Transcript Search for magnetic fields at the surface of Mira stars

Search for magnetic fields
at the surface of Mira stars
Agnès Lèbre (LUPM, Fr)
Michel Aurière & Pascal Petit (IRAP, Fr)
Nicolas Fabas (IAC, Sp)
Denis Gillet (OAMP, Fr)
Fabrice Herpin (LAB, Rr)
Renada Konstantinova-Antova (BAS, Bu)
Mira stars are cool and evolved pulsating stars,
belonging to the tip of the Asymptotic Giant Branch
(AGB), the key evolutionary stage of an intermediate
mass star before its transition toward the Planetary
Nebulae (PNe) stage.
A radially pulsating star…
… with a very extended
atmosphere
Magnitude
period ~ 1 yr
Time
Spectral types :
M (C/O < 1)
C (C/O > 1)
S (C/O ~ 1)
From J. Hron
M57
The
Ring
Nebula
The Eskimo Nebula
NGC6302
The Bug nebula
HST images
The Egg Nebula
IRAS 13208-6020
IRC 10216
R Scl
ALMA (ESO/NAOJ/NRAO)
Observational evidences for magnetic fields in PNe
and around their AGB and post-AGB progenitors
Magnetic fields
throughout the CSE of AGB
Detected and measured
from the polarisation of the
maser emission of
several molecules :
OH at 1000-10000 a.u.
B// ~ 5-20 mG
(Rudnitski et al., 2010)
Water at a few 100 a.u.
B// ~ a few 100 mG
(Vlemmings et al., 2005)
SiO at 5-10 a.u.
B// ~ 3.5 G
(Herpin et al. 2006)
From Reid & Menten, 1997
Magnetic field strength vs. radius relation
as indicated by current maser
polarization observation of a number of Miras
Extrapolating the 1/r law
toward the photosphere
→ the magnetic field
strength at the stellar
surface of Mira stars could
be of the order of a few G.
(From Vlemmings et al., 2011)
solar-type magnetic field config.
toroïdal
Detection with Narval
At the sub Gauss level !
Herpin et al., 2009
The Mira star χ CYG : Pulsating period ~ 400 days ; M = 2 M_sun
Spectral type : from S6 to S10 (presence of ZrO bands)
A magnetic field has already been detected in the inner part of its CSE :
Mean value of the magnetic field along the line of sight : B// = 0 - 8.8 G at 5-10 a.u.
(from SiO maser and Elitzur’s theory, Herpin et al. 2006 )
Max. light on
23 March 2012
174 Stokes V
15 to 28 Mar.12
Intensity (not normalized to the continuum)
BW Vul (B2)- ESPaDOnS
RR Lyr (F5) - ESPaDOnS
OMI Cet (M5) - Narval
Wavelength
LSD analysis with a specific numeric mask (Teff = 3500 K and log g = 0.5)
involving about 14 000 atomic lines
with atomic parameters and Landé factors from VALD
LSD analysis
174 V sequences
A LSD Zeeman
Signature ?
The result is (for the complete profile) :
Bl= -0.25 ± 0.40 G, Definite Detection (chi2=1.81 , fap=5.2 10-10 )
The weak Stokes V signal is present if we split our spectra in two series of equal
significance, and does not appear on the null polarization profiles.
Mask with lines with high landé factor
(mean=1.53)
Definite Detection
Bl= -0.13 ± 0.37 G.
(chi2=1.67, fap=9.3 10-08 )
Zeeman origin ?
The detection of the Stokes V
structure is more significant
when we use a mask with lines
of high Landé factor.
Mask with lines with low landé factor
(mean=0.90)
This supports a (magnetic)
Zeeman effect origin.
No Detection
Bl = -1.52 ±0.67 G.
(chi2=1.18, fap=6.4 10-02 )
The Stokes V signal
is associated to the blue
component of the I profile
The I profile presents
the typical line doubling
of metallic lines (due to the
presence of a shock wave
in the lower atmosphere).
The longitudinal magnetic field computed using the first-order moment method (Rees and Semel
1979) adapted to LSD profiles (Donati et al. 1997) and using the complete Stokes I profile :
Bl = - 0.25 ± 0.4 G → lower limit !
Scaling directly the Stokes V signal and the blue (I) profile to those which are observed in
the K0III star Pollux (classical Zeeman profile), the Chi Cyg detection in the blue profile
could correspond to the detection of a magnetic field with Bl of a few Gauss (2-3 G).
A link with the periodic shock wave ?
The Stokes V signature is associated to the blue component of the I profile,
i.e., to the material which is driven outward by the shock
→ is likely linked to the shock :
- either it may originate directly because of it, or
- it may be due to the amplification -by the shock- of a
weak photospheric magnetic field.
→ Monotoring of Chi Cyg with ESPaDOnS@CFHT
around its 2013 maximum light, focusing on specific
strategic phases:
- before the maximum light, when the shock wave has not
yet emerged from the photosphere;
- at the maximum light, when the star is the brightest and
its photosphere is the hottest;
- after the maximum light, when the shock wave has
emerged from the photosphere, and is accelerated in the
lowest part of the atmosphere.
Light curve of chi Cyg
The Mira star RT CYG : Pulsating period ~ 190 days
Spectral type : M2@max. light (presence of TiO bands)
20 Stokes V
Max. light on
10 Sept. 2012
Bl = -0.09 ± 0.79 G No Detection (20 V sequence)
11 sept. 2012
17 sept. 2012
No signal detected… so far !
We really need the
sub gauss level !
Conclusions
A weak magnetic field may exist at the surface of Mira stars.
It could play a part in the evolution of the morphology from AGB to PN
(besides the role of a companion, for example).
A Stokes V signal has been detected on chi Cyg, a bright S-type Mira,
when observed at its maximum light
However, the possible link with the atmospheric shock wave
needs to be investigated.
This represents the first
detection of a magnetic field
at the surface of a Mira star,
in complement to the
previous detections
of magnetic field
throughout the CSE.