 2010; Borowska et al. 2011).
Attenuation correction in the presence of hail is un-
certain (Vulpiani et al. 2008; Borowska et al. 2011; Gu
et al. 2011) because the relationship between differential
phase and attenuation in a rain/hail mixture has only
recently been examined. Ryzhkov et al. (2013a,b) found
that the coefficients g and « in Eqs. (2) and (4) differ for
rain and melting hail. Therefore, correcting attenuation
in mixed-phase precipitation with coefficients meant for
rain may lead to large errors (Steiner et al. 2009).
Because it is known that attenuation correction
schemes designed for rain may perform poorly in the
presence of hail, disdrometers can be used to detect hail
on the ground and to flag radar observations beyond the
location of the disdrometers as poor quality. Figure 8
shows the disdrometer data accumulated over this event
as a function of particle fall velocity and diameter, with
large hail up to d 5 20 mm present (note that the mea-
FIG. 6. As in Fig. 5b, but for unattenuated S-band WSR-88D Z. surement limit of the PARSIVEL disdrometer is d 5
26 mm, and undercatchment is likely at these large sizes,
as discussed later). Figure 7c shows that between 2218
The disdrometers sampled the forward-flank downdraft and 2233 UTC, up to 15 cm3 of ice were observed per
of the storm (Fig. 1a). The authors observed large hail 1-min time step by CU01. Although the ice volume de-
(d ; 50 mm) between 2220 and 2232 UTC, which se- creases to ,5 cm3 min21 after 2233 UTC, a 5–15-dBZ
verely damaged the windshields of the deployment ve- discrepancy remains in Z, likely because the hail core is
hicles. Figure 7 shows a time series plot comparing Z located between the radar and disdrometer sites. Range
(Fig. 7a) and ZDR (Fig. 7b) values recorded by the radar gates located behind the hail core may exhibit erroneous
and disdrometer CU01 during this event. The total ice Z values due to the cumulative nature of attenuation
volume (assuming spherical particles) and the largest correction errors.
hail size observed by the disdrometer are also plotted An important limitation of this analysis is that large
with time (Figs. 7c,d). One might expect that the radar hail is sparse (number concentration of 1022 m23;
observations would not be heavily attenuated in the Straka 2009) relative to the sample area (54 cm2) of
portion of the storm nearest the radar and that the at- the PARSIVEL disdrometer. Therefore, disdrometer
tenuation could be accurately corrected. However, the data might not indicate that radar data are suspect if
time series data (Figs. 7a,b) reveal that, on average, only a few hailstones fall at the disdrometer site. In
corrected Z and ZDR measured by the X-band radar addition, even in thunderstorms with large amounts
were 8.3 and 1.5 dB larger, respectively, than the cor- of ice, the PARSIVEL disdrometer will underes-
responding measurements from CU01. In fact, during timate the hailstone concentration due to the small
the 27-min period shown, Z recorded by CU01 never sample area. Undercatchment is especially likely for
exceeded that of the radar, and the CU01 ZDR was the largest hailstones, as Fig. 7d demonstrates that the
greater than the radar ZDR for only 2 of the 10 time steps. largest hailstones detected by disdrometer CU01
The trend of generally larger attenuation-corrected radar during matched observations with the radar are d 5
variables relative to all of the disdrometer measurements 13 mm, despite hailstones of d 5 50 mm being ob-
made on 17 May is also shown in Figs. 5b and 5d (red plus served. In an attempt to determine whether under-
signs), a discrepancy that is not reflected in the remainder catchment is the primary cause of the disagreement
of the dataset (blue and black plus signs). (rather than deficiencies in the attenuation correc-
We hypothesize that the disagreement evident in the tion scheme), we compare disdrometer Z to that of
disdrometer and radar observations on 17 May is partly KFDX, the nearest S-band WSR-88D, at 0.58 eleva-
due to large, wet hail causing resonant (Mie) scattering tion angle (Fig. 6). For the hailstorm case (red plus
of the radar beam. Because of Mie scattering, strong signs in Fig. 6), the median disagreement is 21.5 dB
attenuation at horizontal polarization and differential when S-band radar is used, but jumps to 5.8 dB for
attenuation have been observed at C- and X-bands in attenuation-corrected X-band radar (Fig. 5b). Note,

FIG. 7. Time series data recorded by NOXP (solid lines) and disdrometer CU01 (dashed
lines) from the supercell thunderstorm with large hail (d ; 50 mm) observed on 17 May 2010:
(a) attenuation-corrected radar and disdrometer reflectivity, (b) attenuation-corrected
radar and disdrometer differential reflectivity, (c) disdrometer-observed ice volume, and
(d) disdrometer-observed maximum hail size. The error bars in (a),(b) represent the sampling
uncertainty of the PARSIVEL disdrometer.
however, that the two radars are sampling at different Between 0130 and 0140 UTC, the disdrometer
heights (NOXP at 0.6 km AGL and KFDX at 4.2 km recorded Z and ZDR values that are 10–15 dB and 1–
AGL; Table 1), which prevents us from making a de- 3 dB larger, respectively, than the values obtained by the
finitive conclusion about the cause of the improved radar after attenuation correction. During this time, SQI
agreement. ranges from 0.4 to 0.5, which is poor relative to the other
observations considered in this analysis (76% have SQI .
c. 9 June 2010: Supercell thunderstorm with radar Z
0.8). Data with SQI , 0.8 are plotted in blue in Fig. 5 and
and ZDR less than disdrometer values
exhibit smaller radar Z relative to the disdrometer, a
A supercell thunderstorm developed in the late af- trend that is not present in the other data subsets (black
ternoon of 9 June 2010 and moved into the western and red plus signs). We find that the median disagree-
Nebraska panhandle near Scottsbluff. The core of the ment in Z (radar Z 2 disdrometer Z) for SQI , 0.8 is
thunderstorm passed ;5 km south of the disdrometer 212 dB, while the median disagreement in Z for SQI .
deployments, which placed the disdrometers behind the 0.8 is 2.1 dB. These results indicate that it is much more
precipitation core relative to the NOXP radar, which likely for radar Z to be smaller than disdrometer Z when
was deployed south of the storm core (Fig. 9a). Because SQI is small and most of the radar signal has been lost
the heaviest precipitation passed between the radar and due to attenuation. To verify that the disagreement is
disdrometers, the radar data collected near the dis- not solely due to errors in the disdrometer data, we
drometer locations were heavily attenuated. Time series note from Fig. 6 (blue plus signs) that the comparison
of corrected radar and disdrometer (UF01) data are improves markedly for S-band radar data (median
shown in Figs. 10a and 10b. The radar SQI [Eq. (6)] is disagreement of 20.66 versus 213 dB for X-band radar
also plotted (Figs. 9b, 10c). data corrected for attenuation). The radar beam heights

