Conference PaperPDF Available

Distributed Beams: A Technique to Reduce the Scan Time of an Active Rotating Phased-Array Radar System

Authors:

Abstract and Figures

Important requirements for a future generation of weather surveillance radars include improvements in data quality, rapid-update volumetric data, and the ability to perform adaptive weather observations. Phased Array Radar (PAR) is a candidate system capable of providing the required functionality. The National Oceanic and Atmospheric Administration (NOAA) is considering an affordable rotating PAR architecture to improve the capabilities of the current parabolic-reflector-based Weather Surveillance Radar – 1988 Doppler (WSR-88D) operational network. To achieve the current and future needs to support the National Weather Service mission, a concept of operations for the rotating PAR architecture has to be developed. The distributed beams technique introduced in this paper provides a way to reduce the scan-update times by spoiling the transmit beam while receiving multiple digital beams as the radar rotates. Specifically, the azimuthal rotation rate of the platform is matched to the desired sampling (spatial and temporal) such that the coherent processing interval for a given beam position is distributed over several overlapping receive beams. That is, the scan time can be reduced by a factor equal to the transmit beam spoil factor. This exploits the use of spoiled transmit beams in azimuth and allows a faster rotation rate which leads to rapid updates. We illustrate this technique and provide a trade-off analysis in the context of rotating PAR weather surveillance network.
Content may be subject to copyright.
1
DISTRIBUTED BEAMS: A TECHNIQUE TO REDUCE THE SCAN TIME OF AN ACTIVE
ROTATING PHASED ARRAY RADAR SYSTEM
David Schvartzman* and Sebastián M. Torres
Cooperative Institute for Mesoscale Meteorological Studies, University of Oklahoma, and
NOAA/OAR/National Severe Storms Laboratory, Norman, Oklahoma
1. INTRODUCTION
Key requirements for a future generation of
weather surveillance radars include improvements
in data quality, a more rapid update of volumetric
data, and the ability to perform adaptive weather
observations. Although the operational U.S.
Weather Surveillance Radar 1988 Doppler
(WSR-88D) network exceeded its expected life
span, its intrinsic architecture limitations prevent it
from attaining the performance levels required to
meet the set of next generation requirements.
Phased Array Radar (PAR) is a candidate that
may provide the required functionality by exploiting
its unique capabilities. The National Oceanic and
Atmospheric Administration (NOAA) is considering
an affordable rotating PAR (RPAR) architecture to
exceed the capabilities of the current reflector-
based WSR-88D network (Weber 2019). To
achieve current and future needs to support the
National Weather Service (NWS) mission, a
concept of operations (CONOPS) for weather
surveillance using the RPAR has to be developed.
The RPAR architecture has been used for air
surveillance and defense applications since the
late 1970’s (Palumbo and Cucci 1977, Palumbo
1996, Brookner 2002), but was only introduced for
weather surveillance in recent years (Yoshikawa
2013, Ushio 2014, Orzel and Frasier 2018). The
CONOPS for these weather RPAR systems
consisted of either imitating the operation of
conventional reflector radar or performing a
straightforward electronic scan in elevation while
mechanically rotating in the azimuthal direction.
These limited operational concept modes do not
fully exploit RPAR’s unique capabilities and are not
likely to meet demanding functional requirements
such as the more rapid update volumetric data.
The NOAA Radar Functional Requirements
document (NOAA/NWS 2015) specifies the
functionality expected for a future weather
surveillance radar system. The threshold functional
requirements define the minimum expected
performance, while the optimal functional
requirements define the desired performance. One
8B.2
Fig. 1. Illustration of the Distributed Beams (DB) technique.
*Corresponding author address: David Schvartzman, 120
David L. Boren Blvd, Room 4427, Norman, OK, 73072;
e-mail: David.Schvartzman@noaa.gov.
2
of the most demanding optimal requirements is the
1-minute update time to complete a volume scan
with no degradation of the sensitivity, spatial
resolution or standard deviation of measurement
for radar variable estimates. To achieve this
volume-update-time requirement, sophisticated
scanning and digital signal processing techniques
that exploit RPAR’s unique capabilities are likely to
be needed.
This paper introduces a novel distributed
beams (DB) technique which provides a way to
reduce scan update times. It is accomplished by
synthesizing a wide (spoiled) transmit beam and
using digital beamforming on receive as the radar
rotates. Returns from subsequent receive beams
pointing in the same direction are then grouped
coherently. Specifically, the azimuthal rotation rate
of the platform is matched to the desired sampling
(spatial and temporal) such that the coherent
processing interval (CPI) for a given beam position
is distributed over several overlapping receive
beams. In this manner, the scan time can be
reduced by a factor equal to the transmit beam
spoil factor (herein F). This exploits the use of
spoiled transmit beams in azimuth and allows a
faster rotation rate, leading to more rapid updates
without degradation in data quality. A schematic of
this technique is shown in Fig. 1.
This introduction section provided justification
and the concept of a way to reduce the scan times
of RPAR architecture. The rest of the paper is
structured into four more sections. Section 2
provides an overview of the Distributed Beams
CONOPS introduced and offers two DB
applications. Section 3 describes DB practical
implementation and calibration methods. Section 4
takes the theoretical analysis of the DB technique
further by capturing and presenting real data
collection using DB. Data are collected by scanning
a fixed target (for calibration), and an actual
weather front for a comparative demonstration of
the DB CONOPS to that of a conventional weather
radar. Section 5 summarizes the paper.
