Constraining the dynamics of the present-day Alps with 3D geodynamic inverse models.

地球物理方法让我们了解岩石圈和地幔的现今结构，而地质方法则告诉我们这种结构在数百万年的变形中是如何演变的。然而，如果我们想了解阿尔卑斯山脉造山的物理过程，以及深层过程如何与地表变形相互作用，我们还需要对岩石圈动力学进行约束。虽然我们现在可以在3D中数值模拟这些过程，但这些模型的输入参数仍然不确定。此外，地球动力学模型如何很好地适应地球物理约束常常是不清楚的。我们最近开发了一种地球动力学逆建模方法，将当今岩石圈的地球动力学模型与地球物理数据进行比较，并自动改变模型参数（流变学，温度），以减少模型与数据之间的不拟合。通过在贝叶斯蒙特卡罗方法中构建它，我们可以确定最佳拟合参数以及它们的不确定性界限，甚至对于非线性流变。它允许创建力学上一致的山带解释，并显示扩展的各种提出的假设与地球物理数据一致。缺点是计算成本高，而且并不总是有一个最适合的模型。在这里，我们将对阿尔卑斯地区进行全三维地球动力学反演，以了解哪些地幔/地壳结构、热状态和流变学与阿尔卑斯地区的地球物理数据一致。除了GPS、地形和重力异常外，我们还将与4D-MB成员合作，将应变率、地震活动性、震源机制得出的应力方向、三维地震各向异性、冷却年龄和新的层析反演结果纳入我们的地球动力学反演模型中。这可能会产生更好约束的反演结果。要做到这一点，需要技术发展来计算从我们的模拟合成地球物理数据，以及改进，以提高反演方法本身的速度。最初，我们将使用已发表的数据和现有的层析模型进行工作，但新数据将通过与其他4D-MB成员的合作纳入其中。我们的模型将：-显示俯冲极性变化的动态后果，并给出岩石圈流变（RT1）的约束条件。-说明阿尔卑斯山脉的深层结构是如何与地壳和地表变形联系在一起的（RT2, RT3）。-将对地壳中的应力方向及其与更深层次过程的联系（RT4）进行约束。我们将利用DSEBRA网络（AF-A）的结果，阿尔卑斯山的地质研究（AF-E），并为阿尔卑斯山的4d数值模拟（AF-E）做出贡献。此外，我们将获得的岩石圈有效流变学的约束条件对于在百万年时间尺度上建立阿尔卑斯碰撞带的完全动态模型将是重要的。

Geophysical methods give insights in the present-day structure of the lithosphere and mantle, whereas geological methods tell us something about how this structure evolved over millions of years of deformation. Yet, if we want to understand the physics of mountain building in the Alps, and how deep processes interact with surface deformation, we also need constraints on the dynamics of the lithosphere. Whereas we can now numerically simulate such processes in 3D, the input parameters of these models remain uncertain. Moreover, it is often unclear how well geodynamic models fit geophysical constraints. We have recently developed a geodynamic inverse modelling approach in which geodynamic models of the present-day lithosphere are compared with geophysical data and the model parameters (rheology, temperature) are changed in an automatic manner to reduce the misfit between models and data. By framing this in a Bayesian Monte Carlo approach, we can determine the best-fit parameters as well as their uncertainty bounds even for nonlinear rheologies. It allows to create mechanically-consistent interpretations of mountain-belts and shows to which extend various proposed hypotheses agree with geophysical data. The disadvantage is that it is computationally expensive, and that there is not always a single best-fit model. Here, we will perform fully 3D geodynamic inversions of the Alpine region, to understand which mantle/crustal structures, thermal states, and rheologies are consistent with geophysical data of the Alps. In addition to GPS, topography and gravity anomalies, and in collaboration with 4D-MB members, we will also incorporate strain rate, seismicity, stress directions derived from focal mechanisms, 3D seismic anisotropy, cooling ages and results from new tomographic inversions in our geodynamic inverse models. This is likely to yield better-constrained inversion results. Doing this requires technical developments to compute synthetic geophysical data from our simulations, as well as improvements to increase the speed of the inversion method itself.Initially, we will use published data and existing tomographic models for our work, but new data will be incorporated through collaboration with other 4D-MB members. Our models will:- show what the dynamic consequences of subduction polarity changes are and give constraints on lithospheric rheology (RT1).- illustrate how the deep structure of the Alps is linked to crustal and surface deformation (RT2, RT3).- will give constraints on stress orientations in the crust and its link to deeper processes (RT4).We will make use of the results of the DSEBRA network (AF-A), geological studies of the Alps (AF-E) and contribute to 4D-numerical modelling of the Alps (AF-E). Moreover, the constraints we will obtain on the effective rheology of the lithosphere will be important for fully dynamic models of the Alpine collision zone on a million-year timescale.