CN Evershed

1
CN Evershed Ti Polarimetry

• Wenda Cao (NJIT/ Yunnan), S. Walton, G.
  Chapman (CSUN/SFO), W. Livingston (NSO)

2
Thanks References

• Thanks to Claude Plymate, Eric Galayda, Carl
  Henney, Han Uitenbroek and Thomas Rimmele.
• Penn, M.J., Cao, W.D., Walton, S.R., Chapman,
  G.A., Livingston, W., Weak Infrared Molecular
  Lines Reveal Rapid Outflow in Cool Magnetic
  Sunspot Penumbral Fibrils '' ApJ Letters, 2003,
  in press
• Penn, M.J., Cao, W.D., Walton, S.R., Chapman,
  G.A., Livingston, W., Imaging
  Spectropolarimetry of Ti I 2231nm in a Sunspot ''
  Solar Physics, 2003, in press

3
Observations

• Last talk OH molecules and model for line
  strength
• This talk the other molecular lines (CN and
  unidentified) and a line from Ti I.

4
Summary

• CN line ( others) show rapid (6 km/sec)
  Evershed outflow with Doppler shifted line
  profiles.
• Ti line in sunspot fully split and gives
  interesting instrumental polarization check.
• Ti Zeeman splitting gives strange B vs T
• Weak unidentified molecular lines are similar to
  OH (FeH) and CN in EqW vs T.

5
Molecular Evershed Flow

• Why use molecules to study penumbra?
• Fe I lines at 1558nm show fast (supersonic)
  downflows of 16 km/sec at temps only 300K hotter
  than umbra (del Toro Iniesta, Bellot Rubio
  Collados 2001).
• Fe I 1565nm penumbral line profiles can be fit
  with linear combination of quiet Sun and umbral
  profiles (Solanki, Montavon, Livingston 1994).
• CN line has peak equivalent width at penumbral
  temperatures

6
Molecular Evershed Flow
7
Molecular Evershed Flow
8
Molecular Evershed Flow
9
Molecular Evershed Flow

• Spot NOAA 10008 observed at 1565nm 2002 June 21,
  24, 27 at r0.65 (east), 0.27 and 0.66(west) from
  disk center.
• Spots NOAA 9887 and 9888 observed on 2002 April 1
  show similar outflows and velocities at 1565nm
  other scans from CSUN/SFO from 2002 can be
  analyzed.

10
Molecular Evershed Flow
11
Molecular Evershed Flow
12
Molecular Evershed Flow

• 2231nm spectra

13
Molecular Evershed Flow
14
Molecular Evershed Flow

• CN 1565nm line shows absorption shifted from 0 to
  9 km/sec, with 6 km/sec horizontal
  characteristic speed
• Molecular line at 2231nm shows from 2 to 9 km/sec
  horizontal
• Distribution may represent a range of
  inclinations or real speed differences within the
  azimuthal bin

15
Molecular Evershed Flow

• What height is CN formed? z 30km

16
Molecular Evershed Flow

• Height dependence v(h)2.5-(2.5/550)h or
  v(h)4.5e -(h/120) from Doppler shifts or line
  asymmetries of atomic lines (Stellmacher Wiehr,
  1980).

17
Molecular Evershed Flow
18
Molecular Evershed Flow
19
Molecular Evershed Flow

• Outflow speeds agree with line asymmetry function
  from Stellmacher Wiehr 80.
• Roughly agree with Shine et al. 94, Rimmele 94,
  Wiehr 95 and Schlichenmaier Schmidt 00 who
  measure speeds from 3.7 to 7 km/sec using a
  variety of methods.
• Unclear if modern observations support early
  velocity vs. height functions.

20
Molecular Evershed Flow
21
Ti 2231nm data

• Intensity spectra from NOAA 10008 near west limb
  29 June 2002 shows only Ti, molecular lines and
  atmospheric lines there are no photospheric
  lines seen.

