Download Presentation file

LBVs, Hypergiants, and Impostors –
the Evidence for High Mass Loss Events
Roberta M Humphreys
University of Minnesota
Pacific Rim Conference on Stellar Astrophysics. Dec. 2015
The Upper HR Diagram
The evidence for episodic high mass loss events
VY CMa -- the extreme red supergiant, powerful OH/IR source
10”
Distance ~ 1.3 kpc
Luminosity ~ 3 x 105 L sun
Initial Mass ~ 30 M sun
Mass Loss rate 2 -- 4 x 10-4 M sun / year
Size
~ 8 -- 10 A.U., or ~ 1500 -- 2000 Rsun
It is visible as a small red nebula
~ 10 arcsec across
1” = 1500 AU
Due to multiple, asymmetric ejection
episodes possibly from large-scale
convective regions on the star.
Smith, Humphreys, Davidson, Gehrz, & Schuster, 2001
High Resolution, Long-Slit Spectroscopy
--Keck HIRES Spectrograph
2D spectra of strong K I emission lines across the arcs
NW Arc
Arcs 1 and 2
Second epoch HST images
Measure transverse velocities combined with radial velocities long
slit spectra (Keck) of K I em line (resonant scattering)
Total velocity, orientation, direction and age  3D morphology
Feature
NW arc
Arc 1
Vel.
Orientation
km/s relative to sky
46
22 degrees
68
-33
Direction Age (yrs)
of motion
~ west
500
SW
800
Arc 2
SW knots
S knots
SE loop
64
36
42
65
~ south
~ west
SSE
SE
-17
-25
-27
-21
460
250
157
320
Humphreys, Helton & Jones, 2007, Jones, Humphreys, & Helton 2007
Results from LMIRCam (2 – 5mm) on the LBT with AO
The SW clump – in the visible resolved into individual knots but very red and dusty. In
the near-IR an unresolved knot .
Clump is optically thick.
Mass > 5 x 10-3 Msun
Shenoy, et al. 2013
ALMA obs. At 321 & 658 GHz:
O’Gorman et al. (2015)
Mass 2 – 4 x 10-3 Msun
Origin of the Discrete Ejecta
Arcs and Knots are spatially and kinematically distinct;
ejected in different directions at different times; not
aligned with any axis of symmetry.
They represent localized, relatively massive
(few x 10-3 Msun) ejections
Large-scale surface activity -- starspots
Surface asymmetries supergiants, AGBs (Tuthill et al.
1999, Monnier et al. 2004)
Convective activity  Magnetic Fields 200 – 400G (Vlemmings et al. 2005)
magnetic fields measured in ejecta (OH, H2O, SiO masers)
VY CMa, VX Sgr, S Per, NML Cyg etc.
SOFIA imaging 20 – 37 mm
Note asymmetric shape.
Outer ejecta at most about
1200 yrs old
avg mass loss rate 6 x 10 -4
37mm contours on HST visual image
Shenoy et al. 2015
Interesting Chemistry
20 different molecules in the ejecta (Ziurys et al. 2007,
2009) including carbon molecules CO, HCN, CS,
HNC, CN, HCO+ . HNC and sulfur compounds in the
outflows. Suggest C molecules produced in shocks in
the outflows (also NML Cyg)
Kwok (2008)
Another chemical peculiarity – 12C/13C ratio
25– 46 , compared to 3 -14 in O rich supergiants
(Milam et al. 2009)
Ti02 @ 312 - 314 GHz:
de Beck et al. (2015)
Post-RSGs ( and post-AGBs) and a dynamical instability
The Warm Hypergiants and post RSG evolution
IRC+10420
Jones et al 1993
Oudmaijer et al 1994, 1996
Humphreys et al.1997, 2002
IRAS 17163-3907 – 12mm
(the fried egg) Lagedec et al
2011
HD179821 – 11.7 um
(Jura & Werner 1999)
Post RSG or AGB?
Other examples of post RSGs (Var A in M33, r Cas, HR5171a, HR8752 etc.)
The Yellow or Intermediate-type Post Red Supergiant IRC +10420
Strong IR excess
L ~ 5 x 105 Lsun
High mass loss rate 3-6 x 10-4
Warmest maser source
Spectroscopic variation late F  mid A
Complex CS Environment
One or more distant reflection shells
ejected ~3000 yrs ago
Within 2 “ – jet-like structures, rays,
small nearly spherical shells or arcs
Evidence for high mass loss ejections in the
1” = 5300 AU
Jones et al 1993
Oudmaijer et al 1994, 1996
Humphreys, Smith, Davidson, Jones, et al.1997
Humphreys, Davidson & Smith, 2002
past few hundred years
3D Morpholgy of IRC +10420 – 2nd epoch images from HST
spectra from HST/STIS
Numerous, arcs, knots ejected at different times (100400 yrs), directions, and all within a few degrees of
plane –
viewing nearly pole-on
Tiffany, Humphreys, Jones & Davidson 2010
Oudmaijer and de Wit 2013 VLT interferometry
Shenoy et al 2014 polarimetry
Semi-circular arcs – expanding bubbles? or loops?