2. DISTRIBUTED BEAMS (DB) RPAR
CONOPS OVERVIEW
Active PAR technology allows the synthesis of
arbitrary antenna beam patterns on transmission
by varying the magnitude and phase of transmit
signals at each individual array element level (also
referred to as tapering). This capability can be used
to produce a wider transmit beam, effectively
increasing the beam coverage. This comes at the
expense of increased antenna pattern sidelobe
levels, reduced sensitivity, and slightly increased
beamwidth. For example, an active PAR antenna
with an intrinsic radiation pattern that produces a
narrow “pencil” beam (as defined by the one-way 3
dB width), can also be used to synthesize transmit
beams that are wider than the intrinsic narrow
beam of the system. Examples of these pencil and
spoiled transmit patterns are presented in Fig. 2.
Transmission of spoiled beams has been proposed
Fig. 2. Simulated one-way antenna radiation patterns for a narrow pencil beam (left), a
beam spoiled by a factor of three (center), and a beam spoiled by a factor of five (right).
Sectors correspond to azimuthal cuts of the antenna patterns.
3
for stationary PAR systems to reduce the scan time
(Weber et. al, 2005). Depending on the PAR
architecture, several receive beams can be
simultaneously formed digitally within the transmit
beam. The digitally formed receive beams use the
full antenna aperture to produce a narrow pencil
beam. However, these narrow receive beams
combined with the wide transmit beam result in
lowering sidelobe levels of the two-way antenna
pattern.
A scan strategy for the RPAR using the DB
technique is now defined. Assume the antenna is
rotating in azimuth at a constant speed of ω ͦ /s, the
broadside transmit beam is spoiled by a factor F,
and F beams are simultaneously formed within the
transmit beam on receive to sample with a one-
beamwidth (θ1) spacing in azimuth (as illustrated in
Fig. 1 for a factor F = 5). The beamwidth is defined
as the angular width within which the microwave
radiation is one-half of its peak intensity (Doviak
and Zrnić 1993). Finally, let us define the CPI as a
set of M pulses at a pulse repetition time (PRT) of
Ts seconds. With this, the rotation speed can be set
to
 


to transmit the desired CPI (MTs) over the angular
sampling of θ1. There are two applications
considered here for the DB CONOPS. The two
possibilities with the DB technique are: 1) Scan
Time Reduction and 2) Data Quality Improvement.
Both are now briefly described, followed by a
comparison to conventional parabolic-reflector
radar and tradeoff considerations in the context of
a practical implementation.
1. Scan time reduction: The number of pulses
per CPI is reduced MDB=M/F, and the rotation
speed is increased to ωDB = Fω. As the PAR
rotates at ωDB, MDB pulses are received on
each receive beam. In a continuous rotation
regime, the centers of resolution volumes
sampled by F beams received every MDBTs
seconds (coming from distinct transmit
beams) are coming from the same azimuth
location. Samples received on these beams
are combined to recover the MDBF=M pulses
required to obtain the desired data quality.
Operating the radar under this concept results
in reducing the scan time by a factor F.
2. Data quality improvement: The number of
pulses per CPI and the rotation speed are
maintained at M and ω. Similar to the previous
scenario, F beams received every MTs
seconds (coming from distinct transmit
beams) are coming from the same azimuth
location. Samples received on these beams
are combined to obtain MF pulses. Increasing
the number of samples by F results in a
significant reduction in the variance of radar-
variable estimates.
Comparing the process described in the first
application to conventional radar with a parabolic-
reflector antenna, the DB technique exploits the
PAR beamforming capability to reduce the scan
time. This increases rotation speed, but comes at a
price of increased rotation speed, increased
sidelobe levels (i.e., for the same antenna
aperture), reduced sensitivity, and slightly
increased beamwidth. Nevertheless, some of
these limitations can be mitigated. As argued by
Leifer et al. (2013), the rotating machinery has
been around for a long time and has a high
technology-readiness level, which reduces the risk
of deploying PAR pedestals capable of rotating at
higher rates. The sidelobe levels of the two-way
pattern could be reduced to the desired levels by
increasing the aperture size by a small amount,
which would also improve the beamwidth. The
sensitivity loss can be recovered by increasing the
power radiated by each array element.
Section 3 advances the theoretical aspects of
DB by presenting a practical implementation of the
technique and by discussing important antenna
calibration considerations.
3. DB IMPLEMENTATION ON THE ATD
The Advanced Technology Demonstrator
(ATD) is an active S-band planar dual-polarization
PAR that is funded jointly by the National Oceanic
and Atmospheric Administration (NOAA) and the
Federal Aviation Administration (FAA). It is being
developed by the National Severe Storms
Laboratory (NSSL), the Cooperative Institute for
Mesoscale Meteorological Studies (CIMMS) at the
University of Oklahoma, MIT Lincoln Laboratory,
and General Dynamics Mission Systems (Stailey
and Hondl 2016, Conway et al. 2018, Torres et al.
