Fault creep:GPS and InSAR opportunities

Eric LindseyScripps Institution of Oceanography

−117˚ −116˚ −115˚

33˚ 33˚

34˚ 34˚T356

T77

• T356 + T77 processed in ROIPAC and GMTSAR(> 150 iʼgrams/track)

• Compute RMS noise for each date, stack best pairs (~30 iʼgrams/track)

• Assume a combination of fault-parallel and vertical motion only:

• Ignores time dependence, fault-normal motion

BRIEF ARTICLE

THE AUTHOR

v1 v2 (e, n, u)�

v1

v2

�= P

�vf

vz

�, P =

�e1 sin θ + n1 cos θ u1

e2 sin θ + n2 cos θ u2

�

�vf

vz

�= P−1

�v1

v2

�

1

BRIEF ARTICLE

THE AUTHOR

v1 v2 (e, n, u) α�

v1

v2

�= P

�vf

vz

�, P =

�e1 sin α + n1 cos α u1

e2 sin α + n2 cos α u2

�

�vf

vz

�= P−1

�v1

v2

�

µ0 µ1 b ∆�̇

�xy =σ0µ0

�1 + a

b

�

1 + µ1µ0

ab

µ0

µ1=

(a/b)∆�̇

1 + a/b−∆�̇

1

Envisat 2003-2010

Reference

Envisat 2003-2010

LOSLOS

T77 Ascending T356 Descending

Reference

Fault-parallel Vertical

Envisat 2003-2010

Re-occupied UCSB level line 2007-2012

28 sites within 1km

GPS: Creep rate 3.2±1 mm/yr

GPS validation

10

12

14

16

18

20

Faul

tPa

ralle

l rat

e (m

m/y

r)

1.5 1.0 0.5 0.0 0.5 1.0 1.5Across fault distance (km)

Envisat 2003 2010GPS 2007 2012

SABR

PAIN

Creep meters

14

16

18

Faul

tPa

ralle

l (m

m/y

r)

2 1 0 1 2 3 4 5Across fault distance (km)

North Shore <0.1 mm/yr

12

14

16

18

Faul

tPa

ralle

l (m

m/y

r)

2 1 0 1 2 3 4 5Across fault distance (km)

Ferrum 1.9 3 mm/yr

12

14

16

18

Faul

tPa

ralle

l (m

m/y

r)

2 1 0 1 2 3 4 5Across fault distance (km)

Salt Creek 1.5 3 mm/yr

12

14

16

18Fa

ult

Para

llel (

mm

/yr)

2 1 0 1 2 3 4 5Across fault distance (km)

Durmid Hill 1.8 2.9 mm/yr

R. Bilham, cires.colorado.edu/~bilham/

Future directions

• Spatial pattern is well resolved by InSAR

• Where do we need more look directions?

• What spatial resolution / filtering is appropriate?

• Time dependence of creep?

10

12

14

16

18

20

Faul

tPa

ralle

l (m

m/y

r)

1500 1000 500 0 500 1000 1500Distance W E (m)

GPS 2007 2012GMTSAR, 100m filterROIPAC, 200m filter

4 looks (boxcar filter), geocoded stacking100m gaussian filter, stacking in radar coords.