UAVSAR and

1
UAVSAR and Volcano Monitoring
Scott Hensley, Howard Zebker, Cathleen Jones,
Paul Lundgren, Eric Fielding, Thierry Michel
and Bruce Chapman
USEReST Naples, Italy November 11-14, 2008
2
UAVSAR

• UAVSAR is an L-band fully polarimetric SAR
  employing an electronically scanned antenna that
  has been designed to support a wide range of
  science investigations.
• Science investigations supported by UAVSAR
  include solid earth, cryospheric studies,
  vegetation mapping and land use classification,
  archeological research, soil moisture mapping,
  geology and cold land processes.
• To support science applications requiring repeat
  pass observations such as solid earth and
  vegetation applications the UAVSAR design
  incorporates
• A precision autopilot developed by NASA Dryden
  that allows the platform to fly repeat
  trajectories that are mostly within a 5 m tube.
• Compensates for attitude angle changes during and
  between repeat tracks by electronically pointing
  the antenna based on attitude angle changes
  measured by the INU.
• UAVSAR is testing new experimental modes, e.g.
  the multi-squint mode whereby data is collected
  simultaneously at multiple squint angles to
  enable vector deformation measurements with a
  single repeat pass.

3
Need for Airborne RPI Instrument

• Spaceborne repeat pass radar interferometry
  derived deformation measurements has become a
  standard tool for the solid earth science and
  glaciological science communities.
• Repeat times controlled by the the repeat orbit
  cycle of spaceborne SAR systems, e.g. ERS-1,2 (35
  days), Radarsat (24 days), JERS (44 days), and
  Envisat ( 35 days).
• Rapidly deforming features such as some volcanoes
  and glaciers or deformation from post seismic
  rebound require repeat times of a day or less to
  fully study the time varying nature of the
  deformation signal.
• Repeat pass airborne interferometric measurements
  are more difficult to make because
• The difficulty of flying a specified trajectory
  with the required accuracy
• To need to compensate for pointing changes
  between flight tracks

4
UAVSAR and Volcano Observations

• Although a number of the worlds active volcanoes
  have continuous seismic and/or GPS monitoring
  there is still a large number that are not
  persistently monitored.
• Placing in situ monitoring devices on volcanoes
  after they enter more active states can be
  dangerous and usually only a few such devices can
  be deployed leading to an under sampling of the
  resulting deformation.
• UAVSAR is ideally suited to making repeat pass
  observations of volcanic regions
• It has a large swath in excess of 20 km with
  fully polarimetric observations and flies at high
  altitudes (gt 12.5 km) with a resolution of 1.6 m
  in range and 1m in azimuth.
• Operates at L-band to reduce temporal
  decorrelation.
• It can be tasked to make repeat observations on
  time scales as short as 20 minutes from any
  desired look direction.
• It can control its flight path to be within a 10
  m tube (usually within 5 m) and adjust its look
  direction electronically to compensate for
  aircraft attitude changes.
• Has a vector deformation capability whereby
  vector deformation (and atmospheric distortions -
  under investigation) can be simultaneously
  determined.
• Can be rapidly deployed to monitor evolving
  volcano hazards or routinely tasked to monitor
  more quiescent volcanoes.

5
Initial Flight Testing of UAVSAR
6
Antenna Overview

• 24 T/R Modules
• 3 RF Manifold Boards ASN
• 6 power only ESSs
• Differential low voltage daisy
• chain signaling
• TRAC

7
Key Radar Instrument Parameters
Parameter Value
Frequency L-Band 1217.5 to 1297.5 MHz
Bandwidth 80 MHz
Resolution 1.67 m Range, 0.8 m Azimuth
Polarization Full Quad-Polarization
ADC Bits 2,4,6,8,10 12 bit selectable BFPQ, 180Mhz
Waveform Nominal Chirp/Arbitrary Waveform
Antenna Aperture 0.5 m range/1.5 azimuth (electrical)
Azimuth Steering Greater than 20 (45 goal)
Transmit Power gt 3.1 kW
Polarization Isolation lt-25 dB (lt-30 dB goal)
8
UAVSAR Modes
Polarimetric SAR
Strip Mode SAR
Multi-Squint Vector Deformation
CoPol Monopulse
9
Example Repeat Pass Baselines
San Andreas Fault Repeat-Pass Baseline 80 km
Datatakes on February 12 and 20 of 2008.
10 m Tube
10 m Tube
5 m Tube
5 m Tube
10 m Tube
5 m Tube
10
Repeat Pass Processing Challenges

• Repeat pass processing of airborne data is very
  challenging for several reasons
• First, the onboard motion metrology (INU and
  DGPS) are not sufficient to support automated
  processing of the data. Best case the INUGPS
  combination provide 3 cm antenna phase center
  flight path reconstruction, whereas the needed
  accuracy is 1 mm or less. This means residual
  motion between passes must be solved for from the
  data themselves.
• The motion compensation algorithm, which corrects
  the SAR data from an irregular flight path to a
  reference trajectory, is terrain dependent. This
  imposes additional processing complexity and
  reduces throughput.
• Effective phase center changes resulting from
  electronically steering the antenna must be
  compensated on a pulse-by-pulse basis to avoid
  phase discontinuities in the differential
  interferograms.
• Repeat pass processing presently involves a large
  amount of touch labor particularly with respect
  to the residual motion estimation. This problem
  is complicated due to
• Temporal decorrelation results in loss of signal
  which impedes the ability to estimate residual
  motion.
• Deformation signals in the cross line-of-sight
  direction couple with residual motion errors
  hence making estimation of residual motion in the
  region where deformation is occurring extremely
  difficult.

11
Mt St Helens Photo
12
Mt St Helens - UAVSAR March 24, 2008
Flight Direction
gt 20 km
Fully polarimetric image of Mt St Helens
collected on March 24, 2008 by the UAVSAR radar.
A second acquisition was collected on March 31,
2008.
13
Expanded View of Caldera and Dome
AIRSAR December 2004
14
Mt St Helens Interferogram - 4 hour Repeat
This is a first cut interferogram - no
offset measurements - no motion correction
- no topography correction
Since time between observations is 4.2 hours or
.174 days, the estimated rate of motion for an
approximate ? radians of phase change is
Phase (rad)
15
Mt St Helens - UAVSAR March 24, 2008
Flight Direction
gt 20 km
Two fully polarimetric image of Mt St Helens
collected on March 24, 2008 by the UAVSAR radar
separated by 4 hours. A second acquisition was
collected on March 31, 2008.
Deformation of 37 cm/day and 15 cm/day observed
on the two tongues of the glaciers in the
caldera. Refined processing underway to look
for deformation on dome.
16
Multi-Squint Mode Example
Data collected in the UAVSAR multi-squint mode.
Yaw angle of -5.0 with azimuth steering angles
of 13.6 and -6.2. Anthropogenic features
exhibit strong viewing angle scattering
signatures.
Azimuth Steering Angle 13.6
-5
-5
13.6
-6.2
Azimuth Steering Angle -6.2
17
UAVSAR Status

• UAVSAR is an actively scanned fully polarimetric
  L-band SAR designed to meet a range of remote
  sensing applications of interest to the science
  community including specific features to support
  repeat pass interferometry for deformation
  studies.
• UAVSAR platform is completing modifications to
  extend its range and increase the number of
  airports it can be deployed.
• The system is expected to be available to the
  general science community starting in late 2008
  with a number of investigators already funded for
  data collections.