2019). This proof-of-concept system makes use of
pulse compression waveforms to meet sensitivity
and range-resolution requirements (Schvartzman
4
and Torres, 2019). The antenna is composed of 76
panels, where each panel consists of an 88 set of
radiating patch-antenna elements with dual linear
polarization (H and V), for a total of 4864 elements.
This arrangement of antenna elements spaced by
λ/2 results in a ~44 m aperture that produces a
beam that is ~1.58° wide at broadside. On receive,
the antenna is partitioned into overlapped
subarrays (consisting of 8 panels each) to produce
lower sidelobes and suppress grating lobes outside
of the main beam of the subarray pattern (Herd et
al. 2005). The operating frequency band of the
antenna is 2.7-3.1 GHz. Through element-level
control of the magnitude and phase of transmitted
signals, this system is capable of synthesizing
arbitrary beam patterns on transmission.
The spoiled transmit beams produced by the
ATD are synthesized using phase-only coefficients
to maximize the power on transmit (Brown et. al.,
2006). The co-polar main lobes of these antenna
patterns were measured using the calibration
infrastructure installed in the vicinity of the ATD
(Ivić et al, 2019), and are shown in Fig. 3 (axes
were scaled to enhance visual interpretation).
These measurements play an important role for the
calibration and successful implementation of the
DB technique. That is, not only the magnitude of
the two-way patterns has to be calibrated, but also
their phases to ensure a coherent transition across
the F receive beams to be coherently processed.
The DB technique was implemented on the
ATD using a transmit beam spoiled by a factor of
five and forming a set of nine simultaneous receive
beams sampled at ½ θ1 (sampling similar to super-
resolution on the WSR-88D). The uncalibrated two-
way beam patterns were measured for all receive
beams. Measurements were used to calibrate the
gain and phase of the beams. Gain calibration was
straightforward; that is, signal power was
compensated digitally by the relative difference
between the peak gain of the two-way broadside
beam (TW #5 in Fig. 4) and each other receive
beam (Ivić and Schvartzman 2019). This is shown
in the next set of panels in Fig. 4. Achieving phase
calibration required two considerations: (1) similar
to the gain calibration, instantaneous phases of the
two-way peaks were measured and digitally
corrected, and (2) a constant phase compensation
was also needed to align the phase centers for the
two-way distributed beams. Note that the second
phase calibration is a result of the mechanical
rotation of ωMTs degrees, which shifts the phase
centers of the two-way distributed beams because
the center of rotation is not exactly the center of the
antenna.
(a)
(b)
(c)
dB
dB
dB
5
Phase calibration is critical for the DB
technique to achieve the scan-time reduction of F.
If the phases of the signals of the DB samples are
not coherent across two-way beam transitions, the
combined time-series data cannot be coherently
processed. Any loss of coherency from sample to
sample would prevent the use of conventional
pulse-pair or spectral processing methods (e.g.,
clutter filtering). An example of spectral processing
will be shown in Section 4.
While the ATD was not designed for operating
in a constant rotation regime, it is capable of
scanning while rotating to scan in an RPAR
CONOPS. This allows us to explore techniques in
the context of a RPAR concept of operations.
Nevertheless, the pedestal is not prepared to rotate
at high speeds and therefore, this proof-of-concept
implementation demonstrates the second
application described in Section 2 (Data Quality
Improvement). Specifically, sets of data were
collected with the ATD rotating at ω = 4.1 °/s, with
M = 64 at a PRT of Ts = 3 ms over a ~20° sector,
and the DB technique was used to improve the
quality of estimates (beyond the requirements).
The antenna was set at 0° elevation, and the
transmit beam was electronically steered to 0.5° in
elevation and maintained at broadside on azimuth.
We are planning on implementing motion-
Fig. 4. Azimuthal slice of the measured two-way normalized ATD antenna patterns at 0.5°
elevation: uncalibrated (top), calibrated (bottom).
6
compensated steering, by which beams will be
electronically steered by a small angle to maintain
pointing on the center of the resolution volume
being sampled. This would exploit PARs beam
agility and also mitigate beam-smearing effects.
In the next section, illustrations of the DB
technique are presented with a scan over a
stationary point target (i.e., a radio tower), and a
scan of actual weather echoes from a stratiform
precipitation system.
4. DATA COLLECTION USING DB
The first illustration of the DB technique was
accomplished by combining returns from nine
receive beams, as the radar rotates past a
stationary point target. The radar boresight was
commanded to rotate from 290° to 310° azimuth
with respect to North. The target is located at 31.65
km in range, and at 297° azimuth. The purpose of
this test was to verify that the radar executed the
scan by rotating at the commanded speed, and that
the absolute location of the target was the same for
all receive beams. In addition, it was verified that
Fig. 5. Reflectivity from a stationary point target collected with the ATD on 12 Nov 2019. The
bottom panels show the reflectivity computed from the two-way broadside beam (left, RX
Beam #5) and the reflectivity with the DB technique (right).
7
the spread of azimuth angles for pulses across DB
was less than 0.12°.
