Collaborative Research: Development and Application of a Framework for Integrated Geodynamic Earth Models

几十年来，地球科学界一直梦想并致力于建立模拟，以解决全球和区域观测到的固体地球变形模式的时间和长度尺度。这包括地球深部岩石的缓慢运动，构造板块的运动，以及这些板块内部和这些板块之间边界的较小规模的局部变形，时间尺度从数千年到数百万年不等。直到最近，无论是计算工具，还是关于岩石在地球内部温度和压力下表现如何的必要信息，都无法实现合理精度的模拟。然而，随着最近地球科学和计算的进步，我们终于有可能开发出从地幔深处到地表的地球计算模型。该项目旨在开发一个框架来构建地球动力学地球模型，基于PI在过去8年中一直在构建的广泛使用的社区建模代码方面。这些模拟有可能为广泛的主题提供巨大的洞察力，包括构造板块运动和变形的时间和空间变化、岩浆流动和水在地球内部的循环、地球深处的结构和地貌演变。所有工作都将通过开源软件和数据集提供给其他科学家，包括教程和文档模块，以帮助其他人在实践中使用这项工作。此外，该项目将创建无障碍图像、视频和更精致的教育材料，通过外展活动与高中生和大学早期学生分享。除此之外，该项目将通过创建在大型领导层计算设施上运行的复杂、多物理模拟代码来建立的经验，对于解决许多其他重大科学挑战所需的复杂代码也很重要，例如NSF的几个“大想法”。这个项目是关于建立一个综合的地球动力学地球模型，用于在数千年至数百万年的时间尺度上真实地模拟从核-地幔边界到地表的地球。它将解决一系列长期存在的问题，涉及固体地球的物理结构、全球和区域变形模式、由板块边界决定的物质循环以及耦合的表面演化。地球物理数据集的同化和处理将产生一个起始地球模型，提供从地表到核幔边界的地球热-化学-流变性状态的详细描述。高分辨率的全球模拟将使用这种对地球物理状态的详细描述来确定脆性和韧性流变学如何控制观察到的板块边界内变形和断层相互作用的划分。在起始地球模型和全球模拟的基础上，全球嵌入的区域模拟将使其能够确定美国西部地区内的流变和浮力变化如何控制观测到的变形模式。结合完全耦合的两相流体输送和反应，该项目将采用全球嵌入的区域模拟，以估计俯冲大洋板块内挥发分输送的全球速率和量级，并深入了解板块边界和地幔深处的挥发分通量模式。此外，该项目将促进将这些模拟与地貌演变模型相结合，以确定地表过程如何改变俯冲动力学中的时间变化。最后，这个项目将导致开发新的工具来可视化这些模拟结果，并将它们用于各种外展活动。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

For decades, the geosciences community has dreamed of and worked towards building simulations that can resolve the time and length scales of deformation patterns in the solid Earth observed both globally and regionally. This includes the slow motion of rocks in the Earth's deep interior, the motion of tectonic plates, and smaller scale localized deformation in the interior of and at the boundaries between these plates on time scales ranging from thousands to millions of years. Until recently, neither the computational tools, nor the requisite information about how rocks behave at the temperature and pressures of the Earth's interior were available to allow such simulations with reasonable accuracy. However, with recent advances in the Earth sciences and computing, we are finally at a point where it is possible to develop computational models of the Earth from the deep mantle to surface. This project is aimed at developing a framework for building a Geodynamic Earth Models, based on the widely used community modeling code ASPECT that the PIs have been building for the past 8 years. These simulations have the potential to provide enormous insight into a wide range of topics, including temporal and spatial variations in the motion and deformation of tectonic plates, the flow of magma and the cycling of water through the Earth's interior, the structure of the deep Earth, and landscape evolution. All work will be made available to other scientists through open source software and data sets, including tutorials and documentation modules to help others use this work in practice. In addition, the project will create accessible images, videos, and more elaborate educational material that will be shared with high school and early college students through outreach events. Beyond that, the experience this project will build by creating a complex, multi-physics simulation code running on large leadership-level computing facilities is also important for complex codes needed to address many other scientific grand challenges, such as several of NSF's "Big Ideas". This project is about the creation of an Integrated Geodynamic Earth Model for the realistic simulation of the Earth from the core-mantle boundary to the surface on time scales of thousands to millions of years. It will address a series of long-standing questions regarding the physical structure of the solid Earth, global and regional deformation patterns, material cycles determined by plate boundaries, and coupled surface evolution. Assimilation and processing of geophysical data sets will generate a Starting Earth Model providing a detailed description of the Earth's thermal-chemical-rheological state from the surface to the core-mantle boundary. High-resolution global simulations will use this detailed description of Earth's physical state to determine how brittle and ductile rheology controls the partitioning of deformation and fault interaction within observed plate boundaries. Building on the Starting Earth Model and global simulations, globally embedded regional simulations will allow it to determine how rheological and buoyancy variations within the Western USA control observed deformation patterns. In combination with fully coupled two-phase fluid transport and reactions, the project will employ globally-embedded regional simulations to estimate global rates and magnitudes of volatile transport within subducting oceanic plates and provide insight into plate boundary and deep mantle volatile flux patterns. Additionally, the project will facilitate the coupling of these simulations to landscape evolution models to determine how surface processes modify temporal variations in subduction dynamics. Finally, this project will lead to the development of new tools for the visualization of these simulation results, and use them for a variety of outreach activities.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.