TY - JOUR
T1 - Coseismic surface deformation from air photos
T2 - The Kickapoo step over in the 1992 Landers rupture
AU - Michel, Rémi
AU - Avouac, Jean Philippe
PY - 2006/3/4
Y1 - 2006/3/4
N2 - Coseismic deformation of the ground can be measured from aerial views taken before and after an earthquake. We chose the area of the Kickapoo-Landers step over along the 1992 Landers earthquake zone, using air photos (scale 1:40,000) scanned at 0.4 m resolution. Two photos acquired after the earthquake are used to assess the accuracy and to evaluate various sources of noise. Optical distortions, film deformation, scanning errors, or errors in viewing parameters can yield metric bias at wavelength larger than 1 km. Offset field at shorter wavelength is more reliable and mainly affected by temporal decorrelation of the images induced by changes in radiometry with time. Temmporal decorrelation and resulting uncertainty on offsets are estimated locally from the correlation degree between the images. Relative surface displacements are measured independently every about 15 m and with uncertainty typically below 10 cm (RMS). The offset field reveals most of the surface ruptures mapped in the field. The fault slip is accurate to about 7 cm (RMS) and measured independently every 200 m from stacked profiles. Slip distribution compares well with field measurements at the kilometric scale but reveals local discrepancies suggesting that deformation is generally, although not systematically, localized on the major fault zone located in the field. This type of data can provide useful insight into the fault zone's mechanical properties. Our measurements indicate that elastic coseismic strain near the fault zone can be as large as 0.5 × 10-3, while anelastic yielding was attained for strain in excess of about 1-2 × 10-3.
AB - Coseismic deformation of the ground can be measured from aerial views taken before and after an earthquake. We chose the area of the Kickapoo-Landers step over along the 1992 Landers earthquake zone, using air photos (scale 1:40,000) scanned at 0.4 m resolution. Two photos acquired after the earthquake are used to assess the accuracy and to evaluate various sources of noise. Optical distortions, film deformation, scanning errors, or errors in viewing parameters can yield metric bias at wavelength larger than 1 km. Offset field at shorter wavelength is more reliable and mainly affected by temporal decorrelation of the images induced by changes in radiometry with time. Temmporal decorrelation and resulting uncertainty on offsets are estimated locally from the correlation degree between the images. Relative surface displacements are measured independently every about 15 m and with uncertainty typically below 10 cm (RMS). The offset field reveals most of the surface ruptures mapped in the field. The fault slip is accurate to about 7 cm (RMS) and measured independently every 200 m from stacked profiles. Slip distribution compares well with field measurements at the kilometric scale but reveals local discrepancies suggesting that deformation is generally, although not systematically, localized on the major fault zone located in the field. This type of data can provide useful insight into the fault zone's mechanical properties. Our measurements indicate that elastic coseismic strain near the fault zone can be as large as 0.5 × 10-3, while anelastic yielding was attained for strain in excess of about 1-2 × 10-3.
UR - http://www.scopus.com/inward/record.url?scp=33646366140&partnerID=8YFLogxK
U2 - 10.1029/2005JB003776
DO - 10.1029/2005JB003776
M3 - Article
AN - SCOPUS:33646366140
SN - 2169-9313
VL - 111
JO - Journal of Geophysical Research: Solid Earth
JF - Journal of Geophysical Research: Solid Earth
IS - 3
M1 - B03408
ER -