Physics-based machine learning for mantle convection simulations

• URL: http://arxiv.org/abs/2505.16041v1
• Date: Wed, 21 May 2025 21:47:43 GMT
• Title: Physics-based machine learning for mantle convection simulations
• Authors: Siddhant Agarwal, Ali Can Bekar, Christian Hüttig, David S. Greenberg, Nicola Tosi,
• Abstract summary: We propose a machine learning approach that predicts creeping flow velocities as a function of temperature while conserving mass.<n>A finite-volume solver then uses the predicted velocities to advect and diffuse the temperature field to the next time-step, enabling autoregressive rollout at inference.<n>Overall, our model is up to 89 times faster than the numerical solver.
• Score: 2.4608654148475235
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Mantle convection simulations are an essential tool for understanding how rocky planets evolve. However, the poorly known input parameters to these simulations, the non-linear dependence of transport properties on pressure and temperature, and the long integration times in excess of several billion years all pose a computational challenge for numerical solvers. We propose a physics-based machine learning approach that predicts creeping flow velocities as a function of temperature while conserving mass, thereby bypassing the numerical solution of the Stokes problem. A finite-volume solver then uses the predicted velocities to advect and diffuse the temperature field to the next time-step, enabling autoregressive rollout at inference. For training, our model requires temperature-velocity snapshots from a handful of simulations (94). We consider mantle convection in a two-dimensional rectangular box with basal and internal heating, pressure- and temperature-dependent viscosity. Overall, our model is up to 89 times faster than the numerical solver. We also show the importance of different components in our convolutional neural network architecture such as mass conservation, learned paddings on the boundaries, and loss scaling for the overall rollout performance. Finally, we test our approach on unseen scenarios to demonstrate some of its strengths and weaknesses.

Related papers

• Hybrid Neural-MPM for Interactive Fluid Simulations in Real-Time [57.30651532625017]
  We present a novel hybrid method that integrates numerical simulation, neural physics, and generative control.<n>Our system demonstrates robust performance across diverse 2D/3D scenarios, material types, and obstacle interactions.<n>We promise to release both models and data upon acceptance.
  arXiv  Detail & Related papers  (2025-05-25T01:27:18Z)
• A Physics-Informed Convolutional Long Short Term Memory Statistical Model for Fluid Thermodynamics Simulations [0.0]
  Direct numerical simulations of fluid thermodynamics are computationally prohibitive.<n>We present a physics-informed architecture for RBC, a canonical example of convective flow.<n>Inference is penalized with respect to the governing partial differential equations to ensure interpretability.
  arXiv  Detail & Related papers  (2025-05-16T06:47:00Z)
• Accelerating the discovery of steady-states of planetary interior dynamics with machine learning [2.4608654148475235]
  We present a concept for accelerating mantle convection simulations using machine learning. We train a feedforward neural network on 97 simulations to predict steady-state temperature profiles. The number of time steps required to reach steady-state is reduced by a median factor of 3.75.
  arXiv  Detail & Related papers  (2024-08-30T13:55:19Z)
• Inferring turbulent velocity and temperature fields and their statistics from Lagrangian velocity measurements using physics-informed Kolmogorov-Arnold Networks [3.7130378544005844]
  We propose the Artificial Intelligence Velocimetry-Thermometry (AIVT) method to infer hidden temperature fields from experimental turbulent velocity data. Specifically, AIVT is based on physics-informed Kolmogorov-Arnold Networks (not neural networks) and is trained by optimizing a combined loss function. We demonstrate that we can reconstruct and infer continuous and instantaneous velocity and temperature fields from sparse experimental data at a fidelity comparable to direct numerical simulations (DNS) of turbulence.
  arXiv  Detail & Related papers  (2024-07-22T15:30:21Z)
• Accurate melting point prediction through autonomous physics-informed learning [52.217497897835344]
  We present an algorithm for computing melting points by autonomously learning from coexistence simulations in the NPT ensemble. We demonstrate how incorporating physical models of the solid-liquid coexistence evolution enhances the algorithm's accuracy and enables optimal decision-making.
  arXiv  Detail & Related papers  (2023-06-23T07:53:09Z)
• Machine learning of hidden variables in multiscale fluid simulation [77.34726150561087]
  Solving fluid dynamics equations often requires the use of closure relations that account for missing microphysics. In our study, a partial differential equation simulator that is end-to-end differentiable is used to train judiciously placed neural networks. We show that this method enables an equation based approach to reproduce non-linear, large Knudsen number plasma physics.
  arXiv  Detail & Related papers  (2023-06-19T06:02:53Z)
• Robust Extraction of Thermal Observables from State Sampling and Real-Time Dynamics on Quantum Computers [49.1574468325115]
  We introduce a technique that imposes constraints on the density of states, most notably its non-negativity, and show that this way, we can reliably extract Boltzmann weights from noisy time series. Our work enables the implementation of the time-series algorithm on present-day quantum computers to study finite temperature properties of many-body quantum systems.
  arXiv  Detail & Related papers  (2023-05-30T18:00:05Z)
• Physics-informed machine learning with differentiable programming for heterogeneous underground reservoir pressure management [64.17887333976593]
  Avoiding over-pressurization in subsurface reservoirs is critical for applications like CO2 sequestration and wastewater injection. Managing the pressures by controlling injection/extraction are challenging because of complex heterogeneity in the subsurface. We use differentiable programming with a full-physics model and machine learning to determine the fluid extraction rates that prevent over-pressurization.
  arXiv  Detail & Related papers  (2022-06-21T20:38:13Z)
• Machine learning accelerated computational fluid dynamics [9.077691121640333]
  We use end-to-end deep learning to improve approximations inside computational fluid dynamics for modeling two-dimensional turbulent flows. For both direct numerical simulation of turbulence and large eddy simulation, our results are as accurate as baseline solvers with 8-10x finer resolution in each spatial dimension. Our approach exemplifies how scientific computing can leverage machine learning and hardware accelerators to improve simulations without sacrificing accuracy or generalization.
  arXiv  Detail & Related papers  (2021-01-28T19:10:00Z)
• Machine learning for rapid discovery of laminar flow channel wall modifications that enhance heat transfer [56.34005280792013]
  We present a combination of accurate numerical simulations of arbitrary, flat, and non-flat channels and machine learning models predicting drag coefficient and Stanton number. We show that convolutional neural networks (CNN) can accurately predict the target properties at a fraction of the time of numerical simulations.
  arXiv  Detail & Related papers  (2021-01-19T16:14:02Z)
• Simulating quantum vibronic dynamics at finite temperatures with many body wave functions at 0K [0.0]
  We present numerical simulations that exploit a recent theoretical result that allows dissipative environmental effects at finite temperature to be extracted efficiently from a single, zero-temperature wave function simulation. We provide insight into the practical problems lurking behind the elegance of the theory, such as the rapidly growing numerical demands that can appear for high temperatures over the length of computations.
  arXiv  Detail & Related papers  (2021-01-04T17:15:16Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.