22
Ti 2231nm data
23
Ti 2231nm data

• Ti eq. width agrees with Ruedi et al. (1998) RT
  model with TiO formation

24
Ti 2231nm data

• B vs T relationship using Ti line (29 June)
  compared to Fe 1565nm measurements in same spot
  (27 June) present difficulties
• 2231nm continuum formed higher than 1565nm
  continuum so should be cooler (100K cooler)
• Ti formed higher than Fe so should be weaker B,
  assuming vertical gradient (should be 200G
  weaker)
• T is lower by 300K
• B is greater by 100G

25
Ti 2231nm data
26
Ti 2231nm data

• Bifurcation seen from 4200-5000 the upper branch
  corresponds to region of steep intensity gradient
  and is overestimation of temperature due to blur.

27
Ti 2231nm data
28
Instrumental Polarization

• The fully resolved Ti I 2231nm Stokes V profiles
  in umbra show strong crosstalk since they show an
  unshifted central p component.

29
Instrumental Polarization

• Without Mtel or knowledge of full Stokes vector a
  proper solution for Vtrue is not possible.
• However, assuming mostly U to V crosstalk (M34
  dominates) and by constructing a proxy for Utrue
  a solution can be attempted.

30
Instrumental Polarization

• At each spatial position a value for M34 (M44 set
  to 1.0) can be computed and Vtrue calculated

31
Instrumental Polarization

• A proxy for Utrue is constructed using the
  intensity profile
• Red and blue wing p components of Stokes I
  profile are multiplied by -1.
• Ratio between central s component and p
  components from Stokes I preserved for Stokes U
  proxy (this assumes field is tangent to line of
  sight Sakurai (2000))

32
Instrumental Polarization

• M34 is determined at each spatial position by
  ratio of central V to U

33
Instrumental Polarization

• The Stokes V p component is then determined
  although no symmetry is required, the resulting
  profiles are very symmetric

34
Instrumental Polarization
35
Ti 2231nm data

• Integrated Stokes V shows B reversal between limb
  and center penumbra

36
Ti 2231nm data

• Systematic horizontal umbral Doppler flow limit
  of lt200 m/sec (inward or outward)

37
Ti 2231nm data

• Ruedi et al. (1999) report a discontinuity in
  magnetic field, with the penumbral field much
  weaker than umbral field from inversion of NIM
  observations of Ti 2231nm Stokes V.

38
Ti 2231nm data

• Such dual B spectra are not seen, but a
  discontinuity can be manufactured from the
  Evershed Doppler shift of the Stokes V profile

39
Ti 2231nm data
40
Ti 2231nm data

• With spatial averaging a triple-peaked Stokes V
  profile can be produced due to the Evershed
  Doppler shift, but a four-peaked profile cannot
  be produced.

41
Ti 2231nm data

• Azimuthal averaging of the Stokes V spectra show
  Evershed sinusoid as well as limb reversal of
  magnetic sign plus other features that arent
  understood.

42
Ti 2231nm data
43
2231nm molecular lines

• There are four very weak molecular lines which
  can be measured near the Ti I 2231nm line in the
  sunspot and surrounding region.
• 2230.67 (1.5 mA max)
• 2230.77 (1.5 mA)
• 2231.70 (3 mA)
• 2232.21 (0.8 mA).

44
2231nm molecular lines

• The 2230 - 2231 lines behave like OH or FeH in
  their temperature dependence

45
2231nm molecular lines

• The 2232.21nm line behaves more like CN

46
2231nm molecular lines

• These lines need to be identified.

47
Summary

• CN line ( others) show rapid (6 km/sec)
  Evershed outflow with Doppler shifted line
  profiles.
• Ti line in sunspot fully split and gives
  interesting instrumental polarization check.
• Ti Zeeman splitting gives strange B vs T
• Weak unidentified molecular lines are similar to
  OH (FeH) and CN in EqW vs T.