The Dependence of Surface Deformation on Rheology Throughout the Seismic Cycle

岩石圈动力学中的一个基本问题是，下地壳是比上地幔更强还是更弱。这个问题不仅对于了解岩石圈的物理状态是重要的，而且对于能够描述主要断裂上的地震周期(即在数百到数千年的时间内断裂上的应力反复积累，然后在地震中最终释放应力)也是至关重要的。地震周期的观测主要包括对地震周期中地球表面有时的变形的测量。为了将这些地表形变观测转化为孕震地壳中的应力，我们需要准确描述岩石圈的力学性质(即流变性)的力学模型。下地壳比上地幔更强还是更弱，对地震周期中断层上的应力演化有很大的影响。由于地震周期比现代测量地球形变的时代要长得多，我们没有整个地震周期的完整观测记录。如果没有完整的观测，直接从观测中唯一验证力学模型并不总是可能的。此外，具有显著流变复杂性的地震周期模型通常计算昂贵，因此不适合探索与观测一致的所有允许的模型参数范围。地震周期的力学模型通常只包含Maxwell粘弹性，而不考虑材料性质的稳健深度相关性或下地壳或地幔内的局部剪切带。然而，岩石圈的流变性可能比麦克斯韦粘弹性更复杂，岩石圈的材料属性预计与深度有关。在整个地震周期中，可能对岩石圈变形有重大影响的流变复杂性包括深度相关的粘度、幂定律蠕变、Burgers粘弹性和深部的局部蠕变。在这个项目中，我们将：(1)量化整个地震周期中地表变形对不同深度流变学的敏感性。(2)确定Burgers粘弹性和幂定律蠕变在地震循环模型中是否具有相同的影响。(3)系统地测试由深度局部或分布蠕变引起的地震间地表变形的相似性。(4)建立理想化模型中的地表变形与深度依赖粘性、幂定律蠕变、瞬时粘弹性或局部蠕变模型中的变形之间的时间依赖关系。最后一点将有助于理解从地表形变观测中验证地震周期模型所固有的非独特性。此外，通过知道理想化模型和更复杂模型的类别之间的对应关系，人们将能够将计算效率高的理想化模型约束到大地测量数据，然后直接从理想化模型推论中确定更复杂模型的范围。

A fundamental question in lithospheric dynamics is if the lower crust is stronger or weaker than the upper-most mantle. This question is not only of importance to understanding the physical state of the lithosphere, but is also essential to being able to describe the earthquake cycle on major faults (i.e., the repeated cycle of stress build-up on faults during a period of hundreds to thousands of years, and then the eventual release of that stress during an earthquake).Observations of the earthquake cycle are largely composed of measurements of the deformation of Earth?s surface at times during the earthquake cycle. In order to translate those observations of surface deformation to stresses in the seismogenic crust, we require mechanical models that accurately describe the mechanical properties (i.e., rheologies) of the lithosphere. Whether the lower crust is stronger or weaker than the upper-most mantle has strong implications for the evolution of stresses on faults during the earthquake cycle. As the earthquake cycle is significantly longer than the era of modern measurements of Earth deformation, we do not have a complete record of observations throughout an earthquake cycle. Without complete observations, it is not always possible to uniquely validate mechanical models directly from observations. Furthermore, models of the earthquake cycle with significant rheologic complexity are often computationally expensive, and thus not suited to explore the full range of permissible model parameters that are consistent with observations.Mechanical models of the earthquake cycle most often only contain Maxwell viscoelasticity, and do not consider either a robust depth- dependence of material properties or localized shear zones within the lower crust or mantle. However, the rheology of the lithosphere is likely more complicated than Maxwell viscoelasticity and material properties of the lithosphere are expected to be depth-dependent. Rheologic complexities that likely have significant impact on deformation of the lithosphere throughout the earthquake cycle include depth-dependent viscosity, power-law creep, Burgers viscoelasticity, and localized creep at depth. In this project we will: (1) Quantify the sensitivity of surface deformation throughout the earthquake cycle to rheologies at various depths.; (2) Determine if Burgers viscoelasticity and power-law creep have the same affect in models of the earthquake cycle.; (3) Systematically test the similarity of interseismic surface deformation due to localized or distributed creep at depth.; (4) Establish the time-dependent correspondence between surface deformation in idealized models and deformation in models with depth dependent viscosity, power-law creep, transient viscoelasticity, or localized creep at depth. The last point will contribute to an understanding of the non-uniquenesses inherent in trying to validate models of the earthquake cycle from observations of surface deformation. Additionally, by knowing the correspondences between idealized models and classes of more complicated models, one will be able to constrain computationally efficient idealized models to geodetic data, and then determine the range of the more complicated models directly from the idealized model inferences.