Resonance scattering in the X-ray emission line profiles of ζ Pup

Maurice Leutenegger

With David Cohen, Steve Kahn, Stan Owocki, and Frits Paerels

Resonance scattering in the X-ray emission line profiles of ζ Pup

X-ray emission from O star winds X-ray line profiles: summary of theory and

observation Data suggesting resonance scattering Profile formation with resonance scattering Application of RS profile model to data Implications

X-ray emission from O stars

The driving force in radiative lines is unstable

Tenuous streams undergo runaway acceleration before colliding with dense clumps, resulting in reverse shocks

Snapshot from simulation of Feldmeier (1995)

X-ray Doppler profiles(Owocki & Cohen 2001)

Wind modeled as a two-component fluid Cool bulk of wind (absorbs X-rays) Small fraction is heated in shocks (emits X-rays)

X-ray Doppler profiles: emission from thin shells

Example profiles

Parameter dependence

Two parameters influence radial distribution of X-ray emitting plasma: “Turn-on” radius (expected to be ~ 1.5 stellar radii) Filling factor (power law in radius)

Parameter dependence

The cool part of the wind absorbs X-rays as they leave the wind Characteristic continuum optical depth:

Example profiles

Qualitative summary of model profile behavior:

Degree of blueshift measures characteristic continuum optical depth to X-rays

Width measures the onset radius of X-ray emission

Comparison with data

Possible explanations for X-ray profile shapes

(more than one may apply)

Mass loss rates too high – characteristic optical depths really are low

Porosity reduces macroscopic effective optical depth

Resonance scattering causes emission to be intrinsically shifted towards line center

Empirical evidence suggesting resonance scattering?

Empirical evidence suggesting resonance scattering?

Resonance scattering (Ignace et al 2002, Leutenegger et al 2007)

For an optically thick resonance line in a moving stellar atmosphere (Sobolev theory): radial photon escape is due to the radial velocity

gradient (dv/dr) lateral photon escape is due to the spherical

divergence of the wind (v/r) Far out in the wind dv/dr goes to zero, so lateral

escape is favored If the observed X-ray emission comes from far

out in the wind, the profiles are more symmetric

Angular dependence of normalized escape probability (optically thick)

How does resonance scattering affect the model profiles?

Including resonance scattering in N VI leads to much better fit

Including resonance scattering in N VI leads to much better fit

Also improves fit to O VII

Also improves fit to O VII

What about other lines?

We can only infer the importance of RS by comparing two lines from the same ion – one must be a resonance line, and the other must not

But if RS is important in N VI and O VII, it should be important for other strong resonance lines as well!

If resonance scattering is important how do we measure anything?

Problem: profile shape is roughly degenerate for high continuum optical depth with resonance scattering and low continuum optical depth without resonance scattering

Use non-resonance lines Even some resonance lines will not be optically

thick (Make predictions for line optical depth)

Summary

Different profile shapes in resonance and intercombination lines from the same ion can only be explained by resonance scattering

Resonance scattering can explain at least some of the unexpected lack of asymmetry in other profiles

Either porosity or reductions in mass-loss rates (relative to density-squared diagnostics) are still likely to be important in addition to RS

Characteristic optical depth to resonance scattering

Expected values of optical depth