FIG. 8. Accumulated particle counts recorded by disdrometer CU01 on 17 May 2010 binned by the observed fall
speed and diameter. The black lines represent the empirical fall speed–diameter relationships for rain, graupel, and
hail that are shown in Fig. 4. Hail bins are outlined in red.
are also more similar than in the 17 May hailstorm d 5 13 mm. From 2130 to 2136 UTC, hail resulted in
(0.5 km AGL for NOXP and 1.5 km AGL for KCYS; larger radar Z and ZDR values relative to the dis-
Table 1). This evidence supports our claim that a negative drometer data (Fig. 11), which is the same result found
bias exists in the attenuation-corrected NOXP radar data in section 4b. A radial stripe of suspiciously large radar Z
when SQI , 0.8. (.55 dBZ) and ZDR (.6 dB) values is also present in the
radar imagery at 2136 UTC (red circles in Fig. 12) in the
d. 12 June 2010: Squall line with radar Z and ZDR
southwestern portion of the squall line, possibly caused
similar to disdrometer values
by the presence of hail there.
At approximately 2100 UTC on 12 June 2010, a squall Following the passage of the convective cell and the
line developed near Gruver, Texas. Radar and dis- hail at ;2137 UTC at the disdrometer site, agreement in
drometer (UF05) time series data are shown in Fig. 11. Z and ZDR improves and the discrepancies are generally
The median attenuation-corrected radar and disdrometer within 5 and 0.5 dB, respectively. During the period of
Z and ZDR values are in closer agreement than in pre- agreement (2137–2155 UTC), the radar imagery depicts
vious case studies (5.0 and 1.2 dB, respectively). The best precipitation that is nearly uniform in time and space
agreement (to within 5 and 0.5 dB) is found from 2139 near the disdrometer site. Additionally, the observed ice
to 2155 UTC, with greater disagreement (.15 and volume remains small (median 0.7 cm3 min21), suggest-
.1.5 dB) during the first two time steps (i.e., 2130 and ing that hail is not biasing the attenuation correction.
2136 UTC). From 2130 to 2136 UTC, the radar data The large radar SQI (.0.95, not shown) and the relative
(Fig. 12) show that a small convective cell (;2 km in absence of hail are the likely reasons for the improved
north-to-south extent) near the leading edge of the agreement.
squall line moves over disdrometer UF05. The ice vol-
e. Radar and disdrometer hydrometeor classification
ume recorded by the disdrometer (Fig. 11c) is largest
comparisons
during this time period, peaking at 5.6 cm3 min21 at
2130 UTC, before dropping below 4 cm3 min21 after We now compare output from the radar and dis-
2139 UTC. The time series of maximum hail size drometer hydrometeor classification schemes. All of the
(Fig. 11d) and accumulated PSD (Fig. 13) also depict the disdrometer and radar data from Fig. 5 are included. The
hail from the convective cell and show hailstones up to disdrometer hydrometeor classes consist of rain (RA),

The comparison results are shown in Fig. 14. Sectors

that represent disagreement [i.e., the disdrometer (ra-
dar) observes hail, but the radar (disdrometer) does not]
are separated from the remainder of the chart. Of the
179 observations, 113 (63%) are in agreement. The most
common scenarios are that both schemes detect hail
(36% of the time) or that both detect rain (27% of the
time). Of the 60 observations that disagree, 13 (7%) of
them disagree because the radar scheme identifies hail
when the disdrometer does not observe any. One likely
explanation for this disagreement is that due to the
sparseness of the hail, no hailstones passed through the
sample volume of the disdrometer during the same time
step as observed by the radar. In addition, it is also
possible that the hail melted as it fell from the height of
the radar volume to the disdrometer. Since the center of
the radar beam was always 0.2–1 km AGL for the ob-
servations considered here, this possibility is most likely
when small hail was present at the height of the radar
beam, which would be more susceptible to complete
melting.
Disagreement in the remaining 53 observations (29%)
results from the disdrometer detecting hail when the
radar does not. It is important to note that the dis-
drometer ice scheme does not account for the number of
hailstones present; if just one hailstone is recorded
during a 60-s time step, a classification of hail is still
assigned by the disdrometer scheme, even though hail
may not be the dominant contributor to the corre-
sponding radar measurements. The median number of
FIG. 9. Plan position indicator observations of (a) attenuation- hailstones and median hailstone size observed in a 60-s
corrected radar reflectivity and (b) SQI for the supercell thun- time step by the disdrometer is greater when both the
derstorm observed by NOXP at 18 elevation angle at 0130 UTC radar and the disdrometer schemes agree that hail is
10 Jun 2010. Black open circles denote disdrometer locations.
present (median of two hailstones and 8.5 mm when the
The arrow shows the direction of storm motion. The distance
between each labeled tick mark is approximately 8 km in X and schemes agree versus one hailstone and 6.0 mm). The
11 km in Y. radar classification scheme uses the dual-polarization
radar variables to identify the dominant hydrometeor
class within the radar volume, but other hydrometeor
small hail (S. Hail), and large hail (L. Hail), while the types may nevertheless be present. Thus, it may be that
radar classes of interest are BD, RA, HR, and RH. the radar scheme does not assign the hail class when the
Because the radar scheme does not have firm rain rate or radar volume is dominated by rain and there are rela-
reflectivity thresholds that can be applied to the dis- tively few small hailstones present. This is possible be-
drometer data to discriminate between BD, RA, and cause the membership functions for rain and hail in the
HR, these radar classes are combined into a general rain radar fuzzy logic scheme overlap, but it is only the class
class. Therefore, the purpose of the comparison is to with the maximum rule strength (i.e., the maximum
analyze the agreement between the liquid and ice classes weighted sum of the membership function values) that is
of the two schemes. Because of error sources that in- assigned to the radar range gate. When the disdrometer
clude overlapping rain and hail signatures in the radar classifies hail but the radar does not, the median rule
data, errors in attenuation and differential attenuation strength of the radar hail (rain) class is 0.6 (1.5), com-
correction, and particle advection, we do not expect pared to 0.5 (1.5) when both instruments classify rain.
perfect agreement; however, we can nevertheless use However, the radar algorithm will not classify hail in
the data to explain under what conditions we expect either case because rain is the dominant scatterer in the
disagreement and why. radar volume.