The reflectivity fields for the receive beams
collected with the ATD on 12 Nov 2019 are shown
in the top nine panels of Fig. 5. A dotted line was
drawn at 300° azimuth for reference. The two
bottom panels show the reflectivity computed from
the two-way broadside beam (left, RX Beam #5)
and the reflectivity with the DB technique (right). It
is apparent that the increased number of samples
obtained with DB processing improved the SNR
and the recovery of weaker reflectivity echoes (i.e.,
recovered sensitivity). Given that the resolution of
the target’s location did not change after DB
processing, it can be inferred that spatial resolution
was not degraded.
The Doppler spectrum for the stationary target
was computed using all 576 samples combined
using DB (9 × 64), both before and after phase
correction. The resulting spectra are shown in Fig.
6. It is apparent on the uncorrected spectrum that
there was no coherency on the set of pulses,
resulting in replicas of a distorted spectrum. After
applying the described phase calibration and the
correction for the shift on phase centers the
spectrum now appears to have been corrected.
Next, Fig.7 shows the reflectivity fields for the
receive beams collected with the ATD observing a
stratiform precipitation system on 20 Nov 2019.
These are shown in the top nine panels of Fig. 7.
The radar boresight was commanded to rotate
counterclockwise from 100° to 80° in azimuth. The
two bottom panels show the reflectivity computed
from the two-way broadside beam (RX Beam #5)
and the reflectivity with the DB technique.
Comparing these two reflectivity fields, note the
appreciable reduction in the variance of estimates
when using the DB technique. The field looks
smoother and with better-defined weather features.
This is due to the SNR gain obtained from
combining many additional samples. Furthermore,
weaker echoes on the edge of the precipitation
system closer to the radar are also recovered.
This example illustrates the DB application for
improving data quality as described in Section 2. If
the radar had been operated at five times the
speed (~20.5 °/s), the estimated DB reflectivity field
would be similar to that from the two-way broadside
beam (RX Beam #5) with the same data quality but
in 20% of the time.
Fig. 6. Doppler spectra for the stationary point-target without phase correction (orange
curve) and with phase correction (black curve).
8
Fig. 7. Reflectivity from weather echoes collected with the ATD on 20 Nov 2019. The bottom
panels show the reflectivity computed from the two-way broadside beam (RX Beam #5) and
the reflectivity with the DB technique.
9
5. SUMMARY
Achieving the optimal requirements introduced
in the NOAA/NWS Radar Functional Requirements
for future weather surveillance radar will likely
require exploiting capabilities of advanced radar
systems. The RPAR architecture may be an
affordable candidate radar to replace the WSR-
88D and meet the demanding requirements. By
exploiting PARs unique capabilities in conjunction
with advanced signal processing techniques, it may
be possible to design a CONOPS to meet these
requirements.
Further investigation of techniques to exploit
PARs unique capabilities in the context of a
rotating CONOPS is ongoing. The novel
Distributed Beams technique described could be
one of the tools that helps the RPAR achieve the
required update times (~1 min). In this paper, the
DB technique was introduced and two applications
of it were described. A discussion on some of its
limitations and possible ways of addressing them
was given, and illustrations of the technique for a
stationary point-target and a weather system were
presented through an implementation on NSSL’s
ATD system. Even though achieving high reduction
factors may be challenging due to the high rotation
speeds required, an operational implementation
with only a small time-reduction factor (e.g., F = 2),
would still bring a significant improvement in
reducing the scan time, and would function at lower
rotation speeds (~8-15 °/s).
The preliminary results presented show that
the DB technique can be used to reduce the scan
time or to improve the data quality (i.e., reduce the
variance of estimates) in a RPAR. There are plans
to study the performance of the technique to
produce polarimetric-variable estimates, and the
possibility of integrating it with other techniques to
reduce the scan time while maintaining data quality
and spatial sampling. It is hoped that the outcome
of our research efforts will eventually inform the
design of a future U.S. Weather Surveillance Radar
Network.
ACKNOWLEDGMENT
The authors would like to thank Dr. Dušan Zrnić
(NOAA-NSSL), Dr. Antone Kusmanoff (University
of Oklahoma, JHLP), and Jami Boettcher (CIMMS-
OU) for useful discussions and comments that
improved this paper. We would also like to extend
our appreciation to the entire ATD team. Special
thanks go to Christopher Schwarz (University of
Oklahoma, CIMMS), Daniel Wasielewski (NOAA-
NSSL), Rafael Mendoza (NOAA-NSSL), John
“Chip” Murdock (GD-MS), and Henry Thomas
(MIT-LL) for their support configuring and operating
the ATD.
The authors would also like to acknowledge the
contributions of numerous engineers, students,
scientists, and administrators who have supported
these developments over the last decade.
This conference paper was prepared with funding
provided by NOAA/Office of Oceanic and
Atmospheric Research under NOAA-University of
Oklahoma Cooperative Agreement
#NA16OAR4320115, U.S. Department of
Commerce. The statements, findings, conclusions,
and recommendations are those of the authors and
do not necessarily reflect the views of NOAA or the
U.S. Department of Commerce.
