SAR interferometry is widely used for dense measurements of surface displacements caused by earthquakes, but the method cannot be applied if displacements are too large. The near-epicentral area of the Iwate–Miyagi Nairiku Earthquake in 2008 is an inapplicable case. Therefore, we applied photogrammetry to measure surface displacement caused by the earthquake. The maximum horizontal and vertical displacements were found to be 5.3 m and 2.9 m, respectively.
We recognized three lines where displacement changes abruptly. The displacement distribution is like that of a reverse fault along the first line (A–B), an east-rising fault along the second line (F–G–H; west of line A–B), and a left-lateral fault along the third line (B–C; between line A–B and line F–G–H).
The earthquake source fault reaches or approaches the ground surface at line A–B, with slippage decreasing toward the ground surface. The fault-like large surface deformation found north of the Aratozawa Dam is on the first line. The fault-like deformation was caused by the motion of the earthquake source fault, and the relative displacement of the fault-like deformation was enlarged by local causes. A gravitational mass movement found north of the fault-like deformation is one cause.
Because the width of the rising area is small, only 3.5 km, at the southwestern side of line B–C, the slip of the earthquake source fault is mainly distributed near the ground. Line F–G–H suggests the existence of a geological structure that causes the abrupt changes of vertical displacement without a horizontal displacement, for example a high-angle fault.
We assumed: (1) the slip on the main fault is distributed only in a shallow area at the southern part of the main fault and only in a deep area at the northern part; and, (2) the difference of slip caused two lateral faults between southern and northern parts. The assumption qualitatively explains many observation results, such as why there is an abrupt change of horizontal displacement along line B–C and why line F–G–H has a convex part to the east.
We found a correlation between the occurrence of large landslides and abrupt changes of displacement, in other words large surface strain. The following mechanisms are possible causes of the correlation: (1) stress from surface strain increased large landslides; (2) faults (not only the main fault) may exist under the focused areas, rupture of faults caused both large surface strain and large seismic motion, and seismic motion induced large land slides. We also found that landslides and slope failures occurred densely over the slipping area on the main fault, based on the assumptions in the previous paragraph.
Because photogrammetric measurements need interactive observations, we could avoid observations on possible embanking areas. Because photogrammetry allows intensive measurements at interesting areas, we revealed a two flexure-like distribution of vertical displacement. Therefore, photogrammetry is an effective method for measuring surface displacement caused by an earthquake.