Concept and Design for a Pilot Demonstration

1
Concept and Design for a Pilot Demonstration
Ground-based Remote Icing Detection System
Roger. F. Reinking, Robert A. Kropfli,
Sergey Y. Matrosov, W.Carroll Campbell, M.J.
Post, Duane A. Hazen, Janet S. Gibson,
Kenneth P. Moran, and Brooks E. Martner
NOAA / OAR / Environmental Technology
Laboratory, Boulder, Colorado, USA
CIRES/NOAA-ETL
Hydrometeors depolarize a radar signal
according to their shape and orientation,
and the polarization state of the
transmission. Ice particles Non-spherical,
oriented. Droplets
Spherical. Depolarizations caused by
hydrometeors can be linked with identity by
measuring the magnitude of DR and its dependence
on antenna elevation angle, when a polarization
state is chosen that is not sensitive to
variability of particle orientation. To detect
and uniquely identify SLD, the depolarization
must be measurable in low-reflectivity clouds (5
- -20 dBZ).

• Objective Develop an operational-
• grade remote sensing technology that
• will detect in-flight aircraft icing hazards.
• For more than a decade, NOAA/ETL has participated
  in an FAA Aviation
• Weather Research Program to develop icing
  diagnostic remote sensors. Our most
• important on-going task has been to develop a
  dual-polarization Ka-band (8.66 mm)
• radar that will detect clouds of hazardous
  supercooled large droplets (SLD, 50-500-
• microns) and distinguish them from clouds with
  ice particles. The team has
• succeeded.
• In tests including the 1999 Mt. Washington Icing
  Sensors Project (MWISP), ETL
• was able to demonstrate a radar capability for
  deterministic hydrometeor identification.
• By measuring one parameter, a depolarization
  ratio, DR, the technology was shown
• to be capable of distinguishing among the
  regular types of ice crystals and even the
• more spherical and irregular ice particles, and
  to differentiate all of these from clouds
• of SLD. The state of the transmitted polarization
  was key. Proof of concept was
• provided in the excellent agreement between
  scattering theory and field

Ground-Based Remote Icing Detection System Faa
Icing Remote Sensors Testbed (GRIDS /
FIRST)
Slant-45 quasi-linear Polarization A very good,
field-tested selection
(quasi- slightly elliptical)
Theory
Unattended Continuous Operation Automated warning
s
Under Development
Min. sensitivity to crystal canting Very good
separation by shape Droplets vs. ice differ
through wide arc of antenna elevation Determin
istic droplet I.D. above antenna
cross-talk Sensitive to lower reflectivity
clouds (Not true of horizontal linear
polarization)
planar crystals
std. dev. of cant 3 deg
columnar crystals
(droplets)
Radar Ka-band (8.7 mm) Doppler
Dual-polarization Dual-channel Microwave
radiometer
planar crystals
std. dev. of cant 15 deg
columnar crystals
(droplets)

But a (quasi-)circular polarization will be
slightly better (see the GRIDS Design.).
The GRIDS Design
The Algorithm
GRIDS Product A continuously-updated profile of
the potential icing hazard
Theory and field measurements match, and new
theory shows that a circular or near-circular
polarization will differentiate particles by 1-3
dB more than slant-linear. GRIDS will measure
CDR. A fixed-beam measurement at an elevation
angle near 40 degrees will accomplish the
necessary differentiation of SLD from ice
particles while minimizing path length and
propagation effects.
GRIDS will include the most sensitive K-band
cloud radar ever built. 90 of icing events
occur at 0C lt T lt -20C, below 6 km MSL, or
within 10 km range at the selected 40-deg ant.
elev. (An alternating zenith-pointing option
will add a 2nd DR point for ice particle I.D. and
vertical velocity to show particle sorting and
LW-producing convection). SLD clouds range down
to about 20 dBZ, where the cross-polar
reflectivity is about 50 dBZ. So GRIDS is
required to have a main-channel sensitivity of
55 to 60 dBZ at 10 km Range, which will be
achieved with a large antenna (3 m) long dwell
time (60 s) and long pulse width(1.0 1.5
micro- sec). A 150-m range resolution will
isolate hazardous cloud layers.
The simplest GRIDS algorithm uses four decision
points based on the real-time slant-path
measurements of column-liquid (LW), cloud
reflectivity (Ze), and depolarization ratio
(DR), plus the temperature profile ingested
hourly from the rapid update cycle (RUC) or
other operational model. A potentially hazardous
cloud is identified as one that exhibits (1)
Measurable LW, (2) T lt 0C, (3) A reflectivity
Ze gt -20 dBZ, and (4) a depolarization ratio DR
equaling the minimum, spherical hydro-
meteor signature (/- 2dB), so indicating
droplets, not ice crystals.
Clouds with ice crystals aloft, melting layer,
warm drizzle below
DR
LW gt 0, T gt 0C
DRgt-28 dB at Tlt0C
Ze
Ze gt -20 dBZ
SLDR-45 vs. CDR
To the web
The 2-channel mailbox microwave radiometer will
be tilted to match the elevation of the radar
antenna to provide continuous, independent
verification of icing potential from the measured
presence or absence of LW and the quantity, which
can be allocated to the radar-observed cloud
layer(s).
This is the basis for a more sophisticated
algorithm. The quantity of liquid from the
radiometer will be assigned within the measured
vertical boundaries of the cloud layers (if
totally liquid clouds, to the layers of
minimum-DR droplets and if mixed phase,
distributed through the depths of the existing
layers). A scaled icing-hazard warning will be
added (red for probable, yellow for caution,
green for no hazard). Detection and use of the
uniquely high DR of any bright band will confirm
the altitude of the 0C boundary, and indicate
rain below. Appropriate combinations of
measured parameters will provide data quality
control.
Hazardous supercooled large droplets
DRlt-28dB at Tlt0C
DR
LW gt 0, T lt 0C
RADS, the systems integrator, will control the
instruments, ingest external data, process
data, compute products, and transmit the
display of determined potential icing hazards to
the users.
Difference in CDR (dB), ice particles (aspect
ratio lt 1) vs. SLD (a.r. 1) at ant. elev. of
40 deg.
Acknowledgments. The Federal Aviation
Administration funded this research. The views
are those of the authors and do not necessarily
represent the official policy or position of the
FAA. A description of GRIDS is published in the
Proceedings of the International Conference on
Radar Meteorology, Munich, 2001.
Ze
Ze gt -20 dBZ