CME-Driven Joule Heating in Planetary Systems

• CME-Driven Joule Heating is the resistive energy dissipation from CME-induced electric fields that heats both the ionosphere and planetary interiors.
• MHD simulations reveal that 1-hour CME events can yield interior heating of 20 TW on Trappist-1b and up to 10³–10⁴ TW in the ionosphere of Trappist-1e.
• Stronger magnetic fields reduce ionospheric heating while enhancing interior induction, as the inward Poynting flux scales cubically with magnetic field strength.

CME-driven Joule heating denotes the resistive dissipation that occurs when a coronal mass ejection drives electric fields and currents in a coupled plasma–planet system, or more generally when CME-associated magnetic restructuring generates current dissipation in eruptive plasma. In the exoplanetary setting developed for the Trappist-1 system, the process has two principal channels: ionospheric Joule heating in a partially ionized upper atmosphere and induction heating in an electrically conducting interior. Recent magnetohydrodynamic modeling of Trappist-1b and Trappist-1e shows that single 1-hour CME events can dissipate about $20\ \mathrm{TW}$ and $1\ \mathrm{TW}$ in the interiors of the two planets, while ionospheric Joule heating on Trappist-1e reaches $10^{3-4}\ \mathrm{TW}$ during an event, far above steady-wind and stellar-XUV inputs (Elekes et al., 18 Jun 2025).

1. Formal definition and energy channels

In the Trappist-1 CME interaction framework, ionospheric Joule heating is defined as the rate of work done by the ionospheric electric field on the Pedersen currents,

$$Q_{\rm ion}=\int_V \sigma_P E^2\,dV=\int_V J\cdot E\,dV,$$

where $\sigma_P(r,t)$ is the local Pedersen conductivity, $E(r,t)=-v\times B$ is the convective electric field, $J(r,t)=\sigma_P E$ is the current density, and $V$ is the volume of the ionospheric shell. Interior induction heating is written as

$$Q_{\rm int}=\int_V \eta |J|^2\,dV=\int_V \sigma |E|^2\,dV,$$

where $\eta(r)=1/\sigma(r)$ is the electrical resistivity, $1\ \mathrm{TW}$0 is the electrical conductivity of the interior, $1\ \mathrm{TW}$1 is the induced current density, $1\ \mathrm{TW}$2 is the induced electric field, and $1\ \mathrm{TW}$3 is the planetary volume relevant for induction (Elekes et al., 18 Jun 2025).

These definitions distinguish two dissipation reservoirs. The ionospheric term measures direct conversion of electromagnetic forcing into atmospheric heat through Pedersen currents. The interior term measures eddy-current dissipation driven by time-varying external magnetic fields. In the same formulation, the magnetospheric energy supply is tracked with the Poynting flux density

$1\ \mathrm{TW}$4

and the total Poynting power crossing the planetary boundary

$1\ \mathrm{TW}$5

The simulations compute the inward radial component,

$1\ \mathrm{TW}$6

so that energy delivery toward the planetary surface can be separated from the eventual partition into ionosphere and interior (Elekes et al., 18 Jun 2025).

2. Magnetospheric coupling and magnetic-field scaling

A central result of the Trappist-1 study is that magnetospheric compression is the main driver of magnetic variability. Within the range of studied intrinsic equatorial magnetic field strengths $1\ \mathrm{TW}$7, the time-average of the inward Poynting flux scales approximately as

$1\ \mathrm{TW}$8

The paper attributes this cubic dependence to three effects: a stronger dipole supports a larger magnetopause cross-section $1\ \mathrm{TW}$9, it sustains higher Alfvén speeds, and it stores more magnetic energy for a given perturbation (Elekes et al., 18 Jun 2025).

This scaling establishes a nontrivial relation between shielding and energy absorption. Planetary magnetic fields reduce the direct impact of CME-embedded fields, but they enhance induction heating in the interior, although the effect is weaker with flux rope CMEs. At the same time, ionospheric Joule heating decreases for stronger magnetic fields $10^{3-4}\ \mathrm{TW}$0. The consequence is channel-dependent rather than monotonic: stronger magnetization does not simply suppress CME energy deposition, because the same field that buffers direct penetration also increases magnetospheric Poynting-flux capture (Elekes et al., 18 Jun 2025).

3. MHD realization for Trappist-1b and Trappist-1e

The interaction model is a single-fluid, ideal MHD system solving mass continuity, momentum, energy, and induction, including source terms for ionospheric production/loss. The computational domain is a spherical grid from $10^{3-4}\ \mathrm{TW}$1 to $10^{3-4}\ \mathrm{TW}$2 around the planet, with an upstream boundary fixed to a steady-state stellar wind from Dong et al. (2018). The planetary magnetic field is a dipole aligned with the rotation axis, with equatorial surface strength $10^{3-4}\ \mathrm{TW}$3 (Elekes et al., 18 Jun 2025).

