Towards Understanding the Solar Atmosphere

1
Towards Understanding the Solar Atmosphere

• by
• G. A. Doschek
• presented at the 6th Solar-B Science Meeting
• 8-11 November 2005
• Kyoto, Japan

2
Fundamental Questions 1- What drives solar
activity? 2- What structures and heats the solar
atmosphere? 3- How does solar energy affect the
heliosphere and the Earths atmosphere?
1 x 105 K
5,800 K
2.0 x 10 K inner corona
2.0 x 106 K outer corona
2.0 x 106 K solar wind
3
The Connectivity of Different regions of the
Solar Atmosphere

• The transition region is unresolved
• Spicule-like structures are seen on the limb
• The disk features vary approximately from
  loop-like structures to small spots
• The actual emitting regions are only a few tenths
  of an arc-second in size
• How does the lower transition region (T lt 2.0 x
  105 K) interface with the upper transition region
  and corona (2 x 105 K 106 K)?
• Cool loops and coronal funnels?
• Explanation of differential emission measure
  distribution
• In this work I approach this question by
  attempting to correlate upper and lower
  transition region dynamics, specifically flows,
  by analyzing spectra from SUMER/SOHO. If there
  is connectivity between the upper and lower
  transition region, then in the simplest case
  radial flows in each region should be in the same
  direction. This is ongoing work, just recently
  begun.
• C IV (1 x 105 K), S V (1.6 x 105 K), and Ne VIII
  (6.3 x 105 K) lines appear on a single SUMER
  wavelength window, assuring simultaneous and
  perfectly spatially aligned observations
• Ne VIII images still show the network, allowing
  at least an approximate association with lower
  transition region features.

4
The Magnetic Structure of the Solar Atmosphere
5
A Model of the Transition Region/Corona Interface
This schematic illustrates how the standard
model can explain the appearance of solar
structures at different temperatures
6
Average Temperature-Density Structure of the
Quiet Sun Chromosphere, Transition Region, and
Corona
The standard model of the solar atmosphere.
Does it apply to all solar structures? The
conductive flux makes the transition region
extremely thin Fc aT5/2 dT/dh
Ne VIII
S V
C IV
7
Differential Emission Measure of the Quiet Sun
Ne VIII (5.8)
C IV
8
Model of Sub-Resolution Filamentary Structure of
Transition Region Spicules at 105 K
Solar spectra indicate that the transition region
contains fine structures with dimensions ranging
from small values of about 1 (430 miles) down to
very small values of about 0.043 (19 miles).
Modeling of this fine-structure is now a hot
topic in solar physics.
9
How a Stigmatic Spectrometer Works
NRL-NASA/HRTS rocket spectrum -C IV ( 1549 Å).
C IV is formed at 105 K in the solar atmosphere.
EIS will obtain similar spectra with its narrow
1 and 2 slits. Images can be built up by
rastering.
Wavelength
10
SUMER/SOHO Image and Spectra in the Light of O
VI (3.2 x 105 K), N V (2.0 x 105 K), and O V (2.5
x 105 K)
The O IV image was obtained at Sun center and
covers an area of about 4 x 1010 km2. Each pixel
has an area of 5 x 105 km2 and contains a
measurement of both bulk and random plasma
motion. This image contains about 80,000
measurements of dynamical parameters. Images in
other spectral lines formed at different
temperatures are available for comparison and
correlation.
CAT scan-like slice through the transition region
in light of N V and O V.
11
Optical Depth in Spectral Lines

• Optical depth modifies line profiles by
• resonance scattering
• The shape of the line profile allows an
• estimate of the true path length of emitting
• plasma in the solar atmosphere along a
• given line of sight.
• The actual line shape depends on details
• of the geometry.
• Studies of line profiles in the lower transition
• region indicate path lengths of only 50-
• 100 km, or about 0.1 arcsec.

