From plate tectonic reconstructions to 4D geodynamic models of the Alpine Orogeny

阿尔卑斯山的大多数地质演化情景依赖于板块构造重建、地质数据和对当今阿尔卑斯山的地球物理观测。这种重建给出了阿尔卑斯山可能是如何形成的看似合理的场景，但它们纯粹是基于运动学规则。因此，目前还不清楚它们是否也与岩石圈和地幔的物理学相一致。为了充分理解4D中的造山过程，我们将从两个方面着手：1)我们建议使用时间分辨的3D地球动力学模型来研究构造场景的物理一致性。他们对岩石圈的流变性、板片断裂事件等现象的发生等参数提供了见解，并展示了岩石圈的重组如何影响造山。动态一致的模型还允许我们改进板块构造重建，并指导地震层析成像模型的解释。2)我们将在板块构造重建中考虑不确定性。这将使我们更好地了解哪些特征是健壮的，哪些特征不是健壮的，这是将它们与动态模型进行比较时的关键信息。我们将首先从板块构造重建开始，使用反向和正演相结合的建模方法，对阿尔卑斯山进行3D热力地球动力学模拟。为了使问题易于处理，我们从20 Ma的重建开始，并进行系统的正演模拟，以探索不同流变性、初始几何形状和热结构的替代演化场景。我们的模拟将包括非线性粘弹塑性流变学和自由表面，从而能够以自洽的方式模拟剪切带的自发发生、板片断裂事件和新洋壳的产生。侵蚀以一种简化的方式进行，这允许测试表层和深层岩石圈过程之间的相互作用。下一步，我们将通过对时间相关模型的不同快照的准瞬时模型进行灵敏度分析，从地球动力学模拟中提取板块运动规则和不确定性，以约束模型参数的微小变化对板块速度的影响。这是本项目第三部分开发的一种新的反模拟方法的重要组成部分，该方法利用数据同化技术将板块构造重建描述为反问题。在这方面，我们考虑了数据中的不确定性，以及基于物理的板块速度约束，这将导致一系列与数据一致的板块构造重建。结合4DMB项目的新数据和现有数据，这将为造山过程的物理学提供新的见解。

Most geological evolution scenarios for the Alps rely on plate tectonic reconstructions, geological data and geophysical observations of the present-day Alps. Such reconstructions give plausible scenarios on how the Alps may have formed, but they are purely based on kinematic rules. It is therefore unclear whether they are also consistent with the physics of the lithosphere and mantle. Moreover, uncertainties of the various constraints that are included in the reconstructions are usually not considered.To fully understand mountain building processes in 4D, we will address both aspects:1) We propose to investigate the physical consistency of tectonic scenarios using time-resolved 3D geodynamic models. They give insights in parameters such as the rheology of the lithosphere, the occurrence of phenomena like slab breakoff events, and show how the reorganization of the lithosphere affects mountain building. Dynamically consistent models also allow us to refine the plate tectonic reconstructions and guide the interpretations of seismic tomography models. 2) We will take uncertainties into account in the plate tectonic reconstructions. This will give a better understanding of which features are robust and which less so, a crucial piece of information when comparing them with dynamic models.We will first perform 3D thermo-mechanical geodynamic simulations of the Alps, starting from plate tectonic reconstructions and using a combined inverse and forward modelling approach. To make the problem tractable, we start with reconstructions at 20 Ma and perform systematic forward simulations to explore alternative evolution scenarios for different rheologies, initial geometries and thermal structures. Our simulations will include nonlinear visco-elasto-plastic rheologies and a free surface and will thus be able to simulate the spontaneous occurrence of shear zones, slab breakoff events and the generation of new oceanic crust in a self-consistent manner. Erosion is implemented in a simplified manner, which allows testing the interaction between surface and deep lithospheric processes. In a next step, we will extract plate kinematic rules and uncertainties from the geodynamic simulations by performing a sensitivity analysis on quasi-instantaneous models for different snapshots of the time-dependent models, to constrain how small changes in the model parameters affect plate velocities. This is an important ingredient for a new inverse modelling approach, developed in the third part of this project, that employs data assimilation techniques to formulate plate tectonic reconstructions as an inverse problem. In this, we take uncertainties in the data, as well as physics-based constraints on plate velocities into account which will result in a range of plate tectonic reconstructions that are all consistent with the data. Combined with new and existing data of the 4DMB project, this will give new insights in the physics of mountain building processes.