For Trappist-1e, the atmosphere is represented as a thin $10^{3-4}\ \mathrm{TW}$4 layer with base density $10^{3-4}\ \mathrm{TW}$5 and scale height $10^{3-4}\ \mathrm{TW}$6; the photo-ionization rate $10^{3-4}\ \mathrm{TW}$7 is adjusted to yield an O atom loss rate $10^{3-4}\ \mathrm{TW}$8. Two CME prescriptions are used. The density-pulse CME applies a Gaussian enhancement of stellar-wind density $10^{3-4}\ \mathrm{TW}$9 and speed $$Q_{\rm ion}=\int_V \sigma_P E^2\,dV=\int_V J\cdot E\,dV,$$0 over a 1-hour duration such that the total energy density matches a CME associated with a flare of bolometric energy $$Q_{\rm ion}=\int_V \sigma_P E^2\,dV=\int_V J\cdot E\,dV,$$1. The flux-rope CME is a Gold–Hoyle force-free flux rope of radius $$Q_{\rm ion}=\int_V \sigma_P E^2\,dV=\int_V J\cdot E\,dV,$$2 and twist $$Q_{\rm ion}=\int_V \sigma_P E^2\,dV=\int_V J\cdot E\,dV,$$3, with

$$Q_{\rm ion}=\int_V \sigma_P E^2\,dV=\int_V J\cdot E\,dV,$$4

where $$Q_{\rm ion}=\int_V \sigma_P E^2\,dV=\int_V J\cdot E\,dV,$$5 is set by scaling laws for magnetic helicity $$Q_{\rm ion}=\int_V \sigma_P E^2\,dV=\int_V J\cdot E\,dV,$$6, and the bulk speed $$Q_{\rm ion}=\int_V \sigma_P E^2\,dV=\int_V J\cdot E\,dV,$$7 is derived from Kay et al. (2019) (Elekes et al., 18 Jun 2025).

The simulations span flare energies $$Q_{\rm ion}=\int_V \sigma_P E^2\,dV=\int_V J\cdot E\,dV,$$8. In the reference case $$Q_{\rm ion}=\int_V \sigma_P E^2\,dV=\int_V J\cdot E\,dV,$$9, the density-pulse model gives $\sigma_P(r,t)$0 and $\sigma_P(r,t)$1, while the flux-rope model gives $\sigma_P(r,t)$2 at the same $\sigma_P(r,t)$3 (Elekes et al., 18 Jun 2025).

4. Dissipation rates, event energetics, and annual averaging

For interior dissipation, the study extracts the time-varying surface Gauss coefficients $\sigma_P(r,t)$4 and $\sigma_P(r,t)$5, computes the induced subsurface $\sigma_P(r,t)$6–$\sigma_P(r,t)$7 response via the Grayver & Roberts (2022) induction model for $\sigma_P(r,t)$8, and integrates

$\sigma_P(r,t)$9

For a 1-hour CME with $E(r,t)=-v\times B$0 and $E(r,t)=-v\times B$1, the resulting single-event interior Joule heating is $E(r,t)=-v\times B$2 for Trappist-1b and $E(r,t)=-v\times B$3 for Trappist-1e. Taking CME occurrence rates into account places the annual average interior heating near the lower end of previous estimates (Elekes et al., 18 Jun 2025).

The annual average is obtained from a flare frequency distribution $E(r,t)=-v\times B$4 from Howard et al. (2023), the fraction $E(r,t)=-v\times B$5 of flares that launch CMEs toward the planet, and $E(r,t)=-v\times B$6 days,

$E(r,t)=-v\times B$7

with the sum performed over $E(r,t)=-v\times B$8. This gives $E(r,t)=-v\times B$9 for Trappist-1b and $J(r,t)=\sigma_P E$0 for Trappist-1e (Elekes et al., 18 Jun 2025).

Quantity	Representative value	Conditions
Interior Joule heating, Trappist-1b	$J(r,t)=\sigma_P E$1	1-hour CME, $J(r,t)=\sigma_P E$2, $J(r,t)=\sigma_P E$3
Interior Joule heating, Trappist-1e	$J(r,t)=\sigma_P E$4	Same
Annual average interior heating, Trappist-1b	$J(r,t)=\sigma_P E$5	$J(r,t)=\sigma_P E$6, $J(r,t)=\sigma_P E$7, $J(r,t)=\sigma_P E$8
Annual average interior heating, Trappist-1e	$J(r,t)=\sigma_P E$9	Same
CME ionospheric Joule heating, Trappist-1e	$V$0–$V$1	1-hour CME
Steady-state wind ionospheric heating	$V$2	$V$3
Stellar XUV dayside power	$V$4	$V$5

For the ionosphere, the average CME-driven heating is computed as

$V$6

On Trappist-1e, 1-hour CMEs yield average ionospheric heating rates of order $V$7–$V$8, decreasing from $V$9 at $$Q_{\rm int}=\int_V \eta |J|^2\,dV=\int_V \sigma |E|^2\,dV,$$0 to $$Q_{\rm int}=\int_V \eta |J|^2\,dV=\int_V \sigma |E|^2\,dV,$$1 at $$Q_{\rm int}=\int_V \eta |J|^2\,dV=\int_V \sigma |E|^2\,dV,$$2. Flux-rope CMEs give roughly a factor of 2 higher $$Q_{\rm int}=\int_V \eta |J|^2\,dV=\int_V \sigma |E|^2\,dV,$$3 than density-pulses (Elekes et al., 18 Jun 2025).