FIG. 10. As in Fig. 7, but for the supercell thunderstorm observed by disdrometer UF01 on 9 Jun
2010. The radar SQI is shown in (c).
An additional factor that is likely to cause the the same time. If two 5-mm raindrops were present
PARSIVEL disdrometer to report hail when the radar in the sample volume, for instance, the disdrometer
does not is the possibility that multiple particles may would record a single 10-mm particle, which would
be present within the disdrometer sample volume at be classified as a hailstone (Fig. 4). For a simulated
FIG. 11. As in Fig. 7, but for the squall line observed by disdrometer UF05 on 12 Jun 2010.

FIG. 12. Plan position indicator observations of attenuation-corrected (a) radar reflectivity and (b) differential
reflectivity for the squall line observed by NOXP at 18 elevation angle at 2136 UTC 12 Jun 2010. The location of
disdrometer UF05 is denoted by the black open circle and the location of NOXP is annotated. The convective cell is
outlined in blue, and an area of large radar reflectivity and differential reflectivity is circled in red. The arrow in (a)
indicates the storm motion direction. The distance between each labeled tick mark is approximately 18 km in X and
11 km in Y.
heavy convective rain with intercept parameter N0 5 between the hail detections of the radar and dis-
1400 m23 mm21 and rainfall rate R 5 300 mm h21, drometer schemes is possible due to a combination of
L€offler-Mang and Joss (2000) found that the proba- particle coincidences in the disdrometer sample vol-
bility of coincidences in the PARSIVEL disdrometer ume and the use of the fuzzy logic scheme for classi-
sample volume is ;5%. Therefore, disagreement fying the radar data.
FIG. 13. As in Fig. 8, but for disdrometer UF05 on 12 Jun 2010.

FIG. 14. Pie chart comparing the outputs from the disdrometer and radar hydrometeor
classification schemes. The area of each sector in the pie chart is proportional to the percentage
of the total number of time steps (179) included in each sector. Each sector is labeled with the
class assigned by the disdrometer scheme (i.e., rain, small hail, and large hail) in bold, followed
by a solidus (/) and the class assigned by the radar scheme (i.e., rain and hail) in italics. The
number of time steps in each sector is also listed. Sectors in which the outputs from the two
schemes disagree have been separated from the rest of the chart.
5. Summary and conclusions Fig. 6). The discrepancy between the X-band radar data
and the disdrometer measurements is possibly due to the
In this paper, we applied an attenuation correction attenuation correction scheme overcorrecting the radar
scheme designed for rain (Steiner et al. 2009) to X-band data within and behind the hail core of the supercell
dual-polarization radar data collected by NOXP in five thunderstorm, although undercatchment of large hail-
supercell thunderstorms and one squall line during the stones by the disdrometer may have also contributed
VORTEX2 field campaign in 2010. The attenuation- to the disagreement. A second case study of a super-
corrected radar Z and ZDR were then compared to those cell thunderstorm (Fig. 10) demonstrated that the dis-
derived from PSDs recorded by PARSIVEL disdrometer tends to record larger Z and ZDR (by 13 and
drometers using the T-matrix program, which required 0.61 dB, respectively) than the X-band radar when the
assumptions to be made about the axis ratio, fractional radar signal quality is poor (SQI , 0.8). For the same
water content, and canting angle of the observed hail- data subset, only a 0.66-dB discrepancy in Z exists when
stones (see Table 3 and appendix A). The Snyder et al. the disdrometer and WSR-88D are compared (blue plus
(2010) hydrometeor classification scheme was then ap- signs in Fig. 6). Disagreement between the disdrometer
plied to the corrected radar data, and the results were and attenuation-corrected X-band radar data due to
compared to the output from a hydrometeor classifica- poor SQI is most likely to occur within trailing pre-
tion scheme that was developed for disdrometers in cipitation that is located behind heavy rainfall (relative
convective weather. to the radar location). A third case study analysis
When the disdrometer and attenuation-corrected ra- showed that when 1) large hail is not detected, 2) the
dar data were compared, it was shown that 45% (38%) radar signal quality is good (SQI . 0.8), and 3) the
of the Z (ZDR) observations agree to within the sam- precipitation structure is horizontally homogeneous,
pling uncertainty of the PARSIVEL disdrometer (Figs. Z and ZDR observations from both instruments agree to
5b,d). A case study analysis of a supercell thunderstorm within 5 and 0.5 dB (2137–2155 UTC in Fig. 11).
with large hail demonstrated that the attenuation- When the hydrometeor classification schemes for the
corrected X-band radar Z (ZDR) tends to be larger radar and the disdrometer are compared (Fig. 14), they
than the values recorded by the disdrometer by 8.3 dB agree 63% of the time. Disagreement results when the
(1.5 dB), respectively (Fig. 7a). However, when S-band radar scheme diagnoses hail and the disdrometer
WSR-88D and disdrometer Z are compared, the mea- scheme does not (7% of the observations) and when the
surements differ by only 21.5 dB (red plus signs in disdrometer observes hail and it is not detected by