REFERENCES
E. Brookner, 2002: Phased array radars-past, present
and future, RADAR Conference, Edinburgh, UK.
pp. 104-113.
doi: 10.1109/RADAR.2002.1174663
Brown, G., C. Kerce, and M. Mitchell, 2006: Extreme
beam broadening using phase only pattern
synthesis,” Proc. of the IEEE Workshop on
Sensor Array and Multichannel Processing,
Waltham, MA. doi: 10.1109/SAM.2006.1706079
Conway, M. D., D. Du Russel , A. Morris, and C. Parry,
2018: Multifunction phased array radar advanced
technology demonstrator nearfield test results.
Proc. of the IEEE,
doi: https://doi.org/10.1109/RADAR.2018.837877
1
Doviak, R. J., and D. S. Zrnić, 1993: Doppler Radar and
Weather Observations. Academic Press, 562 pp.
Herd, J. S., S. M. Duffy, and H. Steyskal, 2005: Design
considerations and results for an overlapped
subarray radar antenna. Preprints, IEEE
Aerospace Conf., Big Sky, MT,
USA.doi:10.1109/AERO.2005.1559399
Ivić, I. R., and D. Schvartzman 2019: A first look at the
ATD data corrections. Preprints, 39th
International Conference on Radar Meteorology,
Nara, Japan. Amer. Meteor. Soc., Paper 2-06.
Extended abstract available:
https://cscenter.co.jp/icrm2019/program/data/abs
tracts/Poster2-06_2.pdf
Ivić, I., C. Curtis, , E. Forren, R. Mendoza, D.
Schvartzman, S. Torres, D.J. Wasielewski, and
F.A. Zahrai, 2019: An Overview of Weather
10
Calibration on the Advanced Technology
Demonstrator. IEEE International Symposium on
Phased Array Systems and Technology,
Waltham, MA. Available:
https://www.researchgate.net/publication/338019
168_An_Overview_of_Weather_Calibration_on_t
he_Advanced_Technology_Demonstrator
Leifer, M. C., V. Chandrasekar, and E. Perl, 2013: Dual
polarized array approaches for MPAR air traffic
and weather radar applications, IEEE
International Symposium on Phased Array
Systems and Technology, Waltham, MA, pp. 485-
489.
doi: 10.1109/ARRAY.2013.6731876
NOAA/NWS, 2015: Radar Functional Requirements,
NOAA/NWS Internal Rep., 58pp. Available:
https://www.roc.noaa.gov/WSR88D/PublicDocs/
NOAA_Radar_Functional_Requirements_Final_
Sept%202015.pdf
K. A. Orzel and S. J. Frasier, 2018: Weather Observation
by an Electronically Scanned Dual-Polarization
Phase-Tilt Radar, in IEEE Transactions on
Geoscience and Remote Sensing, 56, pp. 2722-
2734. doi: 10.1109/TGRS.2017.2782480
B. Palumbo and A. Cucci, 1977: Array antenna with
multiple, independently scanned beams for a 3D
radar, 1977 Antennas and Propagation Society
International Symposium, Stanford, CA, USA, pp.
185-188.
doi: 10.1109/APS.1977.1147771
B. Palumbo, 1996: Some examples of systems
developments in Italy based on phased array
technology, Proceedings of International
Symposium on Phased Array Systems and
Technology, Boston, MA, USA, pp. 444-449. doi:
10.1109/PAST.1996.566134
Schvartzman, D., and S. Torres, 2019: Design of
Practical Pulse Compression Waveforms for
Polarimetric Phased Array Radar. Preprints, 39th
International Conference on Radar Meteorology,
Nara, Japan. Amer. Meteor. Soc., Paper 15A-02.
Available at:
https://cscenter.co.jp/icrm2019/program/data/abs
tracts/Session15A-02_2.pdf
Stailey, J. E., and K. D. Hondl, 2016:
Multifunction phased array radar for aircraft and
weather surveillance. Proc. IEEE, 104,
649659, doi: https://doi.org/10.1109/
JPROC.2015.2491179
Torres, S., T. Buckman C. Curtis E. Forren S. Gregg I.
Ivic R. Mendoza D. Schvartzman C. Schwarz D.
Wasielewski A. Zahrai, 2019: Towards an
operational demonstration of the first full-scale
polarimetric phased-array radar. Preprints, 39th
International Conference on Radar Meteorology,
Nara, Japan. Amer. Meteor. Soc., Paper 2A-04.
Available at: https://cscenter.co.jp/icrm2019/
program/data/abstracts/Session2A-04_1.pdf
Ushio, T., and Coauthors, 2014: Development and
observation of the phased array radar at X band.
XXXIth URSI General Assembly and Scientific
Symposium (URSI GASS), IEEE. doi:10.1109
/URSIGASS.2014.6929643
Weber, M., J. Cho, J. Flavin, J. Herd, and M. Vai, 2005:
Multi-function phased array radar for U.S. civil-
sector surveillance needs. Preprint, 32nd Conf. on
Radar Meteorology, Albuquerque, NM, Amer.