5. Atmospheric significance and the limits of magnetic shielding

The atmospheric significance of CME-driven Joule heating follows from its magnitude relative to background forcing. In a steady-state stellar wind, the ionospheric Joule heating is $$Q_{\rm int}=\int_V \eta |J|^2\,dV=\int_V \sigma |E|^2\,dV,$$4 for $$Q_{\rm int}=\int_V \eta |J|^2\,dV=\int_V \sigma |E|^2\,dV,$$5. The stellar XUV dayside power is also $$Q_{\rm int}=\int_V \eta |J|^2\,dV=\int_V \sigma |E|^2\,dV,$$6, using $$Q_{\rm int}=\int_V \eta |J|^2\,dV=\int_V \sigma |E|^2\,dV,$$7. CME ionospheric heating therefore exceeds both steady-wind heating and XUV irradiation by 1–2 orders of magnitude during an event (Elekes et al., 18 Jun 2025).

A common oversimplification is that stronger planetary magnetic fields always improve protection by reducing stellar-wind coupling. The Trappist-1 calculations do not support that statement in a universal form. Ionospheric Joule heating decreases with stronger $$Q_{\rm int}=\int_V \eta |J|^2\,dV=\int_V \sigma |E|^2\,dV,$$8, yet magnetospheric Poynting fluxes scale as $$Q_{\rm int}=\int_V \eta |J|^2\,dV=\int_V \sigma |E|^2\,dV,$$9, and stronger magnetic fields increase CME energy absorption. The same simulations therefore support two statements simultaneously: planetary magnetic fields reduce direct impact of CME-embedded fields, and planetary magnetic fields enhance induction heating in the interior, although the effect is weaker with flux rope CMEs. This suggests that “shielding” is not a scalar property but a redistribution among impact channels (Elekes et al., 18 Jun 2025).

The likely atmospheric consequence is severe. The paper states that such intense, rapid heating can inflate the upper atmosphere and greatly enhance escape and erosion, and that the event-scale ionospheric heating rates might severely impact atmospheric erosion. By contrast, the interior response is energetically smaller but persistent on an annual basis, with $\eta(r)=1/\sigma(r)$0 for Trappist-1b and $\eta(r)=1/\sigma(r)$1 for Trappist-1e (Elekes et al., 18 Jun 2025).

6. Relation to solar-CME heating studies and acronym ambiguity

In solar-CME plasma studies, Joule heating usually refers to resistive dissipation within the eruptive structure or its current sheet rather than to energy deposition in a planetary obstacle. A three-dimensional CME simulation by Reeves et al. writes the ohmic contribution as $\eta(r)=1/\sigma(r)$2 in the internal-energy source term and finds that ohmic heating is the dominant source of thermal-energy input in the current sheet early, reaching localized rates of order $\eta(r)=1/\sigma(r)$3–$\eta(r)=1/\sigma(r)$4. After plasmoid formation, adiabatic compression overtakes Joule heating as the main energizer, while thermal conduction redistributes the heat and produces the broad “thermal halo” around the sheet; the net result is a temperature rise from $\eta(r)=1/\sigma(r)$5 up to $\eta(r)=1/\sigma(r)$6 and a density enhancement from $\eta(r)=1/\sigma(r)$7 up to a few $\eta(r)=1/\sigma(r)$8 in the reconnection region (Reeves et al., 2019).

Other solar-CME studies reach a related but more qualified conclusion. In an expanding spheromak model with anomalous resistivity, Joule heating reproduces moderate heating and final Fe charge states peaking between $\eta(r)=1/\sigma(r)$9 and $1\ \mathrm{TW}$00 in the flux-rope proper, but it cannot account for the rapid low-height heating seen by STEREO/EUVI or the highest in-situ $1\ \mathrm{TW}$01–$1\ \mathrm{TW}$02 states; those observations require reconnection or wave-particle processes not captured by the global spheromak prescription (Rakowski et al., 2011). A UVCS-based analysis of a 1999 May 17 CME similarly finds that the cumulative heating energy of the core material is comparable to its kinetic energy and that the MAS 3D MHD simulation’s heating rate agrees with observationally constrained heating rates when a quadratic power law is assumed to describe a self-similar CME expansion, or when the heating parameterization accounts for the CME flux rope’s magnetic energy being converted directly into thermal energy (Wilson et al., 2021).

The acronym “CME” is not unique. In hot and dense matter, “CME-driven Joule heating” can refer to the Chiral Magnetic Effect rather than to a coronal mass ejection. In that usage, the anomalous current is

$1\ \mathrm{TW}$03

and, for a strong externally maintained magnetic field $1\ \mathrm{TW}$04, the Joule-heating rate per unit volume becomes

$1\ \mathrm{TW}$05

That literature treats magnetized hot and dense media such as proto-neutron stars and neutron-star mergers, not stellar coronal-mass-ejection forcing of planets (Sen et al., 30 Sep 2025).