Subsidence Monitoring in Imperial Valley, California, Using Satellite Radar Interferometry

1
Subsidence Monitoring in Imperial Valley,
California, Using Satellite Radar Interferometry

• Mariana Eneva (Imageair, Inc.)
• Piyush Shanker (Stanford University)

2007 GRC, Sep 30 Oct 3
2
Objective
Monitor surface deformation that may result from
increased geothermal production in Imperial Valley
1. Salton Sea, Heber, East Mesa
2. InSAR cost-effective, covers large areas
3. Sub-cm surface deformation, mm precision
3
Study Region - North
ASTER VNIR Green Production Areas White
KGRAs
4
Study Region - South
ASTER VNIR Green Production Areas White
KGRAs
5
M5.1 - 09/02/2005
6
Why?

• Present installed capacity 538 MW
• Salton Sea (CalEnergy) 350MW
• Heber (Ormat) 115 MW
• East Mesa (Ormat) 73 MW
• Eleven geothermal fields with development
  potential
• Combined potential 2560 MW
• More than 2,000 MW additional capacity expected
  over next decade
• Possibly larger environmental impact
• Increased need for monitoring

7
InSAR

• Successfully used to detect surface deformation
  earthquakes, mining, water pumping, geothermal
• Not yet in routine practice
• Cost-effective coverage of large areas
• Can supply feedback planning and mitigation
• Either detecting surface deformation or asserting
  its absence are important

8
InSAR Advantages

• Sub-cm deformation, mm-precision
• Day-and-night capability
• All-weather
• Semi-continuous spatial coverage of large areas
  (100 km x 100 km)
• Frequent monitoring (e.g., every 35 or 24 days)
• Satellite InSAR cost-effective

9
Platforms/Wavelengths/Resolution

• Satellites spatial resolution 20-30 m
• ERS-1, ERS-2, ENVISAT (C-band ?5.6 cm)
• JERS-1, ALOS (L-band ?23.5 cm)
• RADRSAT-1, RADARSAT-2 (C-band ?5.6 cm)
• C-band dry non-vegetated areas
• L-band can penetrate vegetation
• Airborne 10 cm at 25 km, 30 cm at 55 km
• X-band (?3 cm), Ku-band (?2 cm)
• Future
• L-band on UAV

10
InSAR some definitions

• SAR (synthetic aperture radar)
• Aperturesize of antenna
• Using motion to synthesize larger antenna (10 m
  vs. 4-km!)
• InSAR
• Interferogram from two images at two different
  times (for topography)
• Differential InSAR (DInSAR)
• Comparing two interferograms (for surface
  deformation)
• Permanent Scatterers InSAR (PS InSAR)
• Derived from DInSAR, but using individual points
  best in vegetated areas where classic DInSAR
  may not work

11
DInSAR/PS InSAR

• Surface displacement between two passes
• Phase difference

?f/?? 4p/? where ?f phase change ??
displacement ? wavelength (NO dependence on
altitude!)
12
DInSAR Example
Works in dry, non-vegetated areas
Under good conditions two interferograms can
produce results
South African M5.1 rockburst (1999)
13
PS InSAR example
Can work in vegetated areas Time
series of interferograms are needed
14
Permanent Scatterers

• Buildings, roads, big rocks, etc.
• Satellite InSAR corner reflectors
• (CRInSAR, PSInSAR)
• Airborne InSAR
• e.g., 1 foot x 1 foot boxes with gravel

15
Examples of subsidence in geothermal fields
detected with InSAR

• East Mesa, CA 3.7 cm/yr
• Massonnet et al. (1997)
• Coso, CA up to 3.5 cm/yr
• Fialko and Simons (2000), Wicks et al. (2001)
• Bradys Hot Srings, NV 1.3 cm/yr
• Oppliger and Coolbaugh (see Resource
  Characterization II session, today, 1120 a.m.)
• Cerro Prieto (Mexico) up to 16 cm/yr
• Carnec and Fabriol (1999), Hanssen (2001)
• Wairakei and Tauhara (New Zealand) 7.8 cm/yr
• Chang et al. (2005)
• Euganean field (Italy) 0.4 cm/yr
• Strozzi et al. (1999)

16
DInSAR Example from Coso
From Wicks et al. (2001). Change between June
1992 and September 1996
17
Why all this has not been already done in
Imperial Valley?
BEFORE
NOW

• there has not been
• much subsidence
• C-band does not work well in vegetated areas

• this may change with increased production
• L-band available
• permanent scatterers

18
SAR Data

• ERS (C-band) 1993 - 2000
• 20 interferograms - 7.5 years
• ENVISAT (C-band 5.6 cm) since 2003
• 15 interferograms - 3.8 years
• Currently getting results for Salton Sea and
  Heber
• Comparing with leveling measurements
• Intend to use also
• RADARSAT-1 (C-band)
• ALOS (L-band 23.5 cm)

19
SAR footprint in Imperial Valley
20
ERS PS InSAR time series of interferograms
21
ERS PS InSAR deformation rates
- movement towards satellite
movement away from satellite (subsidence)
Displacement, mm
22
PS InSAR in Salton Sea
79 benchmarks 200 ERS PS 450 Envisat PS
23
Subsidence rate 1
Rate 3 mm/yr
24
Subsidence rate 2
Rate 4 mm/yr
25
Subsidence rate 3
Rate 5.4 mm/yr
26
Comparison with leveling data
27
Summary

• InSAR has high potential to provide
  cost-effective monitoring of environmental impact
  (surface deformation)
• Assure compliance with regulations
• Possibly identify measures to mitigate adverse
  impact and reduce cost of mitigation efforts
• Mitigate secondary effects (e.g., agriculture)
• Good public relations/environmentally friendly
  operations

28
Further work

• Detailed comparison of ERS and Envisat PS InSAR
  results with leveling data
• Salton Sea
• Heber
• Analysis of data from Radarsat (C-band/ PS InSAR)
  and ALOS (L-band/DInSAR)

29
Acknowledgements

• California Energy Commission
• CalEnergy, Inc. - Brian Berard
• ORMAT Nevada, Inc.- Charlene Wardlow, Erik Osbun
• Geo Hills Associates Jim Combs
• Stanford University Howard Zebker