Some details for lovers of radiative
transfer tau F(Te) A Ne L f fraction of
total opacity in which the source function
is zero Dashed profiles are optically thin lines
Gaussians
O III line profiles
12
O II 834.465 Å opacity/km 1.08 x 10-2 O III
833.749 Å opacity/km 8.22 x 10-4
13
SUMER/SOHO 23 March 1996, about 19 hr UT 69 E
268 W 50 N 350 N 297 spectra, 21 s
exposures Slit 1 x 300 Scan East to West in
steps of 1.125. Center of slit 200 North of
equator.
14
EIT - 171
EIT - 195
EIT - 304
EIT - 195
15
(No Transcript)
16
Analysis Procedure

• The three spectral lines are C IV 1548.21 Å (1.0
  x 105 K), S V 1572.98/2 Å (1.6 x 105 K), and Ne
  VIII 1540.85/2 Å (6.3 x 105 K).
• Sum the Ne VIII, S V, and C IV line profiles for
  all the spectra in the three images, weighted by
  the total line intensities. There is one
  spectrum per line per 1x1 pixel. Each image
  contains 93,852 spectra, of which 17,292 are
  statistically useful.
• Determine by eye the central wavelengths of the
  three lines in the summed spectra. This is a
  somewhat arbitrary reference wavelength for
  defining redshifts and blueshifts.
• Fit the three line profiles in each individual
  spectrum with Gaussian profiles, subject to the
  condition that the total intensities of all three
  lines exceed a certain value. This is done to
  ensure reasonable statistics. Unfitted spectra
  due to not meeting this condition are defined as
  having zero Doppler shifts.
• The wavelengths of the lines determined by the
  Gaussian fits are subtracted from the reference
  wavelengths to yield the Doppler shifts.

17
Summed weighted spectra for the raster images.
The vertical lines are the central
wavelengths estimated by eye.
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
r
22
(No Transcript)
23
(No Transcript)
24
dashed profile Gaussian fit
shift 6.3 km/s
25
S V
C IV
Doppler shifts in Å
Ne VIII
Downflow in Ne VIII (6.3 x 105 K) and S V (1.6 x
105 K). There is no evidence of a downflow in C
IV (1.0 x 105 K). The downflow appears to be at
the footpoint of what could be described as a
plume in Ne VIII. A similar downflow appears in
another Ne VIII plume-like feature. The
down- flow is quite small, about 6.3 km/s.
26
Doppler Shifts in Angstroms
C IV
S V
C IV
The C IV spectrum is one of the spectra that pass
through the upflow region, showing that it is a
real feature, often seen in HRTS spectra.
Ne VIII
C IV
0.1 Å 20 km/s at 1548 Å 0.17 A 32 km/s
27
Hist. C IV Solid S V Dashed Ne VIII
0.17 Å 33 km/s
28
Doppler shifts in Angstroms
C IV
S V

• The entire image only
• data with total line intensities greater than 100
  counts are
• considered. Ne VIII 0.05 Å is a result of
  changing the
• reference Ne VIII wavelength by 0.05 Å.

Ne VIII
Ne VIII 0.05 Å
29
S V C IV
S V Ne VIII
C IV Ne VIII
Doppler shift differences Note that perfect
agreement between Doppler shifts gives zero net
differences. Only S V C IV approximates
this result.
30
Conclusions The Lower Transition Region

• The lower transition region is unresolved (at the
  best orbiting spacecraft data now available (1-2
  arcsec imaging).
• Different temperature regions are correlated in
  intensity with correlation coefficients of 0.9
• Non-thermal mass motions correlation coefficients
  are 0.5-0.8
• Path lengths from optical depth estimates are
  50-100 km
• High spatial spectrometers like VAULT are needed
  to make further progress in understanding the
  smallest structures in the transition region.
  SUMER images show many elongated structures,
  suggesting that cool loop models are indeed at
  least partly applicable to the real Sun.
• Future Directions Continue to attempt to use
  SUMER spectra to determine the percentage of cool
  loops relative to classical transitions regions
  in the solar atmosphere analyze results from
  continuing VAULT rocket program use He II and Si
  VII lines in EIS Solar-B to attempt a similar
  analysis.