the radar scheme (29%). Hail may be detected solely by expressed in this publication are those of the authors
the radar because small hail is present in the radar vol- and do not necessarily reflect the views of the National
ume that melts prior to reaching the surface or because Science Foundation.
the hail was especially sparse and not detected by the
disdrometer. When the opposite situation occurred (i.e., APPENDIX A
the disdrometer observed hail but the radar scheme
did not), it was shown that fewer, smaller hailstones
Sensitivity to Hailstone Characteristics in the
were observed by the disdrometer than when the two
T-Matrix Program
instruments both observed hail. In these cases, the
nonzero rule strength (i.e., the weighted sum of the To calculate Z and ZDR from disdrometer data that
membership function values) of the hail class suggests contain hail, characteristics of the hail must be specified,
that hail may have been present in the radar volume, but including axis ratio, fractional water content, and fall
because the radar hydrometeor classification scheme behavior. Table 3 provides the default values used in this
identifies the most dominant hydrometeor in the sam- analysis. However, since no measurements of axis ratio,
pling volume, the hail was not classified. fractional water content, or canting angle were made,
The data quality analysis presented here may be the chosen values represent a source of uncertainty.
particularly valuable to those who undertake future Here, we examine the sensitivity of the results to some of
VORTEX2 microphysical process and data assimila- the hailstone characteristics in the T-matrix scattering
tion studies. We have shown that attenuation of radar calculations.
data in severe thunderstorms can be substantial even To quantify the sensitivity, we select the PSD ob-
in the portion of the thunderstorm that is initially served by disdrometer CU01 on 17 May 2010 (Fig. 8),
penetrated by the radar beam. Further, since the as- since these data are from a hailstorm and should exhibit
sumptions in the attenuation correction scheme used the greatest sensitivity to the hailstone characteristics.
in this study are not valid in ice, large errors may result The quantities Z and ZDR are calculated for two values
in and beyond hail cores. This research may also be of the fractional water content of small hail (0.35 and
helpful to those who use the VORTEX2 disdrometer 0.65), axis ratio of large hail (0.5 and 0.65), and canting
observations to validate model-predicted surface angle standard deviation of all hail (358 and 658). These
precipitation types. With the hydrometeor classifica- values, together with those provided in Table 3, provide
tion scheme in Fig. 4, median diameter and number a reasonable range over which these characteristics can
concentration can be derived from the disdrometer be expected to vary (Knight 1986; Lesins and List 1986;
data for each hydrometeor class and compared to Ryzhkov et al. 2011). The resulting sensitivities in Z and
those from numerical models. ZDR for the entire time series recorded by CU01 are
shown in Fig. A1, relative to Z and ZDR obtained using
Acknowledgments. This research was sponsored by the default values of the parameters in Table 3. No
NSF ATM-0910424 and NSF DGE-1144083. Funding sensitivity is evident with respect to the small hail frac-
for Donald Burgess came from NOAA/Office of Oceanic tional water content (Fig. A1a). In contrast, up to 2-dB
and Atmospheric Research under NOAA–University of sensitivity in Z and 0.9 dB in ZDR is present when the
Oklahoma Cooperative Agreement NA11OAR430072. large hail axis ratio is varied (Fig. A1b), although these
Funding for NOXP data came from NSF ATM-0802717. sensitivities are only present in ;5 time steps of the
We thank Gwo-Jong Huang and Prof. Bringi (Colorado disdrometer data. The Z and ZDR display more consis-
State University) for the T-matrix program, in addition to tent sensitivity to hail canting angle standard deviation,
George Fernandez, Carlos Lopez, and Forrest Masters although this sensitivity remains within 0.2 and 0.6 dB,
(University of Florida), who designed the articulating respectively (Fig. A1c). Considering only the data in-
disdrometers and supplied four of the stationary dis- cluded in Fig. 5 from CU01 on 17 May (Table A1), the
drometers. We also thank Stephanie Higgins (University mean sensitivity in Z (ZDR) to large hail axis ratio and
of Colorado) for writing many of the routines that hail canting angle standard deviation, respectively,
process the disdrometer data. We are grateful to is just 20.01 dB (0.03 dB) and 0.05 dB (0.1 dB). Simi-
George Fernandez, Stephanie Higgins, Rachel Humphrey, lar results (Table A1) were obtained from the PSD
Scott Landolt, Carlos Lopez, Daniel Nuding, and Cameron observed by disdrometer UF05 on 12 June 2010
Redwine for deploying instruments during VORTEX2. (Fig. 13).
The helpful feedback from two anonymous reviewers In summary, because the PSDs in this study contain
substantially improved an earlier version of this manu- little hail relative to rain, Z and ZDR exhibit limited
script. Any opinions, findings, or recommendations sensitivity to the hailstone axis ratio, fractional water

FIG. A1. Sensitivity of Z (red lines) and ZDR (blue lines) from disdrometer CU01 on
17 May 2010 to (a) fractional water content of small hail, (b) axis ratio of large hail, and
(c) canting angle standard deviation of small and large hail. The sensitivities are relative
to the Z and ZDR obtained by using the default values of small hail fractional water
content (0.5), large hail axis ratio (0.8), and small and large hail canting angle standard
deviation (508).
content, and canting angle. This result does not mean 1987). In the disdrometer hydrometeor classification
that Z and Z DR are generally insensitive to hailstone scheme of Friedrich et al. (2013b), fall speed curves
characteristics. Rather, the small sample volume of for dry graupel (Locatelli and Hobbs 1974) are used
the PARSIVEL disdrometer caused relatively few to determine the fall speed thresholds for the small
hailstones to be observed, which reduced the sensi- hail region with 2 , d , 5 mm (Fig. 4). Because small
tivity of the results to the hailstone characteristics hailstones in supercell thunderstorms have a bulk
and meant that rain was the dominant contributor to density greater than that of graupel and are likely
Z and Z DR . embedded in a torus of liquid water, it is reasonable to
expect that these particle fall speeds are greater than
APPENDIX B the graupel curves offered by Locatelli and Hobbs
(1974), but somewhat less than the fall speeds for
pure rain given by Atlas et al. (1973). In the Friedrich
Sensitivity to Disdrometer Hydrometeor
et al. (2013b) classification scheme, such particles are
Classification Scheme
left unclassified (see Fig. 4) and are excluded from
A partially melted hailstone of a given diameter can T-matrix computations of Z and ZDR, since there is
exhibit a range of fall speeds depending on its fractional uncertainty about whether they should be modeled as
water content and bulk density (Rasmussen and Heymsfield rain or small hail.

