PosterPDF Available

Combining MinXSS and RHESSI X-ray Spectra for a Comprehensive View of the Temperature Distribution in Solar Flares

Authors:

Abstract and Figures

Solar flares explosively release large amounts of magnetic energy, a significant fraction of which goes into transient heating of coronal plasma to temperatures up to tens of MK. Decades of observations have shown that flares are multi-thermal, exhibiting broad temperature distributions or "differential emission measures" (DEMs). Recent studies suggest that the hottest parts of the DEM evolve differently from, and are heated by a different physical mechanism than, the DEM bulk. For example, the peak temperature of the hot, likely in-situ-heated plasma observed by RHESSI correlates significantly differently with flare intensity (GOES class) than does the cooler, likely chromospherically evaporated plasma observed by GOES XRS and/or Yohkoh BCS. These studies, however, used discrete (iso-/bi-) thermal approximations, in part because temperature determinations by the ratio of 2-channel GOES photometer data or selected BCS lines necessitated such methods. Consequently, the exact DEM profile, its evolution, and how these correlate with other flare parameters, remain poorly known. The MinXSS CubeSat deployed from the ISS in May 2016, and since June has observed (at least) 7 M-class and over 40 C-class flares. MinXSS's X123 spectrometer measures solar soft X-rays (SXRs) from ~0.5 to ~30 keV with ~0.15 keV FWHM resolution; this energy range entirely covers both GOES XRS passbands, and overlaps with and extends the RHESSI observing range with ~5x better resolution. It includes the thermal continuum emission from plasmas with temperatures down to ~2 MK, as well as a number of mid- and high-temperature spectral lines from various low- and high-FIP ion species, providing critical temperature diagnostics for studying flare DEMs with far greater fidelity than is possible with GOES, or using RHESSI alone. We present spectral analyses of several flares observed simultaneously by MinXSS and RHESSI. We compare and contrast the observations of each instrument separately, and present the results of a joint-instrument DEM analysis that forward-fits a parametrized DEM model -- including variable elemental abundances -- to the combined spectra of both instruments simultaneously. We discuss the DEM evolution and its correlation with other flare parameters, and discuss the implications for plasma heating in solar flares.
No caption available
… 
Content may be subject to copyright.
Combining MinXSS and RHESSI X-ray
Spectra for a Comprehensive View of the
Temperature Distribution in Solar Flares:
First Steps
Amir Caspi1†, James M. McTiernan2,
James P. Mason3, Christopher S. Moore3,
Albert Y. Shih4, Harry P. Warren5, Thomas N. Woods3
1 Southwest Research Institute, Boulder amir@boulder.swri.edu
2 University of California, Berkeley; 3 University of Colorado, Boulder
4 NASA Goddard Space Flight Center; 5 Naval Research Laboratory
SH13A-2288
MinXSS SXR Observations
MinXSS covers a crucial observational gap from ~0.2 to
~3 keV (~0.4 to ~6 nm) that has had very few spectrally-
resolved observations in previous decades (Fig. 1, left)
This range is:
Rich with medium- and high-temperature line and
continuum emission for diagnostics of coronal plasma
temperatures and elemental abundances (Fig. 1, right)
Extremely sensitive to plasma temperature
Not photon-starved – high photon fluxes allow good
statistics from small apertures
2
MinXSS SXR Observations (cont.)
But:
MinXSS’s spectral response is optimized for sensitivity
<10 keV, thus for temperatures of ~2–15 MK
Solar flare temperatures can reach ~30–50 MK, also with
significant non-thermal emission, dominating >10 keV
No single spectrometer currently observes the entire ~1–
100 keV range, to sample the full range of temperatures
in flares, ~2–50 MK
3
Fig. 1: (Left) Timeline of spectrally resolved soft (SXR) and hard (HXR)
X-ray observations since 1976 (non-exhaustive). MinXSS observes up to
~30 keV, but is optimized for ~1–10 keV. (Right) The SXR range is
extremely temperature-sensitive and includes both continuum emission
and numerous spectral lines from various ions.