31
Solar Flare Science Issues

• Solar flare flux tubes at multimillion degree
  temperatures do not have the appearance expected
  from 1D numerical simulations. Why? X-ray
  spectral line profiles do not show evidence for
  expected chromospheric evaporation. Why not?
  What are the implications for the standard
  reconnection models of flares?
• X-ray images of solar flares at multi-million
  degree temperatures can show surprising
  morphological complexity that is difficult to
  interpret. Not all flares look like simple loop
  arcades (Masuda Bastille Day events) or
  long-duration events (Tsuneta flare).
• What are the physical conditions in the
  reconnection region above the soft X-ray flare
  loops?
• Can EIS see hard X-ray emission as well as the
  soft X-ray signatures?

32
Ca XIX Resonance Line X-ray Spectra
33
A Multi-Thread Flare Loop Model

• The overall soft X-ray magnetic envelop is
  assumed to be composed of sub-resolution magnetic
  threads.
• The threads are modeled as individual flare loops
  using the NRL 1D solar flux tube model.
• The flare onset is modeled as a succession of
  independently heated threads.
• The length of the flare loop is determined by
  observations of the overall magnetic envelop.
• The energy deposited in each thread and its
  cross-sectional area are related to the GOES
  fluxes (Warren Antiochos 1994). The energy in
  each successively heated thread is determined
  such that the X-ray flux matches the GOES light
  curves.
• The BCS spectrometers on Yohkoh serve as a
  completely independent test of the model.
• The BCS data support the multi-thread model
  (Warren Doschek 2005, ApJ, 618, L157).

34
Multi-Thread Results for Yohkoh/BCS

• Observe evolution and distribution of
  multi-temperature plasma
• Requires images that span 104 107 K.

From Warren Doschek, ApJ, 2005
35
EIS The Use of Slits and Slots
Solar flare time sequence

Skylab flare spectral images and EIS 40 wide slot
Slit spectra give line profiles
36
Images of Multi-million Degree Flare Loops

• Do flare loops at temperatures of 12-25 MK look
  like what we expect???
• No, they dont, but as they cool to 1-3 MK they
  look more and more like respectable 1D loops
  should.
• TRACE observations confirm the gross flare loop
  morphology seen by Yohkoh.

37
Solar Flare Reconnection Model
Solar flare 45,000 km
This schematic flare model provides theoretical
guidance for analyzing solar flare data.
38
Other Objectives Reconnection Region

• Observe motions of ejecta and inflows above
  arcade
• Requires cooler lines and longer exposures
  summing of images

From Shibata et al., ApJ, 1995
From Sheeley, Warren, and Wang, ApJ, 2004
39
Conclusions

• The first observed BCS Ca spectra are not
  blue-shifted in disk flares. However, the
  multi-thermal thread model can reasonably
  reproduce the observations. A sudden impact
  evaporation model cannot reproduce the
  observations.
• TRACE images indicate the presence of
  substructure (i. e., threads in soft X-ray flare
  loops (see Doschek, Strong, Tsuneta, 1995, ApJ,
  440, 370 for a discussion of Yohkoh flare loops).
• BCS Ca XIX/S XV initially observed electron
  densities are about 3.5 x 1010 cm-3 and emission
  measures are about 6 x 1046 cm-3, assuming the
  multi-thread scenario and a nominal electron
  temperature of 12 x 106 K (see Doschek Warren,
  ApJ, 629, 1150).
• The appearance of multi-million degree flare
  loops exhibits hot knots at the loop-tops and
  sometimes asymmetric loop brightening. The knots
  and asymmetric brightening are not projection
  effects and need to be explained in terms of our
  conventional views of plasma confinement in
  magnetic flux tubes.
• The soft X-ray flare onset for some events begins
  before the hard X-ray flare onset.
• EIS can use He II 256 Å emission as a hard X-ray
  proxy for footpoint emission.