Machine learning helps solve problems in heliophysics

Machine learning helps solve problems in heliophysics

Spread the love

Voice of editors is a blog from the AGU Publications Department.

Heliophysics is a broad subject that covers the study of the interior of the Sun, solar physics, the interplanetary medium, solar wind-magnetosphere interactions, the dynamics of the magnetosphere and its coupling with the ionosphere-thermosphere system, or the effects of space weather on space and terrestrial infrastructures.

Like all other disciplines, heliophysics is not immune to the machine learning revolution currently taking place in science.

Like all other disciplines, heliophysics is not immune to the machine learning revolution currently taking place in science. Indeed, many now recognize that the techniques and tools developed by computer scientists (often to address completely different applications), combined with big data and easy access to accelerated computing, are becoming the fourth pillar of scientific discovery. More importantly, it is now abundantly clear that data-driven approaches investigated through machine learning techniques will become increasingly mainstream and, therefore, are here to stay.

The heliophysical community now faces the challenge of overcoming the barrier of technical skills posed by machine learning that are not usually mastered by the typical scientist. We must therefore fully appreciate and critically understand what is within reach in a few years and what could be achieved in a decade.

Heliophysics is a highly interdisciplinary subject, where the focus of a research effort can range from basic understanding of physical phenomena to designing an operational space weather capability capable of forecasting specific events. The number of ways in which machine learning can be generally exploited in earth and space sciences has been presented in Bortnik and Camporeale (2021).

The new cross-journal special collection “Machine Learning in Heliophysics” follows the 2n/a edition of a community conference of the same name successfully held in Boulder, Colorado in March 2022. The conference program and many oral and poster presentations are still available for download. The main takeaway from the conference was that the field is rapidly moving from an exploratory phase where machine learning techniques have often been unsuccessfully attempted on problems that were not well suited for such approaches to a more mature which gives a much higher success rate and convincing results. .

Among all the possible uses of machine learning techniques in heliophysics, we highlight a few that we believe have the potential for a real scientific breakthrough in the next few years:

  • Reduced-order modeling or acceleration/emulation of computationally expensive physics-based models
  • Physics-based machine learning, where physics-based constraints are encoded in a machine learning architecture. This is often referred to as a “grey box” approach that bridges purely data-based (black box) and physics-based (white box) approaches, with the aim of leveraging the advantages of both approaches and limiting their weaknesses.
  • Data-guided discovery of new physical laws and/or new parameterizations of physical quantities

Other topics relevant to the special collection include inverse estimation of physical parameters, automatic event identification, feature detection and tracking, time series analysis of dynamical systems, combination of models based on physics with machine learning techniques, surrogate models and uncertainty quantification.

This is a joint special collection between Space weather, JGR: Space Physics, Geophysical Research Lettersand Earth and Space Sciences. Manuscripts may be submitted to any of these journals, depending on their fit with the scope and requirements of the journal.

—Enrico Camporeale (; 0000-0002-7862-6383), University of Colorado, USA; Veronique Delouille (0000-0001-5307-8045), Royal Observatory of Belgium, Belgium; Thomas Berger (0000-0002-4989-475X), University of Colorado, USA; and Sophie Murray (0000-0002-9378-5315), Dublin Institute for Advanced Studies, Ireland

Quote: Camporeale, E., V. Delouille, T. Berger and S. Murray (2022), Machine Learning Helps Solve Problems in Heliophysics, Eos, 103, Posted November 3, 2022.
This article does not represent the opinion of AGU, Eos, or any of its affiliates. This is solely the opinion of the author or authors.
Text © 2022. The authors. CC BY-NC-ND 3.0
Unless otherwise stated, images are subject to copyright. Any reuse without the express permission of the copyright holder is prohibited.

Leave a Comment

Your email address will not be published.