4
FlareQuiescent
CubIXSS
Yohkoh HXT (4 ch)
MinXSS-1
MinXSS-2 CubIXSS
(proposed)
Multi-Instrument Temperature Coverage
Caspi et al. (2014; ApJL 788, L2) combined SDO/EVE
EUV spectra and RHESSI HXR spectra to calculate joint
differential emission measure (DEM) distributions
covering the entire ~2–50 MK range in flares
But:
EVE MEGS-A data ends May 2014
Cross-calibration with X-ray data is difficult and strongly
model-dependent (direct comparison not possible)
EUV hot coronal lines are plentiful but strongly blended,
including with cool chromospheric lines (Fig. 2, top)
EUV lines may suffer from non-equilibrium ionization
EUV continuum emission is weak and difficult to analyze5
Multi-Instr. Temp. Coverage (cont.)
MinXSS SXR observations are sensitive to a similar
temperature range as SDO/EVE (Fig. 3)
But:
SXR continuum immune to non-equilibrium ionization
SXR line blends are generally from ion species with
similar temperature sensitivities (e.g., Caspi et al. 2015;
ApJL, 802, L2; see also Fig. 2, right)
MinXSS data available from June 2016 to present,
including multiple strong (M-class) flares
Cross-calibration of MinXSS SXR and RHESSI HXR data
much more straightforward and well defined
6
Fig. 2: (Top) Blending of hot coronal
and cool chromospheric spectral lines
complicates analysis of EUV spectra,
particularly for joint DEM
reconstruction with RHESSI.
(Right) SXR spectra are sensitive to
similar temperatures as EUV, but with
fewer difficulties stronger
continuum, “simpler” line blends, and
easier cross-calibration with HXRs. 7
0.000
0.005
0.010
0.015
0.020
Irradiance (erg cm2 s1 Å1)
EVE Observation
DEM MODEL
χ2 = 4.71
100 150 200 250 300 350
Wavelength (Å)
10
5
0
5
10
I/σ
6.0 6.5 7.0 7.5 8.0
Log T (K)
1
10
100
1000
DEM (1040 cm3 K1)
Recovered DEM
Total χ2 = 5.27
0
50
100
150
Intensity (count s1)
RHESSI Obs
DEM MODEL
χ2 = 0.57
A0: 4.79
Ebr: 10.16
γ: 5.11
Fe: 5.79
FeNi: 4.87
0 20 40 60 80
Energy (keV)
10
0
10
I/σ
(Caspi et al. 2014)
Example blend:
Fe XXIV (255 Å; ~20 MK)
He II (256 Å; ~0.01 MK)
X123 rocket spectrum, 2012-Jun-23 @ 19:33 UT
Energy [keV]
102
104
106
108
1010
Flux [photon s-1 cm-2 keV-1]
0.5 1.0 2.0 5.0
Observed: Full
DEM model = ~1.82–10.7 MK @ 𝜏 ≈ 6.31
(5.0×1043cm–3 keV–1 @ 23.2 MK)
Low-FIP mult: 0.8 × coronal
Observed: Henke
XPS L4 Model
Model (DEM, χ2 = 6.548)
X123 rocket spectrum, 2013-Oct-21 @ 18:03 UT
Energy [keV]
0.5 1.0 2.0 5.0
Observed: Full
DEM model = ~1.94–15.0 MK @ 𝜏 ≈ 6.04
(5.0×1044cm–3 keV–1 @ 23.2 MK)
Low-FIP mult: 0.4 × coronal
Observed: Henke
XPS L4 Model
Model (DEM, χ2 = 23.086)
Mg XI
Mg XII
Si XIII
Si XIV
Si XII d / XIII
Si XIV S XIV d / XV
S XVI
S XIV d / XV
S XV / XVI
Ar XVI d / XVII
Ar XVIII
Ar XVII
Ca XVIII d / XIX
X123 rocket spectrum, 2012-Jun-23 @ 19:33 UT
Energy [keV]
102
104
106
108
1010
Flux [photon s-1 cm-2 keV-1]
0.5 1.0 2.0 5.0
Observed: Full
DEM model = ~1.82–10.7 MK @ 𝜏 ≈ 6.31
(5.0×1043cm–3 keV–1 @ 23.2 MK)
Low-FIP mult: 0.8 × coronal
Observed: Henke
XPS L4 Model
Model (DEM, χ2 = 6.548)
X123 rocket spectrum, 2013-Oct-21 @ 18:03 UT
