T44B-08
Rheology of Asthenosphere in the Kuril-Kamchatka Subduction Zone from Postseismic GPS Observations after Great 2006-2007 Earthquakes
Thursday, 17 December 2015: 17:45
104 (Moscone South)
Mikhail G Kogan1, Nikolai F Vasilenko2, Dmitry I Frolov3, Jeffrey Todd Freymueller4, Grigory M Steblov5, Goran Ekstrom6 and Alexandr S Prytkov2, (1)Columbia University of New York, Palisades, NY, United States, (2)Institute of Marine Geology and Geophysics FEB RAS, Yuzhno-Sakhalinsk, Russia, (3)Ioffe Physical-Technical Institute RAS, St Petersburg, Russia, (4)University of Alaska Fairbanks, Fairbanks, AK, United States, (5)Institute of Physics of the Earth RAS, Moscow, Russia, (6)Lamont-Doherty Earth Observatory of Columbia University, Palisades, NY, United States
Abstract:
The 1,200-km-long Kuril subduction zone is one of the most seismically active regions on the Earth. It was the last major subduction zone totally unexplored by methods of space geodesy. The Kuril GPS Array was installed in summer 2006. Several months later, a doublet of great (MW > 8) earthquakes ruptured the central segment of the Kuril arc where events of such size had not occurred for about a century. We test mechanisms of viscoelastic relaxation with both Maxwell and power-law rheologies to explain the postseismic motion for 8 years following the earthquake doublet. For the period after summer 2007, we have assumed that the contributions from afterslip are negligible. We model viscoelastic relaxation caused by coseismic slip using the open-source software package RELAX of S. Barbot. It allows us to consider the three-dimensional rheology, including a dipping elastic slab and a low-viscosity mantle above the slab. In 2007–2008, the observed postseismic movement was at a speed of several tens of millimeters per year and directed trenchward. Seven years later, the speed was reduced by an order of magnitude and still directed trenchward. Viscoelastic relaxation with the Maxwell viscosity of about 1 × 1018 Pa s reasonably explains the observed deformation for initial 3–5 years with a tendency to an increase in apparent viscosity. However, the question arises: will the apparent viscosity continue to grow with post-earthquake time? A power-law rheology would predict such growth over the decades following an earthquake. According to laboratory experiments, a power-law rheology of olivine is associated with dislocation creep (stress exponent n = 3.4–4.5) or diffusional creep (n = 0.9–1.5). We modeled deformation for a range of values of n and of apparent initial viscosity. We found the power-law models that agree well with the observed postseismic deformation, showing a significantly better fit in later years (2012–2015) than a Maxwell rheology with viscosity 1 × 1018 Pa s.