TABLE A1. Mean sensitivity in disdrometer Z and ZDR to small tests: one that assumes the previously unclassified
hail FWC, axis ratio (AR) of large hail, small and large hail canting particles are rain, and a second that assumes the par-
angle standard deviation (sCA), and the disdrometer hydrometeor
ticles are hail. Figure B1a illustrates the sensitivity in
classification scheme for two subsets of the data in Fig. 5: obser-
vations from disdrometer CU01 on 17 May 2010 and observations Z (red lines) and ZDR (blue lines) for both the rain
from disdrometer UF05 on 12 Jun 2010. All values are relative to (solid lines) and hail (broken lines) tests. Up to 13-dB
those obtained using the default parameters listed in Table 3 and (1.5 dB) sensitivity in Z (ZDR) is evident for the time
the disdrometer hydrometeor classification scheme shown in Fig. 4. step at 2212 UTC. However, in all but one of the other
17 May 2010 (CU01) 12 Jun 2010 (UF05) 86 1-min time steps, negligible sensitivity is present.
DZ (dB) DZ (dB) The reason for the large sensitivity at 2212 UTC is due
Sensitivity test DZDR (dB) D ZDR (dB) to small disdrometer reflectivity (23 dBZ, not shown),
FWC 5 0.65 0.000 471 0.001 77 coupled with a relatively large particle of d 5 4.25 mm
20.000 195 20.000 917 within the unclassified region. Because of the small
FWC 5 0.35 20.000 391 20.001 36 total number of particles observed at 2212 UTC, the
0.000 138 0.000 702
particle within the unclassified region has marked in-
AR 5 0.65 20.0110 0.0152
0.009 54 0.0114 fluence on Z and ZDR. In heavy rainfall typical of the
AR 5 0.5 20.004 70 0.0399 thunderstorms observed in this study, however, Z and
0.0340 0.0110 ZDR are relatively insensitive to whether particles in
sCA 5 358 0.0412 0.0669 the unclassified region are discounted or assumed to
0.130 0.190
be rain or hail. Table A1 indicates that for the data
sCA 5 658 0.0507 0.0687
0.116 0.141 included in Fig. 5 from CU01 on 17 May, the mean
Unclassified particles 0.193 0.0401 sensitivity to the unclassified region is less than 0.2 and
are rain 0.0116 20.001 68 0.02 dB for Z and ZDR, respectively. In fact, when all
Unclassified particles 0.121 0.0105 small hail particles are removed from the 17 May PSD,
are small hail 20.0133 20.006 87
Fig. B1b shows that Z and ZDR change by less than
Small hail particles 20.004 43 20.0158
removed 0.001 94 0.0103 0.2 dB for the entire time series recorded by CU01.
For the data included in Fig. 5, the mean change in Z
(ZDR) is 20.004 dB (0.002 dB; see Table A1). These
To examine the sensitivity of Z and ZDR to these tests demonstrate that despite the uncertainty in small
excluded particles, we again use the data recorded by hailstone characteristics, the impact on the results is
disdrometer CU01 during the 17 May 2010 hailstorm small, likely because the PSDs examined here contain
(Fig. 8) and compute Z and ZDR for two different little small hail compared to rain.
FIG. B1. As in Fig. A1, but for (a) the precipitation type of the particles in the unclassified region
in Fig. 4 and (b) the small hail region in Fig. 4. The sensitivities are relative to the Z and ZDR
obtained by (a) excluding the unclassified particles and (b) including the small hail particles.

