Using machine learning to characterize solar wind driving of convection in the terrestrial magnetotail lobes

Aug 4, 2026

Upload by Patrick

Quantitative investigation of how magnetospheric convection is driven in Earth’s magnetotail lobes is conducted using ARTEMIS spacecraft measurements in the deep tail and Cluster spacecraft data in the near-to-mid tail. Building on prior results that convection can be estimated from ARTEMIS lunar ion velocities, machine learning models are trained to identify which upstream solar wind parameters most strongly drive lobe convection. The models achieve prediction-to-measurement correlations above 0.75, outperforming multiple linear regression (~0.23–0.43) in the testing dataset. Systematic analysis indicates that the IMF and magnetospheric activity significantly influence global magnetotail lobe plasma convection.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages7

Reading Time18 min

AccessFree to download

Introduction

Plasma convection in Earth’s tail regions

Solar wind and IMF driving mechanisms

Lunar exosphere and observations

Lunar ions in magnetotail lobes

Machine learning approach

Training with upstream solar wind parameters

Model performance vs linear regression

Results and systematic analysis

IMF and magnetospheric activity effects

TYPE Original Research PUBLISHED 14 August 2023 DOI 10.3389/fspas.2023.1180410 OPEN ACCESS EDITED BY Fadil Inceoglu, National Atmospheric and Oceanographic Administration (NOAA), United States REVIEWED BY Sai Gowtam Valluri, University of Alaska Fairbanks, United States Savvas Raptis, Johns Hopkins University, United States *CORRESPONDENCE Xin Cao, [xin.cao@lasp.colorado.edu](xin.cao@lasp.colorado.edu) RECEIVED 06 March 2023 ACCEPTED 02 June 2023 PUBLISHED 14 August 2023 CITATION Cao X, Halekas JS, Haaland S, Ruhunusiri S and Glassmeier K-H (2023), Using machine learning to characterize solar wind driving of convection in the terrestrial magnetotail lobes. Front. Astron. Space Sci. 10:1180410 . doi: 10.3389/fspas.2023.1180410 COPYRIGHT © 2023 Cao, Halekas, Haaland, Ruhunusiri and Glassmeier. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Using machine learning to characterize solar wind driving of convection in the terrestrial magnetotail lobes Xin Cao 1*, Jasper S. Halekas 2, Stein Haaland 3, Suranga Ruhunusiri 2 and Karl-Heinz Glassmeier 4 1 Laboratory for Atmospheric and Space Physics (LASP), University of Colorado, Boulder, CO, United States, 2 Department of Physics and Astronomy, University of Iowa, Iowa City, IA, United States, 3 Birkeland Centre for Space Science, University of Bergen, Bergen, Norway, 4 Institut für Geophysik und Extraterrestrische Physik, Technische Universität Braunschweig, Braunschweig, Germany In order to quantitatively investigate the mechanism of how magnetospheric convection is driven in the region of magnetotail lobes on a global scale, we analyzed data from the ARTEMIS spacecraft in the deep tail and data from the Cluster spacecraft in the near and mid-tail regions. Our previous work revealed that, in the lobes near the Moon’s orbit, the convection can be estimated by using ARTEMIS measurements of lunar ions’ velocity. Based on that, in this paper, we applied machine learning models to these measurements to determine which upstream solar wind parameters significantly drive the lobe convection in magnetotail regions, to help us understand the mechanism that controls the dynamics of the tail lobes. The results demonstrate that the correlations between the predicted and measured convection velocities for the machine learning models (>0.75) are superior to those of the multiple linear regression model (∼ 0.23–0.43) in the testing dataset. The systematic analysis shows that the IMF and magnetospheric activity play an important role in influencing plasma convection in the global magnetotail lobes. KEYWORDS convection, magnetosphere, solar wind, tail lobes, ARTEMIS, machine learning 1 Introduction Characterizing the plasma convection in Earth’s tail regions is important to help us understand global magnetospheric dynamics. Haaland et al. (2008) and Haaland et al.(2009) used Cluster data (Escoubet et al., 1997) to show that the plasma convection at ∼ 10 RE downtail has opposite lateral patterns in the southern and northern lobes. For instance, the convection shows a pattern such that the north-south convection moves towards the current sheet in the magnetotail. Ohma et al. (2019) revealed that the asymmetry of the convection flow could also be affected by magnetic reconnection in the tail, which relates to magnetospheric activity. Cao et al. (2020b) used the two Acceleration, Reconnection, Turbulence, and Electrodynamics of Moon’s Interaction with the Sun (ARTEMIS) lunar ion data (Angelopoulos, 2011) to show that the dawn-dusk component of plasma

FAQs

What data sources are used to study lobe convection driving?

The analysis combines ARTEMIS spacecraft measurements from the deep tail with Cluster spacecraft data from near and mid-tail regions.

How is convection estimated in the lobe regions?

Convection is estimated using ARTEMIS measurements of lunar ions’ velocity, leveraging the relationship between lunar ion observations and lobe convection.

Which upstream factors are found to drive solar wind-related convection in the magnetotail lobes?

The results show that the interplanetary magnetic field (IMF) and magnetospheric activity play important roles in influencing plasma convection across the global magnetotail lobes.

Aug 4, 2026

Upload by Patrick