Energy [keV]
0.5 1.0 2.0 5.0
Observed: Full
DEM model = ~1.94–15.0 MK @ 𝜏 ≈ 6.04
(5.0×1044cm–3 keV–1 @ 23.2 MK)
Low-FIP mult: 0.4 × coronal
Observed: Henke
XPS L4 Model
Model (DEM, χ2 = 23.086)
Mg XI
Mg XII
Si XIII
Si XIV
Si XII d / XIII
Si XIV S XIV d / XV
S XVI
S XIV d / XV
S XV / XVI
Ar XVI d / XVII
Ar XVIII
Ar XVII
Ca XVIII d / XIX
(Caspi et al. 2015)
Example blend:
Si XI (303 Å; ~1.6 MK)
He II (304 Å; ~0.01 MK)
Fig. 3: Temperature coverage of Fe ion species observed by SDO/EVE;
above ~20 MK, EUV lines are scarce and EVE is not sensitive. MinXSS
is sensitive from ~2 to >15 MK via observations of SXR bremsstrahlung
continuum and lines from Mg XI, Si XII, S XIV, Fe XXV, and others (see
Fig. 2). RHESSI continuum and Fe XXV line observations extend
sensitivity from ~10 to >50 MK. Together, MinXSS and RHESSI cover
the full range of coronal temperatures in solar flares.
8
many EUV lines in this range EUV not sensitive
(Fe ions except where noted)
RHESSI
MinXSS
Methodology
We picked 2 intense (M-class) flares well-observed by both
MinXSS and RHESSI. As a first cut at comparing the two
instruments, for each flare, we:
Picked one time interval for comparison
Determined incident photon spectrum (instrument
response deconvolved) for each instrument
Fit dual-isothermal model to RHESSI (similar pre-
processing as Caspi et al. 2014), and single or double
isothermal to MinXSS (similar to Caspi et al. 2015)
Compared “unaltered” observations and photon models
9
10
11
12
13
Discussion
NOTE: All calibrations are preliminary. MinXSS data is
Level 1, version 1. RHESSI uses detector 3 only, with
calibration modifications from Caspi & Lin (2010).
Nominal RHESSI detector efficiency is assumed – we do
not yet account for possible efficiency loss from radiation
damage. No attempts have yet been made to inter-
calibrate the two instruments – that is the next step.
With that said…
For one flare, MinXSS and RHESSI agree within ~30%,
within their respective uncertainties.
For the other flare, agreement is off by a factor of ~10,
although overall spectral shape is nearly identical.
14
Discussion (cont.)
Cause of disagreement is unclear – potentially livetime
due to radio or other low-energy noise (no livetime
correction is currently applied to MinXSS)
Correction currently in progress
Excellent agreement for one flare, and spectral shape
agreement for the other, is encouraging for future cross-
calibration and joint DEM calculations.
15
Next Steps
MinXSS corrections to be implemented, including
livetime and pulse pileup
Additional data in-fill required; L1 data not currently complete
RHESSI corrections to be implemented, including
efficiency loss, attenuator corrections for detector 3
Cross-calibration of both instruments to yield single joint
spectrum from ~0.5–100 keV
Implementation of joint DEM calculations, and extension
of SXR spectra to ~0.1 keV from model fits
16
Downloads:
MinXSS science data
MinXSS first results paper
(in press, accepted today!)
17
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.