REFERENCES during VORTEX2. Mon. Wea. Rev., 141, 1182–1203,

doi:10.1175/MWR-D-12-00116.1.
Anagnostou, M. N., E. N. Anagnostou, and J. Vivekanandan, 2006: Geotis, S. G., 1978: Comparison of reflectivity measurements by
Correction for rain path specific and differential attenuation of radar and by disdrometer. J. Appl. Meteor., 17, 1403–1405,
X-band dual-polarization observations. IEEE Geosci. Remote doi:10.1175/1520-0450(1978)017,1403:CORMBR.2.0.CO;2.
Sens., 44, 2470–2480, doi:10.1109/TGRS.2006.873204. Goddard, J. W. F., S. M. Cherry, and V. N. Bringi, 1982: Com-
Atlas, D., R. C. Srivastava, and R. S. Sekhon, 1973: Doppler radar parison of dual-polarization radar measurements of rain
characteristics of precipitation at vertical incidence. Rev. Geo- with ground-based disdrometer measurements. J. Appl.
phys. Space Phys., 11, 1–35, doi:10.1029/RG011i001p00001. Meteor., 21, 252–256, doi:10.1175/1520-0450(1982)021,0252:
Aydin, K., S. H. Park, and T. M. Walsh, 1998: Bistatic dual- CODPRM.2.0.CO;2.
polarization scattering from rain and hail at S- and C-band Gorgucci, E., V. Chandrasekar, and L. Baldini, 2006: Correction of
frequencies. J. Atmos. Oceanic Technol., 15, 1110–1121, X-band radar observation for propagation effects based on the
doi:10.1175/1520-0426(1998)015,1110:BDPSFR.2.0.CO;2. self-consistency principle. J. Atmos. Oceanic Technol., 23,
Balakrishnan, N., and D. S. Zrnic, 1990: Use of polarization to 1668–1681, doi:10.1175/JTECH1950.1.
characterize precipitation and discriminate large hail. J. Atmos. Grzych, M. L., B. D. Lee, and C. A. Finley, 2007: Thermodynamic
Sci., 47, 1525–1540, doi:10.1175/1520-0469(1990)047,1525: analysis of supercell rear-flank downdrafts from Project
UOPTCP.2.0.CO;2. ANSWERS. Mon. Wea. Rev., 135, 240–246, doi:10.1175/
Barthazy, E., S. G€oke, R. Schefold, and D. H€ ogl, 2004: An optical MWR3288.1.
array instrument for shape and fall velocity measurements of Gu, J.-Y., A. Ryzhkov, P. Zhang, P. Neilley, M. Knight, B. Wolf,
hydrometeors. J. Atmos. Oceanic Technol., 21, 1400–1416, and D.-I. Lee, 2011: Polarimetric attenuation correction in
doi:10.1175/1520-0426(2004)021,1400:AOAIFS.2.0.CO;2. heavy rain at C band. J. Appl. Meteor. Climatol., 50, 39–58,
Beard, K. V., and C. H. Chuang, 1987: A new model for the doi:10.1175/2010JAMC2258.1.
equilibrium shape of raindrops. J. Atmos. Sci., 44, 1509–1524, Gunn, R., and G. D. Kinzer, 1949: The terminal velocity of fall for
doi:10.1175/1520-0469(1987)044,1509:ANMFTE.2.0.CO;2. water droplets in stagnant air. J. Meteor., 6, 243–248,
Borowska, L., A. Ryzhkov, D. Zrnic, C. Simmer, and R. Palmer, doi:10.1175/1520-0469(1949)006,0243:TTVOFF.2.0.CO;2.
2011: Attenuation and differential attenuation of 5-cm- Huang, G. J., V. N. Bringi, S. van den Heever, and W. Cotton, 2005:
wavelength radiation in melting hail. J. Appl. Meteor. Climatol., Polarimetric radar signatures from RAMS microphysics. Pre-
50, 59–76, doi:10.1175/2010JAMC2465.1. prints, 32nd Int. Conf. on Radar Meteorology, Albuquerque,
Bringi, V. N., and V. Chandrasekar, 2001: Polarimetric Doppler NM, Amer. Meteor. Soc., P11R.6. [Available online at http://
Weather Radar: Principles and Applications. Cambridge Uni- ams.confex.com/ams/pdfpapers/96261.pdf.]
versity Press, 662 pp. ——, ——, and M. Thurai, 2008: Orientation angle distributions after
——, T. D. Keenan, and V. Chandrasekar, 2001: Correcting an 80-m fall using a 2D video disdrometer. J. Atmos. Oceanic
C-band radar reflectivity and differential reflectivity data for Technol., 25, 1717–1723, doi:10.1175/2008JTECHA1075.1.
rain attenuation: A self-consistent method with constraints. ——, ——, R. Cifelli, D. Hudak, and W. A. Petersen, 2010: A
IEEE Trans. Geosci. Remote Sens., 39, 1906–1915, doi:10.1109/ methodology to derive radar reflectivity–liquid equivalent
36.951081. snow rate relations using C-band radar and a 2D video dis-
——, V. Chandrasekar, J. Hubbert, E. Gorgucci, W. L. Randeu, drometer. J. Atmos. Oceanic Technol., 27, 637–651, doi:10.1175/
and M. Schoenhuber, 2003: Raindrop size distribution in dif- 2009JTECHA1284.1.
ferent climatic regimes from disdrometer and dual-polarized Hubbert, J., and V. N. Bringi, 1995: An iterative filtering technique for
radar analysis. J. Atmos. Sci., 60, 354–365, doi:10.1175/ the analysis of coplanar differential phase and dual-frequency
1520-0469(2003)060,0354:RSDIDC.2.0.CO;2. radar measurements. J. Atmos. Oceanic Technol., 12, 643–648,
Burgess, D. W., E. R. Mansell, C. M. Schwarz, and B. J. Allen, 2010: doi:10.1175/1520-0426(1995)012,0643:AIFTFT.2.0.CO;2.
Tornado and tornadogenesis events seen by the NOXP ——, S. M. Ellis, M. Dixon, and G. Meymaris, 2010a: Modeling,
X-band, dual-polarization radar during VORTEX2 2010. error analysis, and evaluation of dual-polarization variables
Preprints, 25th Conf. on Severe Local Storms, Denver, CO, obtained from simultaneous horizontal and vertical polariza-
Amer. Meteor. Soc., 5.2. [Available online at http://ams. tion transmit radar. Part I: Modeling and antenna errors. J.
confex.com/ams/25SLS/techprogram/paper_176164.htm.] Atmos. Oceanic Technol., 27, 1583–1598, doi:10.1175/
Carey, L. D., S. A. Rutledge, D. A. Ahijevych, and T. D. Keenan, 2010JTECHA1336.1.
2000: Correcting propagation effects in C-band polarimetric ——, ——, ——, and ——, 2010b: Modeling, error analysis, and
radar observations of tropical convection using differential evaluation of dual-polarization variables obtained from si-
propagation phase. J. Appl. Meteor., 39, 1405–1433, doi:10.1175/ multaneous horizontal and vertical polarization transmit ra-
1520-0450(2000)039,1405:CPEICB.2.0.CO;2. dar. Part II: Experimental data. J. Atmos. Oceanic Technol.,
Dolan, B., and S. A. Rutledge, 2009: A theory-based hydrometeor 27, 1599–1607, doi:10.1175/2010JTECHA1337.1.
identification algorithm for X-band polarimetric radars. Illingworth, A. J., and C. J. Stevens, 1987: An optical disdrometer
J. Oceanic Atmos. Technol., 26, 2071–2088, doi:10.1175/ for the measurement of raindrop size spectra in windy con-
2009JTECHA1208.1. ditions. J. Atmos. Oceanic Technol., 4, 411–421, doi:10.1175/
Friedrich, K., S. A. Higgins, F. J. Masters, and C. R. Lopez, 2013a: 1520-0426(1987)004,0411:AODFTM.2.0.CO;2.
Articulating and stationary PARSIVEL disdrometer mea- Iwanami, K., K. Kusunoki, N. Orikasa, M. Maki, R. Misumi, and
surements in conditions with strong winds and heavy rainfall. M. Murakami, 2007: Hydrometeor type classification in winter
J. Atmos. Oceanic Technol., 30, 2063–2080, doi:10.1175/ clouds using X-band polarimetric radar measurements—
JTECH-D-12-00254.1. Comparison of X-band polarimetric radar data with in-situ
——, E. A. Kalina, F. J. Masters, and C. R. Lopez, 2013b: Drop-size measurements by HYVIS. Preprints, 33rd Conf. on Radar
distributions in thunderstorms measured by optical disdrometers Meteorology, Cairns, QLD, Australia, Amer. Meteor. Soc.,