Meteor. Soc.
http://ams.confex.com/ams/pdfpapers/96905.pdf
Weber, M., 2019: Meteorological Phased Array Radar
Research at NOAA’s National Severe Storms
Laboratory, 2019 IEEE International Conference
on Microwaves, Antennas, Communications and
Electronic Systems (COMCAS), Tel-Aviv, Israel.
doi: 10.1109/COMCAS44984.2019.8958067
Yoshikawa, E. et al., 2013: MMSE Beam Forming on
Fast-Scaning Phased Array Weather Radar. IEEE
Transactions on Geoscience and Remote
Sensing, 51, no. 5, pp. 3077-3088, doi:
10.1109/TGRS.2012.2211607
... Panels are organized as follows: the top row corresponds to scan 1, the middle row corresponds to the forward-looking beams in scan 2, and the bottom row corresponds to the back-scanning beams in scan 2; the columns from left to right show fields of Zh, ZDR, DP, and hv. The distributed beams (DB) technique [40] [41] was used to exploit the simultaneously received beams in azimuth and to produce data from the forward-looking beams with reduced variance. Qualitative comparison of radarvariable fields shows no apparent data artifacts from either scan. ...
Article
Full-text available
The Rotating Phased Array Radar (RPAR) is an architecture that could improve the capabilities of the current Weather Surveillance Radar-1988 Doppler (WSR-88D) operational network and is likely to be more affordable than other candidate PAR architectures. However, continuous antenna rotation coupled with the need to perform coherent processing of multiple samples results in a degraded effective beamwidth (referred to as beam smearing) compared to architectures based on stationary antennas. The RPAR's beam agility can be exploited to reduce beam-smearing effects by electronically steering the beam on a pulse-to-pulse basis within the coherent processing interval. That is, the motion of the antenna can be compensated to maintain the beam pointed at the center of resolution volume being sampled. This motion-compensated steering (MCS) could reduce the effects of antenna motion and lead to a reduction in the effective beamwidth. The purpose of this article is to present and demonstrate the MCS technique for a dual-polarization RPAR system. We provide a formulation for the MCS technique, simulations to quantify its performance in mitigating beam-smearing effects, its impacts on the quality of dual-polarization radar-variable estimates, and a practical implementation on the National Severe Storms Laboratory's Advanced Technology Demonstrator (ATD) system. Experiments were carried out using two alternative Concepts of Opearations (CONOPS) described in this article. Results show that a system designed with sufficient pointing accuracy can be operated as an RPAR using MCS, and the impact on radar-variable estimates is comparable to that obtained when operating the same system as a stationary PAR.
... It allows a faster rotation speed to be maintained, leading to more rapid updates potentially without degradation in data quality. Alternatively, if the rotation speed is maintained, the number of samples can be increased by a factor of RF, which leads to reduced variance of radar-variable estimates [25]. The DB technique introduced in this article was implemented on NSSL's recently deployed Advanced Technology Demonstrator (ATD) PAR system and it is the first practical demonstration of digital beamforming in azimuth for dual-polarization weather RPAR. ...
Article
Full-text available
Important requirements for a future generation of weather surveillance radars include improvements in data quality and more rapid update of volumetric data. Phased Array Radar (PAR) is a candidate technology capable of providing the required functionality. The rotating PAR (RPAR) is a potential architecture that could improve the capabilities of the current parabolic-reflector-based US Weather Surveillance Radar – 1988 Doppler (WSR-88D) operational network and is more affordable than other candidate PAR architectures. However, RPAR concept of operations that support observational needs have to be developed. The Distributed Beams (DB) technique introduced in this paper provides a way to either reduce the scan times or to reduce the variance of radar-variable estimates by azimuthally spoiling the transmit beam while receiving multiple digital beams as the radar rotates in azimuth. Specifically, the rotation speed of the pedestal is derived from the duration of the coherent processing interval (CPI) to produce the desired spatial sampling. This results in beams from subsequent CPIs in approximately the same directions, which increases the number of available data samples for processing. The increased number of available samples can be coherently processed to reduce the variance of estimates. Alternatively, by reducing the number of samples per CPI and increasing the RPAR’s rotation rate, the scan time can be reduced without increasing the variance of estimates. Results presented demonstrate both applications of the DB technique for dual-polarization observations. Given that this technique makes use of spoiled transmit beams, its benefits come at the expense of degraded angular resolution (beamwidth and sidelobe levels), and reduced sensitivity compared to the use of pencil beams. The technique could be implemented as part of an RPAR concept of operations to meet requirements for the future weather surveillance network if certain tradeoffs are accounted for in the radar design process.
... In other words, the DB technique exploits the use of digital beamforming in azimuth and allows a faster rotation rate to be maintained, leading to more rapid updates potentially without degradation in data quality. Alternatively, if the rotation rate is maintained, the number of samples can be increased by a factor of R F , which leads to reduced variance of radar-variable estimates [104]. Furthermore, scanning a cluster of receive beams in both azimuth and elevation could provide a larger scan-time reduction factor, at the price of reduced sensitivity and angular resolution. ...
