Fresh Outlook on Numerical Methods for Geodynamics. Part 2: Big Data, HPC, Education

Morra, Gabriele; Yuen, David A.; Tufo, Henry M.; Knepley, Matthew G.

doi:10.1016/b978-0-08-102908-4.00111-9

Cited by 4 publications

(3 citation statements)

References 102 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Some other examples of Machine learning (ML) studies include Baumann (2016), who approached the problem of constraining the dynamics and rheology of the lithosphere in collision zones using an unsupervised classification algorithm called self‐organizing map (Vesanto & Alhoniemi, 2000) and Shahnas et al. (2018), who used support‐vector machines to estimate the magnitude of density anomalies from mantle temperature fields (see Morra et al., 2020, for a recent review on the application of data science techniques in geodynamics).…”

Section: Introductionmentioning

confidence: 99%

Toward Constraining Mars' Thermal Evolution Using Machine Learning

Agarwal

Tosi

Kessel

et al. 2021

Earth and Space Science

View full text Add to dashboard Cite

• Mixture Density Networks provide a probabilistic framework for inverting observables to infer parameters of Mars' interior evolution. 10 • Reference viscosity, crustal enrichment in heat-producing elements and initial mantle 11 temperature can be well constrained. 12 • Activation energy of diffusion creep can be weakly constrained; constraining activation 13 volume requires new observational signatures.

show abstract

Section: Introductionmentioning

confidence: 99%

Toward Constraining Mars' Thermal Evolution Using Machine Learning

Agarwal

Tosi

Kessel

et al. 2021

Earth and Space Science

View full text Add to dashboard Cite

show abstract

“…To our best knowledge, this is the first time a parameterized 2D surrogate is proposed in the context of mantle convection (see Ref. [31] for a recent review of applications of data science methods in geodynamics). We use the term "parameterized" to stress that we are interested in predicting a variety of mantle flows based on different combinations of input parameters and not, for example, given a certain amount of time steps of a single simulation with fixed parameters, predicting the subsequent time steps.…”

Section: Introductionmentioning

confidence: 99%

Deep learning for surrogate modeling of two-dimensional mantle convection

et al. 2021

View full text Add to dashboard Cite

“…To our knowledge, this is the first time a parameterized 2D surrogate is proposed in the context of mantle convection (see [27] for a recent review of applications of data science methods in geodynamics). We use the term parameterized to stress that we are interested in predicting a variety of mantle flows based on different combinations of input parameters and not, for example, given a certain amount of time-steps of a single simulation with fixed parameters, predicting the subsequent time-steps.…”

mentioning

confidence: 99%

Deep learning for surrogate modelling of 2D mantle convection

Agarwal,

Tosi,

Kessel

et al. 2021

Preprint

View full text Add to dashboard Cite

Mantle convection, the buoyancy-driven creeping flow of silicate rocks in the interior of terrestrial planets like Earth, Mars, Mercury and Venus, plays a fundamental role in the long-term thermal evolution of these bodies. Yet, key parameters and initial conditions of the partial differential equations governing mantle convection are poorly constrained. This often requires a large sampling of the parameter space to determine which combinations can satisfy certain observational constraints. Traditionally, 1D models based on scaling laws used to parameterized convective heat transfer, have been used to tackle the computational bottleneck of high-fidelity forward runs in 2D or 3D. However, these are limited in the amount of physics they can model (e.g. depth dependent material properties) and predict only mean quantities such as the mean mantle temperature. A recent machine learning study has shown that feedforward neural networks (FNN) trained using a large number of 2D simulations can overcome this limitation and reliably predict the evolution of entire 1D laterally-averaged temperature profile in time for complex models [1]. We now extend that approach to predict the full 2D temperature field, which contains more information in the form of convection structures such as hot plumes and cold downwellings. Using a dataset of 10, 525 two-dimensional simulations of the thermal evolution of the mantle of a Mars-like planet, we show that deep learning techniques can produce reliable parameterized surrogates (i.e. surrogates that predict state variables such as temperature based only on parameters) of the underlying partial differential equations. We first use convolutional autoencoders to compress the temperature fields by a factor of 142 and then use FNN and long-short term memory networks (LSTM) to predict the compressed fields. On average, the FNN predictions are 99.30% and the LSTM predictions are 99.22% accurate with respect to unseen simulations. Proper orthogonal decomposition (POD) of the LSTM and FNN predictions shows that despite a lower mean absolute relative accuracy, LSTMs capture the flow dynamics better than FNNs. When summed, the POD coefficients from FNN predictions and from LSTM predictions amount to 96.51% and 97.66% relative to the coefficients of the original simulations, respectively. I. INTRODUCTIONStudying the long-term thermal evolution of terrestrial planets like Earth, Venus, Mercury and Mars requires detailed modelling of thermal convection in their rocky mantles (e.g., [2]). Similar

show abstract

Fresh Outlook on Numerical Methods for Geodynamics. Part 2: Big Data, HPC, Education

Cited by 4 publications

References 102 publications

Toward Constraining Mars' Thermal Evolution Using Machine Learning

Toward Constraining Mars' Thermal Evolution Using Machine Learning

Deep learning for surrogate modeling of two-dimensional mantle convection

Deep learning for surrogate modelling of 2D mantle convection

Contact Info

Product

Resources

About