P10.11. [Available online at http://ams.confex.com/ams/ MCS. Wea. Forecasting, 24, 730–748, doi:10.1175/
pdfpapers/123072.pdf.] 2008WAF2222205.1.
Jaffrain, J., and A. Berne, 2011: Experimental quantification of the Park, S. G., V. N. Bringi, V. Chandrasekar, M. Maki, and K. Iwanami,
sampling uncertainty associated with measurements from 2005: Correction of radar reflectivity and differential reflec-
PARSIVEL disdrometers. J. Hydrometeor., 12, 352–370, tivity for rain attenuation at X band. Part I: Theoretical and
doi:10.1175/2010JHM1244.1. empirical basis. J. Atmos. Oceanic Technol., 22, 1621–1632,
Kanofsky, L., and P. B. Chilson, 2008: An analysis of errors in drop doi:10.1175/JTECH1803.1.
size distribution retrievals and rain bulk parameters with Rasmussen, R. M., and A. J. Heymsfield, 1987: Melting and shed-
a UHF wind profiling radar and a two-dimensional video ding of graupel and hail. Part I: Model physics. J. Atmos.
disdrometer. J. Atmos. Oceanic Technol., 25, 2282–2292, Sci., 44, 2754–2763, doi:10.1175/1520-0469(1987)044,2754:
doi:10.1175/2008JTECHA1061.1. MASOGA.2.0.CO;2.
Knight, N. C., 1986: Hailstone shape factor and its relationship to radar Rinehart, R. E., 2004: Radar for Meteorologists. Rinehart Publi-
interpretation of hail. J. Climate Appl. Meteor., 25, 1956–1958, cations, 482 pp.
doi:10.1175/1520-0450(1986)025,1956:HSFAIR.2.0.CO;2. Ryzhkov, A. V., and D. S. Zrnic, 2007: Depolarization in ice
——, and A. J. Heymsfield, 1983: Measurement and interpreta- crystals and its effect on radar polarimetric measurements.
tion of hailstone density and terminal velocity. J. Atmos. J. Atmos. Oceanic Technol., 24, 1256–1267, doi:10.1175/
Sci., 40, 1510–1516, doi:10.1175/1520-0469(1983)040,1510: JTECH2034.1.
MAIOHD.2.0.CO;2. ——, M. Pinsky, A. Pokrovsky, and A. Khain, 2011: Polarimetric
Kruger, A., and W. F. Krajewski, 2002: Two-dimensional video dis- radar observation operator for a cloud model with spectral
drometer: A description. J. Atmos. Oceanic Technol., 19, 602–617, microphysics. J. Appl. Meteor. Climatol., 50, 873–894,
doi:10.1175/1520-0426(2002)019,0602:TDVDAD.2.0.CO;2. doi:10.1175/2010JAMC2363.1.
Kwiatkowski, J. M., A. B. Kostinski, and A. R. Jameson, 1995: The ——, M. R. Kumjian, S. M. Ganson, and A. P. Khain, 2013a:
use of optimal polarizations for studying the microphysics of Polarimetric radar characteristics of melting hail. Part I:
precipitation: Nonattenuating wavelengths. J. Atmos. Oceanic Theoretical simulations using spectral microphysical mod-
Technol., 12, 96–114, doi:10.1175/1520-0426(1995)012,0096: eling. J. Appl. Meteor. Climatol., 52, 2849–2870, doi:10.1175/
TUOOPF.2.0.CO;2. JAMC-D-13-073.1.
Lesins, G. B., and R. List, 1986: Sponginess and drop shed- ——, ——, ——, and P. Zhang, 2013b: Polarimetric radar char-
ding of gyrating hailstones in a pressure-controlled icing acteristics of melting hail. Part II: Practical implications.
wind tunnel. J. Atmos. Sci., 43, 2813–2825, doi:10.1175/ J. Appl. Meteor. Climatol., 52, 2871–2886, doi:10.1175/
1520-0469(1986)043,2813:SADSOG.2.0.CO;2. JAMC-D-13-074.1.
Liu, H., and V. Chandrasekar, 2000: Classification of hydrometeors Schuur, T. J., A. V. Ryzhkov, D. S. Zrnic, and M. Sch€
onhuber, 2001:
based on polarimetric radar measurements: Development of Drop size distributions measured by a 2D video disdrometer:
fuzzy logic and neuro-fuzzy systems, and in situ verifica- Comparison with dual-polarization radar data. J. Appl. Me-
tion. J. Atmos. Oceanic Technol., 17, 140–164, doi:10.1175/ teor., 40, 1019–1034, doi:10.1175/1520-0450(2001)040,1019:
1520-0426(2000)017,0140:COHBOP.2.0.CO;2. DSDMBA.2.0.CO;2.
Locatelli, J. D., and P. Hobbs, 1974: Fall speed and masses of solid Sevruk, B., 1982: Methods of correction for systematic error in
precipitation particles. J. Geophys. Res., 79, 2185–2197, point precipitation measurement for operational use. Opera-
doi:10.1029/JC079i015p02185. tional Hydrology Rep. 21, Publ. 589, 91 pp. [Available from
L€
offler-Mang, M., and J. Joss, 2000: An optical disdrometer for WMO, Case Postale 2300, CH-1211, Geneva, Switzerland.]
measuring size and velocity of hydrometeors. J. Atmos. Oceanic Shabbott, C. J., and P. M. Markowski, 2006: Surface in situ ob-
Technol., 17, 130–139, doi:10.1175/1520-0426(2000)017,0130: servations within the outflow of forward-flank down drafts of
AODFMS.2.0.CO;2. supercell thunderstorms. Mon. Wea. Rev., 134, 1422–1441,
——, and U. Blahak, 2001: Estimation of the equivalent radar re- doi:10.1175/MWR3131.1.
flectivity factor from measured snow size spectra. J. Appl. SIGMET, 2009: Vaisala user’s manual: Digital IF receiver/Doppler
Meteor., 40, 843–849, doi:10.1175/1520-0450(2001)040,0843: signal processor RVP8. SIGMET, 451 pp. [Available online at
EOTERR.2.0.CO;2. ftp://ftp.sigmet.com/outgoing/manuals/RVP8_Users_Manual.pdf.]
Markowski, P. M., J. M. Straka, and E. N. Rasmussen, 2002: Direct Snyder, J. C., H. B. Bluestein, G. Zhang, and S. J. Frasier, 2010:
surface thermodynamic observations within the rear-flank Attenuation correction and hydrometeor classification of
downdrafts of nontornadic and tornadic supercells. Mon. Wea. high-resolution, X-band, dual-polarized mobile radar mea-
Rev., 130, 1692–1721, doi:10.1175/1520-0493(2002)130,1692: surements in severe convective storms. J. Atmos. Oceanic
DSTOWT.2.0.CO;2. Technol., 27, 1979–2001, doi:10.1175/2010JTECHA1356.1.
Ne^spor, V., W. F. Krajewski, and A. Kruger, 2000: Wind- Solheim, F. S., R. Vivekanandan, R. H. Ware, and C. Rocken, 1999:
induced error of raindrop size distribution measurement using Propagation delays induced in GPS signals by dry air, water
a two-dimensional video disdrometer. J. Atmos. Oceanic Tech- vapor, hydrometeors, and other particulates. J. Geophys. Res.,
nol., 17, 1483–1492, doi:10.1175/1520-0426(2000)017,1483: 104, 9663–9670, doi:10.1029/1999JD900095.
WIEORS.2.0.CO;2. Steiner, M., G. Lee, S. M. Ellis, and J. Vivekanandan, 2009:
Palmer, R., and Coauthors, 2009: Weather radar education at the Quantitative precipitation estimation and hydrometeor iden-
University of Oklahoma—An integrated interdisciplinary tification using dual-polarization radar—Phase II. NCAR
approach. Bull. Amer. Meteor. Soc., 90, 1277–1282, doi:10.1175/ Tech. Rep., 74 pp. [Available online at ftp://ftp.eol.ucar.edu/
2009BAMS2738.1. pub/temp/users/sellis/Steiner_etal_2009/2009%20KMA-NIMR
Park, H., A. V. Ryzhkov, D. S. Zrnic, and K. Kim, %20Report.pdf.]
2009: The hydrometeor classification algorithm for the Straka, J. M., 2009: Cloud and Precipitation Microphysics: Principles
polarimetric WSR-88D: Description and application to an and Parameterizations. Cambridge University Press, 406 pp.