Thesis
Full-text available
The WSR-88D network has been operational for over 30 years and is still the primary observational instrument employed by the National Weather Service (NWS) forecasters to support their critical mission of issuing severe weather warnings and forecasts in the US. Nevertheless, the WSR-88Ds have exceeded their engineering design lifespan and technological limitations may prevent the WSR-88D to meet demanding functional requirements for future observational needs. Unique and flexible capabilities offered by Phased Array Radar (PAR) technology support the required enhanced weather surveillance strategies that are envisioned to improve the weather radar products, making PAR technology an attractive candidate for the next generation of weather radars. The single-face Rotating PAR (RPAR) architecture is more affordable than the four-faced PAR system and is capable of exceeding the functionality provided by the WSR-88D. This dissertation is focused on exploring advanced RPAR scanning techniques and concepts of operation in support of meeting future radar functional requirements. Three advanced RPAR scanning techniques are developed exploiting unique RPAR capabilities and their performance is quantified. The proposed techniques are implemented on the Advanced Technology Demonstrator (ATD), a dual-polarization RPAR system at the National Severe Storms Laboratory (NSSL) in Norman, OK. Data collection experiments are conducted with the ATD to demonstrate the performance of the proposed techniques for dual-polarization observations. Results are verified by comparing fields of radar-variable estimates produced using the proposed RPAR techniques with those produced by a well-known collocated WSR-88D radar simultaneously collecting data following an operational volume coverage pattern.
Conference Paper
Full-text available
The Advanced Technology Demonstrator (ATD) is an S-band planar polarimetric phased array radar (PPAR) at the National Severe Storms Laboratory (NSSL) in Norman, OK. It serves as a testbed for evaluating the suitability of PPAR technology for weather observations. The major challenge to PPAR weather observations is the calibration needed to produce polarimetric measurements comparable to radars using parabolic reflectors. This is due to the existence of significant cross-polar patterns, as well as the scan-dependent measurement biases inherent to PPAR. To address the former, a pulse-to-pulse phase coding technique has been proposed. However, it does not affect the scan-dependent system biases. The effects of these variations must be addressed via corrections using appropriate values at each boresight location. In this work, an overview of weather calibration for the Advanced Technology Demonstrator is presented.
Conference Paper
Full-text available
The Advanced Technology Demonstrator (ATD) is an S-band planar polarimetric phased array radar (PPPAR) being funded through a joint collaboration of the National Oceanic and Atmospheric Administration and the Federal Aviation Administration and developed by the National Severe Storms Laboratory (NSSL), MIT Lincoln Laboratory, and General Dynamics. It serves as a testbed for evaluating the suitability of PPPAR technology for weather observations. The major challenge on this path is the calibration needed to achieve PPPAR measurements comparable to those of the systems using parabolic-reflector antennas. This is due to the existence of significant cross-polar antenna patterns, as well as the scan-dependent measurement biases, inherent to PPPAR. The former induces cross coupling between returns from the horizontally and vertically oriented fields resulting in the biases of polarimetric variable estimates. Furthermore, the inductive and capacitive coupling in hardware may exacerbate the cross-coupling effects. Pulse-to-pulse phase coding in either the horizontal or vertical ports of the transmission elements has been proposed to mitigate the cross-coupling effects. However, it does not address the scan-dependent system biases in PPPAR estimates. These are caused by the horizontal (H) and vertical (V) copolar antenna patterns which vary with beamsteering direction. The effects of these variations must be addressed via corrections using appropriate values at each boresight location. If the cross-coupling effects are sufficiently suppressed with pulse-to-pulse phase coding and given sufficiently narrow antenna main beam, the corrections can be conducted using only the beam peak values of the copolar patterns. But, the cross coupling mitigation from pulse-to-pulse phase coding is inversely proportional to the cross-polar pattern levels. Thus, at boresight locations where the cross-polar pattern levels are too high, the corrections using the beam peak values of both the copolar and cross-polar patterns need to be conducted. Furthermore, the effects of active electronic components in transmit and receive paths in PPPAR systems can result in significant differences between transmit and receive patterns. For this reason, it is important to characterize the peaks of both transmit as well as receive copolar and cross-polar antenna patterns. In this work, a first look at the beam-peak-based corrections using data from ATD system is presented.
Conference Paper
Full-text available
The National Oceanic and Atmospheric Administration (NOAA) has identified evolutionary radar mission capabilities aimed at addressing current and future mission needs to support National Weather Service warning operations. Important requirements driving the next generation of weather surveillance radars include improvements in data quality, access to rapid-update volumetric data, and the ability to perform focused and tailored weather observations. Systems based on phased array antennas are promising solutions, and significant R&D investments over the last two decades have contributed to a growing understanding of this technology and to reducing risk associated with Phased Array Radar (PAR) technologies. As a major milestone in this quest, NOAA partnered with the Federal Aviation Administration to develop the Advanced Technology Demonstrator (ATD). The ATD is the first full-scale, S-band, dual-polarization, multifunction, active, electronically scanned PAR for weather observations; it is currently being tested at the National Weather Radar Testbed in Norman, Oklahoma. The ATD leverages several prior investments to provide a flexible and affordable radar system with which to evaluate performance and identify residual risk areas. The ATD will become operational in early 2020 and is expected to provide important information about PAR polarimetric performance (including calibration) and the maturity of this technology and its suitability to meet NWS mission-critical requirements. In this presentation, we will provide an overview of the planned operational capabilities of the ATD, an update on progress to date, and a roadmap for engineering and meteorological research that will help prepare NOAA for the next generation of weather-surveillance radars.
