Plasma Heating During Coronal Mass Ejections

Nicholas A. Murphy,1 Chengcai Shen,1 Remington Rimple,1,2

John C. Raymond,1 and Leonard Strachan3

1Harvard-Smithsonian Center for Astrophysics2California State University San Marcos

3Naval Research Laboratory

SHINE Conference 2016Santa Fe, New Mexico, USA

July 11–15, 2016

Introduction

I Our understanding of astrophysical phenomena begins withthe energy budget

I Magnetic energy dominates coronal mass ejections (CMEs)but is difficult to diagnose remotely

I CME kinetic and potential energies are estimated using whitelight coronagraphs (e.g., SOHO/LASCO)

I Estimates of the thermal and cumulative heating energyrequire non-equilibrium ionization (NEI) modeling

I We present work in progress to model the evolution of chargestates during CMEs to determine heating rates in eventsobserved by the Ultraviolet Coronagraph Spectrometer(UVCS) on SOHO

The 2010 Nov 3 CME observed by SDO/AIA showsevidence for a hot core

I Emission in 94 & 131 channels not present in cooler channelsindicates a core temperature of 5–10 MK and early heating

Reeves & Golub (2011)

Key questions

I How much is CME plasma heated?

I What are the spatial and temporal dependences of heating?

I What physical mechanisms are responsible for CME heating?

I Where does the energy for CME heating come from?

I What are the consequences of CME heating on magneticcloud propagation and space weather?

I Are some parts of CME plasma not heated?

Candidate CME heating mechanisms (e.g., Murphy et al.2011)

I Magnetic reconnection in the CME current sheet

I Relaxation and reconnection inside expanding flux rope

I Dissipation of Alfven waves and turbulenceI Collisions between thermal plasma and flare-accelerated

electronsI 2010 Nov 3 event described by Glesener et al. (2013)

I Shocks

There are three main strategies for observationallyconstraining plasma heating in CMEs

I Ultraviolet spectroscopy1

I Observations from the low corona to heights of a few solar radiiI Usually requires NEI forward modelingI Example instruments: SOHO/UVCS, Hinode/EIS

I In situ charge state observations at 1 AU2

I Requires NEI modelingI Example spacecraft: ACE, Wind, STEREO, DSCOVR

I Multiwavelength EUV and X-ray imaging3

I Observations at the low coronaI Example instruments: SDO/AIA, Hinode/XRT, SOHO/EIT,

STEREO/EUVI, PROBA2/SWAPI Sometimes requires NEI modeling

1Akmal et al. (2001); Landi et al. (2010); Murphy et al. (2011)2Rakowski et al. (2007, 2011); Lepri et al. (2012)3Cheng et al. (2011); Nindos et al. (2015)

Non-equilibrium ionization modeling is required whenionization/recombination timescales . expansion timescale

I The evolution of charge states in an NEI plasma is given by

dfidt

= ne [Ci−1fi−1 − (Ci + Ri ) fi + Ri+1fi+1] (1)

where fi is the ion fraction, ne is the electron density, and Ci

and Ri are the ionization and recombination rate coefficientsfor an ion with charge state i

I The thermodynamic history of NEI plasma is encoded in thecharge state distributions

I Errors associated with NEI modeling include uncertainties inatomic rates (∼10–30%) and the assumption of a Maxwelliandistribution without energetic particles

SunNEI: a non-equilibrium ionization (NEI) pythonpackage in development

I Uses the eigenvalue method to solve NEI equations (Masai1984; Hughes & Helfand 1985; Smith & Hughes 2010)

I Reduces the NEI differential equations to matrix multiplicationand exponential calculations

I Inherently stable at long time steps

I The Python implementation is based off of Fortran routinesby C. Shen et al. (2015)

I The Fortran implementation is useful for computationallyintensive investigations and is available athttps://github.com/ionizationcalc/time dependent fortran

I Example: post-processing analysis of a 3D MHD simulation ofthe solar wind (C. Shen et al., submitted)

I The Python implementation allows easier analysis of 1Dmodels that do not require significant computing time

I Example: a grid of hundreds of 1D modelsI In development at: https://github.com/namurphy/SunNEI

SunNEI: CME heating module

I The velocity evolution is given by

V (t) = V∞(

1 − e−t/τ)

(2)

where V∞ is the final velocity and τ is the accelerationtimescale

I The density evolution is given by

n

n0=

(h

h0

)α(3)

where n0 is the number density of neutral and ionizedhydrogen at initial height above the photosphere h0

I The temperature evolves due to adiabatic expansion andradiative cooling.

I Next step: implementing different heating parameterizations

I Next slide: initial & final charge states for V∞ = 800 km s−1,n0 = 109 cm−3, T0 = 106.4 K, α = −2, and no heating

0 10.0

0.2

0.4

0.6

0.8

1.0Io

niz

ati

on f

ract

ions

H

0. 1R¯5R¯

0 1 20.0

0.2

0.4

0.6

0.8

1.0

Ioniz

ati

on f

ract

ions

He

0 1 2 3 4 5 60.0

0.2

0.4

0.6

0.8

1.0

Ioniz

ati

on f

ract

ions

C

0 1 2 3 4 5 6 70.0

0.2

0.4

0.6

0.8

1.0

Ioniz

ati

on f

ract

ions

N

0 1 2 3 4 5 6 7 80.0

0.2

0.4

0.6

0.8

1.0

Ioniz

ati

on f

ract

ions

O

0 1 2 3 4 5 6 7 8 9 100.0

0.2

0.4

0.6

0.8

1.0

Ioniz

ati

on f

ract

ions

Ne

0 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

Ioniz

ati

on f

ract

ions

Mg

0 1 2 3 4 5 6 7 8 9 10 11 12 13 140.0

0.2

0.4

0.6

0.8

1.0

Ioniz

ati

on f

ract

ions

Si

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 160.0

0.2

0.4

0.6

0.8

1.0

Ioniz

ati

on f

ract

ions

S

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 180.0

0.2

0.4

0.6

0.8

1.0

Ioniz

ati

on f

ract

ions

Ar

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 200.0

0.2

0.4

0.6

0.8

1.0

Ioniz

ati

on f

ract

ions

Ca

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 260.0

0.2

0.4

0.6

0.8

1.0

Ioniz

ati

on f

ract

ions

Fe

Strategy for constraining heating rates for an NEI plasma

I Choose CMEs with appropriate observational constraints likeI SOHO/UVCS observations at several solar radiiI In situ observations at 1 AU

I Perform a grid of models based on the CME’s observedproperties (velocity profile, inferred density, etc.) and usedifferent heating rates and parameterizations

I Find the models that are consistent with the observedspectrum or charge state distributions

I The remaining models will show the allowed heating rates foreach parameterization

Next steps

I Code development: add heating parameterizations andprediction capabilities for UVCS and AIA observations

I Perform a baseline study for models with no heatingI REU project underway by Remi Rimple, who is planning to

present results at the AGU Fall Meeting

I Constrain heating rates for three events for which the sameplasma was observed by UVCS at multiple heights

I This analysis should provide better constraints on plasmaheating further from the eruption site

I Long-term goals include non-Maxwellian distribution functioncapabilities (e.g., Dzifcakova et al. 2015) and photoionization(e.g., Lepri & Landi 2015)

Summary

I Plasma heating is an important component of CME energybudgets

I Diagnosing plasma heating often requires non-equilibriumionization modeling of the erupting plasma

I We are developing a Python implementation fornon-equilibrium ionization to be applied to CMEs

I We are analyzing three events observed by SOHO/UVCS atmultiple heights to better constrain continued heating afterthe plasma leaves the eruption site