The data-driven future of high energy density physics

Abstract

The study of plasma physics under conditions of extreme temperatures, densities and electromagnetic field strengths is significant for our understanding of astrophysics, nuclear fusion and fundamental physics. These extreme physical systems are strongly non-linear and very difficult to understand theoretically or optimize experimentally. Here, we argue that machine learning models and data-driven methods are in the process of reshaping our exploration of these extreme systems that have hitherto proven far too non-linear for human researchers. From a fundamental perspective, our understanding can be helped by the way in which machine learning models can rapidly discover complex interactions in large data sets. From a practical point of view, the newest generation of extreme physics facilities can perform experiments multiple times a second (as opposed to ~daily), moving away from human-based control towards automatic control based on real-time interpretation of diagnostic data and updates of the physics model. To make the most of these emerging opportunities, we advance proposals for the community in terms of research design, training, best practices, and support for synthetic diagnostics and data analysis.

arXiv comment14 pages, 4 figures. This work was the result of a meeting at the Lorentz Center, University of Leiden, 13th-17th January 2020. This is a preprint of Hatfield et al., Nature, 593, 7859, 351-361 (2021) https://www.nature.com/articles/s41586-021-03382-w

References in corpus (43)

• The Challenge of Machine Learning in Space Weather Nowcasting and Forecasting
• Shock Compression of Deuterium and the Interiors of Jupiter and Saturn
• Multi-messenger Bayesian parameter inference of a binary neutron-star merger
• A generalized bayesian inference method for constraining the interiors of super Earths and sub-Neptunes
• Laboratory evidence of dynamo amplification of magnetic fields in a turbulent plasma
• Photometric Supernova Classification With Machine Learning
• Automation and control of laser wakefield accelerators using Bayesian optimisation
• Predicting Coronal Mass Ejections Using Machine Learning Methods
• Astrophysical Black Holes: A Compact Pedagogical Review
• Identifying Solar Flare Precursors Using Time Series of SDO/HMI Images and SHARP Parameters
• Fast modeling of turbulent transport in fusion plasmas using neural networks
• Fundamental Parameters of Main-Sequence Stars in an Instant with Machine Learning
• Coherent control of plasma dynamics
• Classification of Solar Wind with Machine Learning
• Contemporary machine learning: a guide for practitioners in the physical sciences
• Feature ranking of active region source properties in solar flare forecasting and the uncompromised stochasticity of flare occurrence
• A Survey of Deep Learning for Scientific Discovery
• Using Machine Learning Methods to Forecast If Solar Flares Will Be Associated with CMEs and SEPs
• Collimated Ultra-Bright Gamma-Rays from a PW-Laser-Driven Wire Wiggler
• A variational approach to Data Assimilation in the Solar Wind
• Inverse Problem Instabilities in Large-Scale Plasma Modelling
• Laboratory Measurements of White Dwarf Photospheric Spectral Lines: H$\beta$
• The Blind Implosion-Maker - Automated Inertial Confinement Fusion experiment design
• A deep learning model to emulate simulations of cosmic reionization
• Thermonuclear fusion rates for tritium + deuterium using Bayesian methods
• Conservative cosmology: combining data with allowance for unknown systematics
• Designing Accurate Emulators for Scientific Processes using Calibration-Driven Deep Models
• Building high accuracy emulators for scientific simulations with deep neural architecture search
• A gray-box model for a probabilistic estimate of regional ground magnetic perturbations: Enhancing the NOAA operational Geospace model with machine learning
• Exploring the Long-Term Evolution of GRS 1915+105
• Observing Thermal Schwinger Pair Production
• Temporal Feedback Control of High-Intensity Laser Pulses to Optimize Ultrafast Heating of Atomic Clusters
• Spectral feature extraction for DB white dwarfs through machine learning applied to new discoveries in the SDSS DR12 and DR14
• The Revolution in Astronomy Education: Data Science for the Masses
• Development of a Bayesian method for the analysis of inertial confinement fusion experiments on the NIF
• Quantification of MagLIF morphology using the Mallat Scattering Transformation
• Using Sparse Gaussian Processes for Predicting Robust Inertial Confinement Fusion Implosion Yields
• Research towards high-repetition rate laser-driven X-ray sources for imaging applications
• Bayesian inference of inaccuracies in radiation transport physics from inertial confinement fusion experiments
• Bayesian inference of quasi-linear radial diffusion parameters using Van Allen Probes
• Plasma Wakefield Accelerator Research 2019 - 2040: A community-driven UK roadmap compiled by the Plasma Wakefield Accelerator Steering Committee (PWASC)
• Deep neural network initialization with decision trees
• Correlation of Auroral Dynamics and GNSS Scintillation with an Autoencoder

Cited by in corpus (9)

• Accurate Temperature Diagnostics for Matter under Extreme Conditions
• 2022 Review of Data-Driven Plasma Science
• Numerical modeling and simulation on nanosecond laser-target interactions
• A laser-plasma platform for photon-photon physics
• First Proof of Principle Experiment for Muon Production with Ultrashort High Intensity Laser
• Beyond optimization -- supervised learning applications in relativistic laser-plasma experiments
• An electron-beam based Compton scattering x-ray source for probing high-energy-density physics
• Characterization and automated optimization of laser-driven proton beams from converging liquid sheet jet targets
• Physics-enhanced neural networks for equation-of-state calculations