Tabary, P., G. Vulpiani, J. J. Gourley, A. J. Illingworth, and Vivekanandan, J., W. M. Adams, and V. N. Bringi, 1991: Rigorous
O. Bousquet, 2009: Unusually high differential attenuation at approach to polarimetric radar modeling of hydrometeor
C band: Results from a two-year analysis of the French distributions. J. Appl. Meteor., 30, 1053–1063, doi:10.1175/
Trappes polarimetric radar data. J. Appl. Meteor. Climatol., 1520-0450(1991)030,1053:RATPRM.2.0.CO;2.
48, 2037–2053, doi:10.1175/2009JAMC2039.1. ——, R. Raghavan, and V. N. Bringi, 1993: Polarimetric radar
Testud, J., E. Le Bouar, E. Obligis, and M. Ali-Mehenni, 2000: modeling of mixtures of precipitation particles. IEEE Trans.
The rain profiling algorithm applied to polarimetric weather Geosci. Remote Sens., 31, 1017–1030, doi:10.1109/36.263772.
radar. J. Atmos. Oceanic Technol., 17, 332–356, doi:10.1175/ ——, S. M. Ellis, R. Oye, D. S. Zrnic, A. V. Ryzhkov, and J. Straka,
1520-0426(2000)017,0332:TRPAAT.2.0.CO;2. 1999: Cloud microphysics retrieval using S-band dual-polarization
Thomson, A. D., and R. List, 1996: Raindrop spectra and updraft radar measurements. Bull. Amer. Meteor. Soc., 80, 381–388,
determination by combining Doppler radar and disdrometer. doi:10.1175/1520-0477(1999)080,0381:CMRUSB.2.0.CO;2.
J. Atmos. Oceanic Technol., 13, 465–476, doi:10.1175/ Vulpiani, G., P. Tabary, J. Parent du Chatelet, and F. Marzano,
1520-0426(1996)013,0465:RSAUDB.2.0.CO;2. 2008: Comparison of advanced radar polarimetric tech-
Thurai, M., and V. N. Bringi, 2005: Drop axis ratios from a 2D niques for operational attenuation correction at C band.
video disdrometer. J. Atmos. Oceanic Technol., 22, 966–978, J. Atmos. Oceanic Technol., 25, 1118–1135, doi:10.1175/
doi:10.1175/JTECH1767.1. 2007JTECHA936.1.
——, M. Szakall, V. N. Bringi, K. V. Beard, S. K. Mitra, and Wurman, J., D. Dowell, Y. Richardson, P. Markowski, E. Rasmussen,
S. Borrmann, 2009: Drop shapes and axis ratio distributions: D. Burgess, L. Wicker, and H. Bluestein, 2012: The second
Comparisons between 2D video disdrometer and wind tunnel Verification of the Origins of Rotation in Tornadoes Experi-
measurements. J. Atmos. Oceanic Technol., 26, 1427–1432, ment: VORTEX2. Bull. Amer. Meteor. Soc., 93, 1147–1170,
doi:10.1175/2009JTECHA1244.1. doi:10.1175/BAMS-D-11-00010.1.
——, W. A. Petersen, and L. D. Carey, 2010: DSD characteristics of Yuter, S. E., D. E. Kingsmill, L. B. Nance, and M. Loffler-Mang,
a cool-season tornadic storm using C-band polarimetric radar and 2006: Observations of precipitation size and fall speed char-
two 2D-video disdrometers. Proc. Sixth European Conf. on Radar acteristics within coexisting rain and wet snow. J. Appl. Me-
Meteorology and Hydrology, Sibiu, Romania, European Meteo- teor. Climatol., 45, 1450–1464, doi:10.1175/JAM2406.1.
rological Society, 1–6. [Available online at http://www.erad2010. Zhang, G., S. Luchs, A. V. Ryzhkov, M. Xue, L. Ryzhkova, and
org/pdf/oral/tuesday/radpol1/5_ERAD2010_0101.pdf.] Q. Cao, 2011: Winter precipitation microphysics characterized
——, ——, A. Tokay, C. Schultz, and P. Gatlin, 2011: Drop size dis- by polarimetric radar and video disdrometer observations in
tribution comparisons between PARSIVEL and 2-D video dis- central Oklahoma. J. Appl. Meteor. Climatol., 50, 1558–1570,
drometers. Adv. Geosci., 30, 3–9, doi:10.5194/adgeo-30-3-2011. doi:10.1175/2011JAMC2343.1.
Ulbrich, C. W., and N. E. Miller, 2001: Experimental test of the effects Zrnic, D. S., R. J. Doviak, G. Zhang, and A. V. Ryzhkov, 2010:
of Z–R law variations on comparison of WSR-88D rainfall Bias in differential reflectivity due to cross-coupling
amounts with surface rain gauge and disdrometer data. Wea. through the radiation patterns of polarimetric weather
Forecasting, 16, 369–375, doi:10.1175/1520-0434(2001)016,0369: radars. J. Atmos. Oceanic Technol., 27, 1624–1637, doi:10.1175/
ETOTEO.2.0.CO;2. 2010JTECHA1350.1.

Comparison of Disdrometer and X-Band Mobile Radar Observations in Convective Precipitation

Uploaded by

Copyright:

Available Formats

Comparison of Disdrometer and X-Band Mobile Radar Observations in Convective Precipitation

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Comparison of Disdrometer and X-Band Mobile Radar Observations in Convective Precipitation

Uploaded by

Copyright:

Available Formats

2414 MONTHLY WEATHER REVIEW VOLUME 142

Comparison of Disdrometer and X-Band Mobile Radar Observations in Convective

EVAN A. KALINA AND KATJA FRIEDRICH

(Manuscript received 28 January 2014, in final form 11 March 2014)

1. Introduction low-level cold pool, which changes the near-surface

Ó 2014 American Meteorological Society

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

NOXP distance (km) WSR-88D distance (km)

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

FIG. 3. An idealized schematic that shows the disdrometer and

Chandrasekar 2001). Therefore, small and large hail

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

the presence of hail (e.g., Steiner et al. 2009; Tabary

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

The comparison results are shown in Fig. 14. Sectors

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

FIG. 13. As in Fig. 8, but for disdrometer UF05 on 12 Jun 2010.

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

REFERENCES during VORTEX2. Mon. Wea. Rev., 141, 1182–1203,

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

Unauthenticated | Downloaded 12/19/22 05:38 PM UTC

You might also like