Field and theoretical investigations of strain localization: Effects of mineralogy, shear heating and grain size evolution on deformation in the Earth

Viscous and viscoelastic deformation strongly affects the mechanical behavior of the Earth. This style of deformation has consequences for a wide range of geodynamic processes from large scale processes like the formation and maintenance of plate boundaries, to smaller scale processes like postseismic deformation on and near faults. One of the key features of viscous and viscoelastic deformation in the Earth is that it is observed to be self localizing under some circumstances. This is in spite of the tendency for viscous deformation to be pervasive in a deforming system. Many processes are thought to contribute to strain localization in the Earth: (1) viscous dissipation or shear heating (e.g., Braeck and Podladchikov, 2007; Braeck et al., 2009; Kameyama et al., 1999; Kelemen and Hirth, 2007; Ogawa, 1987), (2) grain size reduction (e.g., Braun et al., 1999; Montési and Hirth, 2003; Précigout and Gueydan, 2009), (3) lattice preferred orientation development (LPO) (e.g., Poirier, 1980; Tomassi et al., 2009), mixing of phases (e.g., Skemer et al., 2010a; Toy et al., 2010; Warren and Hirth, 2006) and geometrical interconnection of weak phases and materials (e.g., Handy, 1994). Utilizing both natural samples from Oman (Chapter 2) and theoretical work based on numerical modeling (Chapters 3 and 4) each chapter of this thesis evaluates the effect of a different one of these processes on strain localization, and in the case of Chapter 4 evaluates the additional feedback between two of these processes. In Chapter 2 we examine strain localization in a natural system in which two very rheologically different materials, gabbronorite (predominantly plagioclase) and harzburgite (predominantly olivine), were juxtaposed due to volcanic intrusion and subsequently deformed. We utilized field relationships, pyroxene and amphibole/plagioclase thermometry, metamorphic phase equilibrium, grain size piezometry and electron backscatter diffraction (EBSD) in order to constrain the deformation conditions for the field area. The viscosity of gabbronorite was found to be: (1) consistent with the predicted viscosities based on the extrapolation of experimental flow laws and (2) at least two orders of magnitude lower than the harzburgite while deformation was occurring. This suggests both that a significant viscosity contrast exists at the crust-mantle boundary where the crustal lithology is dominated by plagioclase and the mantle by olivine, and wherever deformation is geometrically allowed to localize within plagioclase rich layers. In Chapter 3 we examine the theoretical effect of shear heating as well as the feedback between viscous dissipation and temperature dependant viscosity on strain localization in a one-dimensional model of a viscoelastic shear zone. This model builds on the work of Kelemen and Hirth (2007) by utilizing a complex dry olivine viscoelastic rheology that includes dislocation creep, diffusion creep, dislocation accommodated grain boundary sliding (disGBS) and low temperature plasticity (LTP). We have found that increasing either the applied strain rate or the grain size system behavior is modified in three significant ways: (1) it causes the maximum stress the system can archive to increase, (2) it results in more unstable system behavior and (3) it causes the system to accommodate more deformation in the background. One consequence of enhanced background deformation is that system exhibits distinct periods of accelerated stress relaxation accompanied by increased strain rates, that do not necessarily go unstable. Consequently, we have shown that shear heating may play an important roll both in viscous deformation in the Earth and potentially in the occurrence of intermediate depth earthquakes and slow slip events. In Chapter 4, we extend Chapter 3 and examine the feedbacks between grain size evolution, viscous dissipation and a complex temperature and grain size dependant viscosity in a one-dimensional model of a viscoelastic shear zone. We evaluated both the grain size evolution models of Austin and Evans (2007) and a modified version of Hall and Parmentier (2003). We find that Austin and Evans predicts unrealistically fine background grain sizes while the predictions based on Hall and Parmentier (2003) are more reasonable. We also find that, based on this model and the experimental work of Mei et al. (2010), LTP may not contribute to grain size reduction in viscously deforming materials. Based on this model grain size evolution does not appear to strongly affect the peak stress or stability of a system for fine initial grain sizes as grain size reduction does not significantly alter the initial viscosity structure. However, in systems with coarser initial grain sizes, grain size evolution does appear to contribute to system instability. Additionally, for both initially coarse and fine systems, grains size evolution results in the emergence of stress evolutions displaying two distinct episodes of stress reduction. Much like Chapter 3, our observations in Chapter 4 suggest that grain size evolution may play an important role in viscous deformation in the Earth and may potentially be a mechanism for some intermediate depth earthquakes and slow slip events. Taken together the chapters in this thesis explore several of the potentially important processes that affect strain localization in the Earth. Thus providing significant insight into this important phenomenon