Conference Paper
Full-text available
The Advanced Technology Demonstrator (ATD) radar system is being developed through a joint collaboration of the National Oceanic and Atmospheric Administration, the Federal Aviation Administration, Lincoln Laboratory, General Dynamics Mission Systems, and the University of Oklahoma. This proof-of-concept radar system consists of an active, S-band, electronically scanned, dual-polarization phased array radar, and it makes use of pulse compression waveforms to meet sensitivity and range-resolution requirements. It is well known that the range weighting functions (RWF) associated with the use of pulse-compression waveforms exhibit range sidelobes that extend beyond the range-gate spacing, and may result in contaminated measurements from adjacent range locations. Even if a pulse-compression waveform is theoretically designed to have acceptable RWF sidelobe levels, practical effects introduced by the system may increase these sidelobe levels beyond acceptable limits. In order to meet strict range resolution requirements, practical system effects should be accounted for in the waveform design process. Typical solutions include the design of pre-distorted waveforms or mismatched filters. On one hand, waveform pre-distortion may result in better performance in terms of RWF sidelobe levels at the price of reduced sensitivity; o. On the other hand, due to the down-sampled nature of the received data, the use of a mismatched filter to compensate practical effects may not be very effective compared to pre-distorting the transmit waveform. Furthermore, system distortions may be different for the horizontal and vertical polarizations, which can impact the quality of polarimetric-variable estimates. In this paper, we characterize the distortions introduced by the ATD on pulse compression waveforms, we provide design considerations to address practical system effects, and we illustrate the performance of practical waveforms with real data. Results of this work are expected to lead to the design of pulse-compression waveforms for the ATD that are able to meet demanding range-resolution requirements by incorporating system effects in the design process.
Article
A dual-polarized X-band solid-state 1-D electronic scanning ''phase-tilt'' weather radar (PTWR) and its scanning geometry are presented. In this architecture, the true elevation angle decreases from the nominal array tilt angle and the true azimuth angle increases from the requested azimuth as one scans off boresight. Additionally, this scanning geometry induces a canting angle effect, which can be significant especially at higher elevation tilts. The predictions of potential biases in selected polarimetric variables due to this effect are derived. For elevation angles below 10°, where polarimetric measurements are of most value, predicted biases are negligible. The PTWR was deployed in Arlington, TX, USA, for an eight-week period during Spring 2014 collecting data on a number of weather events. The direct proximity (250 m away) of a mechanically scanning magnetron-based radar, employing a dual-polarized parabolic antenna allowed for a qualitative and quantitative data comparison. We find the differences in the observations made by the two radar systems are not attributable to the aforementioned biases, but appear primarily due to differences in sampling volumes of the two radars. We also find that at high-elevation tilts, the coupling of the true elevation angle with the array-relative azimuth scan angle complicates the interpretation of features at a constant altitude such as the melting layer. IEEE
Article
The survey was conducted to know the recent developments and future trends in passive, active, bipolar and monolithic microwave integrated circuitry (MMIC) phased arrays. The people's Republic of China has developed wide range of phased arrays such as, passive, active, over-the-horizon, dual band, wide-band, ultra-low sidelobe, synthetic aperture, digital-beamforming, and super-resolution. To cope with ground clutter and sidelobe jamming for airborne radar, an airborne phased array using space-time adaptive processing (STAP), is being developed. MIT Lincoln Laboratory had been developing a high performance, lower power signal processor to do digital beamforming and signal processing for a notional X-band Discover I I space-based radar.
Article
Multifunction phased array radar (MPAR) is a multiagency initiative to investigate the feasibility of replacing the aircraft surveillance and weather radar fleets in the US with a network of phased array radars based on a single, scalable architecture. The Federal Aviation Administration and the National Atmospheric and Oceanic Administration have been collaborating on MPAR risk reduction, which focuses on reducing cost, ensuring that the technology could accomplish all the missions within radar timelines, and developing dual polarization (dual pol) capability. The agencies have completed siting, cost, spectrum, dual pol, and back end studies, among others; have developed three dual pol architectures; and are building operational arrays to demonstrate that the technology can meet the basic needs of the agencies.
Conference Paper
A new Phased Array Radar (PAR) system for meteorological application has been developed by Toshiba Corporation and Osaka University under a grant of NICT, and installed in Osaka University, Japan in 2012. The phased array radar system developed has the unique capability of scanning the whole sky with 100m and 10 to 30 second resolution up to 60 km. The system adopts the digital beam forming technique for elevation scanning and mechanically rotates the array antenna in azimuth direction within 10 to 30 seconds. The radar transmits a broad beam of several degrees with 24 antenna elements and receives the back scattered signal with 128 elements digitizing at each elements. Then by digitally forming the beam in the signal processor, the fast scanning is realized. After the installation of the PAR system in Osaka University, the initial observation campaign was conducted in Osaka urban area with Ku-band Broad Band Radar (BBR) network, C-band weather radar, and lightning location system. The initial comparison with C band radar system shows that the developed PAR system can observe the behavior of the thunderstorm structure in much more detail than any other radar system. The observed high temporal resolution images of the severe thunderstorm are introduced, showing the potential capabilities of the PAR system. The correlation coefficient of the reflectivity in PAR with C band radar ranges from 0.6 to 0.9 as a function of the distance from the PAR.