Maser distribution
IRC +10420 -- circular polarization of OH (Nedoluha & Bowers 1992)
-> 300 – 400 G at the star (based on Vlemmings 2008)
Arcs and loops associated with surface activity
SOFIA, MIRAC, LMIRCam, 2 – 37mm 2 mass loss episodes (Shenoy et al. 2015),
2 x 10-3 Msun up until 2000 yrs ago, ~ 10-4 Msun . Mass lost in earlier ~ 7 Msun
The Evolutionary State and Mass loss mechanism for RSGs , post-RSGs ?
Pulsation
Convection
+?
A correlation among:
Surface asymmetries (spots) -> convective cells?
-> mass loss -> circumstellar dust/ejecta
Enhanced surface activity -> magnetic activity? ->
high mass loss events
-> complex CS ejecta
Some thoughts on the evolutionary state of the RSGs
Why not more VY CMa’s ?
Evidence for increased mass loss, CS ejecta, with higher luminosity and
cooler temperatures
Do RSGs evolve through the red supergiant stage getting apparently
cooler, more extended envelopes and high mass loss episodes?
Like lower mass stars, could there be more than one RSG stage?
 warm hypergiant  RSG again – extreme RSG (VY CMa) ?
Luminous Blue Variables (LBVs)
Conti (1984) suggested -- “luminous blue variable”
eta Car vars, P Cyg stars, S Dor vars, H-S vars
What is an LBV/S Dor variable?
Distinguished by their photometric and spectroscopic variability
In quiescence – hot, luminous star, sp. types late O to mid B, Of/WN7
Some emission lines H, He I, Fe II, P Cyg profiles
mass loss rates – typical
In “eruption” – rapid rise in apparent visual brightness -- weeks – months
apparent shift in sp. type ( late A to early F) or
apparent temp -- shift in bolometric correction
~ constant luminosity
star develops, slow, dense, optically thick wind
mass loss rate increases ~ 10-4 – few x 10-5 Msun/yr)
this optically thick wind stage may last years – decades
R127 (Walborn
et al. 2008)
S Doradus or LBV Instability Strip
Wolf (1989)
Note – in “eruption” – all about same temp ~ 7500 – 8000K
Davidson (1987) – opaque wind model (as opacity and mass loss rate
increase, temperature approaches a minimum)
Examples of reflection nebulae
associated with LBVs (K. Weis)
ejecta and atmospheres are N and He
rich  Evolved post MS
Same linear scale
Weis 2008
The Cause of the Instability?
Most explanations -- the star is near the Eddington Limit
LEdd = 4pcGMsun/k ,
GEdd = const k (L/Lsun) (M/Msun) -1
Opacity modified limit is temperature dependent
1. opacity – modified Eddington Limit (Davidson, Lamers, Appenzeller)
as temp decreases, opacity increases (“bi-stability jump”, Pauldrach & Puls 1990
Lamers et al 1995)
2. Omega limit -- add rotation to the Eddington Limit (Langer)
W = vrot/vcrit > 1, v2crit = (1 –G) GM/R
3. Vibration/Pulsation -- e mechanism (in the core) no longer considered applicable
to evolved stars
-- k mechanism in the envelope periods of weeks to months
4. Sub-photospheric – violent mode or strange mode instabilities Glatzel et al, Guzik,
Stothers & Chin
Caused by increase in opacity due to Fe at base of photosphere leading
to ionization induced instability
For stars > 30 --40Msun -- do not become RSGs, this is a critical high mass loss stage
For lower mass stars – have been RSGs, lowered surface gravity
Problem: – the S Dor instability strip has not been explained !
-- it depends on initial mass of star
-- note there are normal hot stars in the strip and a few to the right
Giant Eruptions and Supernova Impostors
Giant Eruption LBVs (Humphreys & Davidson (1994) -- increase their luminosity
during the eruption
Zwicky’s Type V SN
SN 1954J
Collaborators
VY CMa
Kris Davidson
Andrew Helton
George Herbig
Terry J. Jones
Gerald Ruch
Dinesh Shenoy
Nathan Smith
George Wallerstein
IRC+10420
Kris Davidson
Terry J. Jones
Dinesh Shenoy
Nathan Smith
Chelsea Tiffany
LBVs
Dominik Bomans
Kris Davidson
Kerstin Weis
h Car
Kris Davidson
Bish Ishibashi
John Martin
Andrea Mehner
Variable A in M33 – a warm hypergiant ~ 45 years in eruption!
Var A – Spectrum and Circumstellar IR excess
45 years in “eruption”
Cool dense wind
Looked like an M supergiant
Humphreys et al. 1987, 2006
NML Cyg
Optically obscured RSG embedded in a
small asymmetric bean-shaped nebula,
strong OH/IR source
mass loss rate 6 x 10-5 L ~ 5 x 105 Lsun
HII contours (30” away) due to interaction of
RSG wind with ionizing photons hot stars in Cyg
OB2 (Morris & Jura 1983)
0”.25 = 500 AU
Schuster, Humphreys & Marengo
(2006) , Schuster et al. (2009)
showed this is the molecular
photodissociation boundary
21 cm HI recombination