Plate Coupling and Strain of the Far Western Aleutian Arc Modeled from GPS Data

Monday, 15 December 2014: 2:10 PM
Mikhail G Kogan1, Dmitry I Frolov2, Nikolay N Titkov3, Grigory M Steblov4, Nikolay F Vasilenko5, Jeffrey Todd Freymueller6, Alexandr S Prytkov5, Goran Ekstrom1 and Yury V Gabsatarov7, (1)Lamont-Doherty Earth Observatory of Columbia University, Palisades, NY, United States, (2)Ioffe Physical-Technical Institute RAS, St Petersburg, Russia, (3)Kamchatka Division GS RAS, Petropavlovsk-Kamchatsky, Russia, (4)Institute of Physics of the Earth RAS, Moscow, Russia, (5)Institute of Marine Geology and Geophysics FEB RAS, Yuzhno-Sakhalinsk, Russia, (6)University of Alaska Fairbanks, Fairbanks, AK, United States, (7)Geophys. Survey Russ. Acad. Sci., Kaluga Region, Russia
The Pacific plate (PACI) converges with the Bering plate (BERI) at the Aleutian arc, but the convergence becomes extremely oblique in the western Aleutians. The westernmost segment of the Aleutian arc near the cusp with the Kamchatka-Kuril arc is called the Komandorsky Islands (KOMA), with two islands, Bering and Medny. The strike-slip motion of PACI relative to BERI in the KOMA segment has a speed of ~80 mm/a, but the faulting of strike-slip motion is poorly understood. We analyze GPS velocities at three sites on both islands of KOMA and at two sites on the Pacific coast of Kamchatka. From several years of observations, trench-parallel velocities at all three sites on KOMA are similar: 51, 43, and 50 mm/a relative to BERI. Trench-normal velocities at both sites on Kamchatka are 14 and 21 mm/a relative to the North American plate (NOAM). The question arises: where are the faults that take up the strike-slip motion of PACI relative to BERI? We model elastic lithospheric rotation and strain of KOMA using the DEFNODE software of R. McCaffrey in three scenarios: (1) The active fault is at the Aleutian trench, (2) The active fault is in the backarc north of KOMA, (3) KOMA is a separate lithospheric block, with active faults on both sides. We set the fault locking depth at 20 km because strike-slip earthquakes in the region are not deeper. In Scenario 1, the motion of the Komandorsky Islands is explained by elastic strain at the transform plate boundary to the south. Scenario 1 requires the strike-slip fault to be shallowly dipping (12°) beneath KOMA. For a strike-slip fault nearly vertical in dip, Scenario 1 underpredicts the observed speed of KOMA and Scenario 2 overpredicts the speed. Scenario 3, with KOMA bounded by such faults on the north and on the south, predicts for KOMA a block motion of 58 mm/a, i.e., 120% of the mean observed speed. Velocities of sites on Kamchatka are well explained by elastic strain due to subduction of PACI beneath NOAM.