1. Mass Loss and Winds All massive stars with L>104LSun have winds
All low and intermediate stars in post AGB evolution (M>0.6 MSun or L>103.7LSun have winds
Lower mass, less luminous stars do too, but mass loss rates are lower
Evidence for winds includes
P Cygni profiles
Radio or IR excesses
Thermal x-ray emission

2. The Solar Wind
The only cool dwarf with a directly observable wind, BUT
We do see x-rays from solar type stars (hot gas similar to the Sun’s corona (T~106, ~108 cm-3)
Observed decrease in rotation with age implies torque from a wind
ISM around nearby stars is disturbed (by wind…)
Solar wind discovered from space observations in the early ’60s
Wind inferred earlier from aurora, geomagnetic storms, comet tails
Parker (1958) demonstrated the wind must be there as a consequence of the high temperature of the corona and the low density of the ISM
Coronal gas pressure decreases outward, but tends toward a finite pressure at infinite radius (and that finite pressure is larger than the ambient pressure in the ISM)
Wind can be described by conservation of mass and momentum combined with the ideal gas law:

3. If the Star Rotates…
Early work of Mufson’s (Mufson & Liszt 1975) describes the effect of rotation on an expanding stellar wind assuming the corona rotates uniformly at a given angular velocity and angular momentum is conserved.
Rotation increases the speed of the wind due to centrifugal force
In the case of the Sun, virtually all of the driving force of the wind comes from the pressure gradient
Using appropriate values for the solar corona, we get wind speeds of about the right size for the solar wind.
Can study the evolution of the solar wind by varying the rotational velocity, and using solar evolutionary models to estimate solar radius and temperature, and phenomenological model of magnetic field

4. Evolution of the Solar Wind
Initially high mass loss rate decreases as rotation, magnetic field decline
Increases again in the old Sun due to enlarging radius
Figure from MacGregor, ESO Workshop, 1997

5. Three Solar Winds Wind from open magnetic fields
Terminal velocity= km/sec
nion=2.7 cm-3
1.5 x Msun per year
Wind from closed magnetic fields
Terminal velocity = 400 km/sec
nion=7 cm-3
2 x Msun per year
Coronal mass ejections
Terminal velocity = ~1000 km/sec
nion=0.2 cm-3
Uncertain mass loss rate

6. Hot Star Winds “Hot” means hydrogen is ionized.
Hot star winds are important from several astrophysical contexts:
Evolution of massive stars
Physical state of the ISM
Chemical evolution of galaxies
Interpretation of integrated spectra of extragalactic star bursts
Phenomenon itself is interesting
Early work (1940’s)on spectroscopy of WR stars – mass loss by high speed outflows, stratification of ionization (Beals)
Winds understood in the context of P Cygni profiles
Winds in O stars also inferred from C III l5696A emission in a Cam (O9.5Ia) with wings extending to 1500 km/sec (Wilson)
Expanding Shell
WR Star
Emission
Absorption
Emission

7. Winds in the Far UV
Winds became accessible to study with the beginning of far UV astronomy – discovery of P Cygni profiles for C IV, Si IV and N V in Orion supergiants
Optical lines appear only in the case of extreme winds, but UV lines are formed in the weak-wind approximation (modelling the circumstellar continuum as a geometrically diluted continuum of a static photosphere)
Decoupling radiative transfer from equations of statistical equilibrium and equations of gas dynamics allowed progress
Winds in OB supergiants are spherically symmetric, terminal velocities greater than stars’ escape velocities
Single-fluid approximations are OK (no differential streaming)
NOT analogs of the solar wind – existence of ions in the flow means temperature not too hot. Winds dynamically “cold,” driven by radiation pressure
Non-radiative heating also necessary
Modern treatments advanced to strong-wind limit with full physics included

8. Modeling Hot Star Winds
Radiation driven wind theory developed by Lucy and Solomon (1970) and Castor, Abbot, and Klein (CAK, 1975) to explain FUV line profiles
Unified Model Atmospheres for hot stars - spherically extended, sub- and supersonic atmospheric structure, including mass-loss rate, density, and velocity structure
Calculate energy distributions and spectra simultaneously for photospheric and wind lines, and mixed cases
Fully line-blanketed models
millions of spectra lines
equilibrium solutions with 150 ionic species
hydrodynamic equations for radiation-driven winds
Detailed spectrum synthesis in UV and optical

9. Observed vs. Synthetic Spectrum of LMC Sk –68 137 (O3)

10. Flux Distributions
Free-free thermal emission causes IR and radio excesses (x10)
Bound-free edges in UV and EUV also affected
Change in density stratification
Velocity shifts in resonance lines weakens optical thickness, more UV flux
More ionizing flux in the ISM

11. Wind Momentum-Luminosity Relation
Hot star winds result from radiation pressure
Properties of stellar winds must reflect the luminosities of stars
Wind momentum (mass loss rate x terminal velocity) should be a function of the photon momentum rate (f(L/c))
“It can be derived that”
Here, a is the power law of the line strength distribution function
Wind momentum depends on luminosity and weakly on radius
Wind momentum-luminosity relation can be used as a distance indicator

12. Areas of Current Work Variability of hot star winds
Time scales of less than an hour to several days
Rotational modulation – wind structures tied to photospheric origin
Pulsations
Non-spherical winds

13. Winds in Cool Stars
Observed as blue-shifted cores in the center of optically-thick resonance lines, emission features in UV resonance lines
Basic physical parameters include
Mass loss rate
Terminal velocity
Velocity law
Geometry
Time variability
Described by simple scaling laws: Reimers mass loss formula

14. AGB Mass Loss The “lemming” diagram Derived from Bowen models
Argues that Reimers formula is an artifact of a selection effect
We observe “Reimers” mass loss because stars with higher rates of mass loss are very short-lived

15. Wind Properties in Cool Stars
Geometry – we know winds aren’t spherically symmetric
Shocks, discontinuities, non-monotonic flows, circulation patterns
Magnetic properties – as yet no field measurements
Wind acceleration mechanisms not yet known
Convection driven acoustic waves
Pulsation driven acoustic waves
Alfven waves
Radiation pressure on dust
Parker-type thermal pressure gradients
Steady vs. stochastic?
NLTE in the ionization equilibria, need to infer mass loss rates (e.g Mg II emission strength depends on both the mass loss rate and the ionization state of Mg)
Turbulence, turbulent pressure gradients
Binary systems

16. Wind in a Tra Hybrid-chromosphere star High temperature emission lines
High velocity winds (P Cygni profile in Mg II
Fit with
spherical flow
Chromosphere at base of wind
Simple wind parameterization

17. Complexities in Winds Recall the 3 solar winds
Wind properties may depend on
Source of wind
Corotating interaction regions (when a fast wind overtakes a slower wind)
Properties may depend on viewing angle
Models also too simplistic
NLTE
Not in hydrodynamic equilibrium
More complex radiative transfer needed
Gas not in radiative equilibrium (heating and cooling times longer than dynamical time scales
Must solve radiative transfer and dynamics simultaneously