The Establishment of the Nottinghamshire Corner Reflector Array

Document Sample
The Establishment of the Nottinghamshire Corner Reflector Array Powered By Docstoc
					The Establishment of the Nottinghamshire
  Corner Reflector Array (NCRA) for PSI
An on-going study using corner reflectors with Persistent
           Scatterer Interferometry analysis

    Sponsored by the University of Nottingham and
            The British Geological Survey

  Jon Leighton, Andrew Sowter, Doug Tragheim,
       Richard Bingley and Norman Teferle

         FRINGE 2007, 28 November 2007

• How should one set-up and optimise a corner reflector array?
   – Alignment
   – Siting
   – Network Geometry

• What use could be made of reflector networks in PSI analysis?
   – With GPS for improved geocoding
   – With concurrent GPS for atmospheric determination

• How do PSI processing methods compare?
   – TerraFirma PSIC 4 study

• How can we validate our PSI results?
   – Compare PSI to survey ‘truth’ measurements
                             Aims and

• Establish a corner reflector array for use in InSAR and PSI experiments
• Collect additional data
   – GPS at reflector sites concurrent with ENVISAT passes
   – Levelling data at sites of known motion
• Investigate and compare PSI processing methods, in particular
  treatment of atmospheric terms
   – 3-PaPS, StamPS and Gamma
• Optimally geocode images
   – Apply Small’s bistatic correction
• Analyse atmospheric terms
   – Point-wise analysis
      • Determine absolute phase using GPS data and deconstruct
      • Compare atmospheric phase with GPS Zenith Tropospheric Delay (ZTD):
         Total, Hydrostatic and Wet
   – Image-wise analysis
      • Compare atmospheric phase screen data with ZTD maps from GPS
      • Potential use of GPS ZTD in atmospheric phase screen creation
• Compare PSI analyses with ground ‘truth’ survey measurements

           • 12 reflectors
             established in May
              – 6 UoN sponsored
              – 6 BGS sponsored

           • Spread over

           • 117 ERS 1 & 2 and
             26 ENVISAT scenes
             archived, dating
             back to 1991

           • 12 ENVISAT scenes
             with reflectors
             present, since May
                    Data Collection

• Acquisition planning
  – On-going project with ESA
  – Field teams deployed for
    ENVISAT passes
  – GPS data collected 3 hours
    before and after

• Sites
  – Mixture of dynamic and
    stable ground
  – Mixture of permanent and
    non-permanent* reflector      *Figure showing reflector /
                                  triangulation pillar interface
  Sites with and
without Reflectors

 Triangulation Pillar

  Permanent Site

 West Notts College
 Triangulation Pillar
                   Data Processing

• 3 Pass Persistent Scatterer
  Processor (3-PaPS)
• Developed by Warren et al,
  at University of Nottingham
• Uses the 3 pass Integer
  Ambiguity Search (IAS)
  method (Sowter 2004)
• Good agreement with
  commercial software
  (Gamma) demonstrated over
  the Jubilee Line in London
                     Data Processing
                    StaMPS & Gamma

• Stanford Method for            • Gamma Software
  Persistent Scatterers          • Developed by Werner et al,
                                   Gamma Remote Sensing.
• Developed by Hooper et al,
  Stanford.                      • Commercial software
• Analyses spatial correlation   • Looks for points with
  of interferogram phase to        backscatter that is constant
  find PS points                   or varies temporally, and
• ‘Best’ current method for        looks at 2D spectrum of
  rapid deformations in non-       point targets
  urban areas                    • Results in huge numbers of
• Requires fewer images than       PS points
  other softwares
                    Data Processing

1. Ideally need a non-relative technique
   – No requirement for a large network of stations
2. Need to isolate tropospheric delay terms
   – For comparison with InSAR tropo terms
PPP GPS is the most convenient candidate

• Is GPS a suitable tool for testing PSI?
   – PSIC4 argues against:
   – “they both suffer from the same drawback, that is the
     propagation of an electromagnetic pulse through the atmosphere”
   – Geodesists would argue for:
   – GPS estimates ZTD as an additional unknown
   – Lots of evidence from comparing GPS ZTD (wet and water
     vapour) with microwave radiometers and radiosondes
   – GPS ZTD is being assimilated into Numerical Weather Prediction
                         Reflector Uses:

• Oversample reflector areas
  and determine positions
• Back-geocode GPS positions
  of reflectors & compare
• Check error agrees with
  predicted bistatic error
  (Small et al 2004)          1.Uncorrected image   2.Oversampled image
• Re-process
• Tests have shown
  improvements in geocoding
  of ~20m in ENVISAT ASAR
• See the Sowter et al poster
  for more info
           Visible reflector position      3.Corrected image
           Reflector Position from GPS
                           Reflector Uses:
                            Point Based

                InSAR                            GPS Troposphere
• Main contributor to                          • Zenith Wet Delay (ZWD) is
  atmospheric phase is water                     mainly due to water vapour
  vapour                                       • ZWD is obtained from
• Unwrapped phase components:                    Zenith Total Delay (ZTD)
  Φtotal= Φgeo+ Φnoise+ Φbaseline+ Φdefm+ Φatm   thus:
• Determine Φgeo from GPS data                    ZWD = ZTD – ZHD
• Assume Φnoise & Φdefm to be                     where Zenith Hydrostatic Delay
                                                  (ZHD) is obtained from met data
• Use precise orbits to minimise • ZWD are sparse (spatially)
                                                 compared to InSAR

• Convert Φatm to metres and map
  to zenith
            Are ZWD values correlated to atmospheric phase Φatm?
            Can a sparse network of GPS make an improvement?
                      Reflector Uses:
                       Image Based

                                  +                      =?

       Atmospheric Phase Screen

                              Map of Zenith Wet Delay (ZWD) as
                                  Integrated Water Vapour
                                        (BIGF 2006)

• Can GPS-ZWD/IWV maps add any value to APS?
• If not, what spatial density of observations is required?
                            Levelling Data
                               and PSI

   Nottingham City Centre              Rolls Royce Facility, Hucknall Airfield
Processed from 59 ERS1 and 2 SLC IMS            Clear deformation signal
    Mar 95 – Dec 00 using Gamma                   Of up to 1-2cm / yr

• Area has since shown uplift on more recent ENVISAT data
• Episodic precise levelling is underway along with PSI processing
  using three separate methods and up to date ASAR data
• Talks underway with Rolls Royce to establish a reflector array on site
                      Summary &

• Once sufficient data has been acquired, the NCRA
  should provide a useful tool for validating and
  assessing InSAR and PSI studys.
• The value of any PSI analysis should increase if
  multiple methods are used and a comparison with
  ground truth data is performed.
• Use of GPS at reflector sites may improve
  atmospheric estimation in InSAR and PSI

  Thanks to the British Geological Survey for supporting the
  research, the various software contributors and ESA for the
  Cat-1 project.

Shared By: