A New View of Accretion Shock Structure

Nancy S. Brickhouse

Harvard-Smithsonian Center for Astrophysics

Collaborators: Steve Cranmer, Andrea Dupree, Juan Luna, and Scott Wolk

Accretion Processes in X-rays: FromWhite Dwarfs to Quasars

Boston, MA 14 July 2010

Diverse X-ray Spectra from Young Stars Observed with Chandra HETG

Accretion shock models → Te and Ne for given mass accretion rate

• Kastner et al. (2002) find high Ne in TW Hya

• Chandra Large Observing Program to definitively establish accretion as the source of the emission,

and, if confirmed, bring new diagnostics to bear, using 500 ksec High Energy Transmission Grating

TW Hya

• Classical T Tauri star (accreting)• i=7o (pole-on)• M = 0.8 MSun

• R = 0.7 RSun

• Distance 57 pc• 10 million yr old • Poised to make planets• X-rays from the accretion shock (Kastner et al. 2002)

• X-ray plasma has high Neon abundance (Kastner et al. 2002; Drake, Testa, & Hartmann 2005)

Romanova et al. 2004

Accretion and a Corona

Emission Measure vs Te

Lightcurve

Emission measure distribution and variabilityallow us to isolate the accretion shock.

Brickhouse et al. 2010, ApJ, 710, 1835

He-like Line Ratio Diagnostics

He-like Energy LevelsNe and Te Diagnostic Ratios

(Smith et al. 2009)

Chen et al. 2006

Accurate Atomic Theory:Ne IX G-ratio Diagnostic

G-ratio vs Te

Smith et al. 2009

He-like Ions in TW Hya: O VII, Ne IX, and Mg XI

Diagnostics for Te and Ne

X-Ray Line Ratio Diagnostics for Density and Temperature

Ne = 6 x 1012 cm-3 Mg XI 3 x 1012 Ne IX 6 x 1011 O VII

Te = 2.50 ± 0.25 MK

This looks like the accretion shock!

Spectrum shows strong H-like Ne X and He-like Ne IX, up to n=7 or 8 in Ne X.Series lines are sensitive to absorption

Neon Region of HETG Spectrum

Complex absorption

• O VII: NH = 4.1 x 1020 cm-2

• Ne IX: NH = 1.8 x 1021 cm-2

Series lines rule out resonance scattering:

Tau ~ g f λ, for a given ion

Testing the Accretion Shock Model

Vff = (1 – R*/rt )1/2

~ 510 km/s

Te = 3.4 MK

Macc = f A* ρpre vff

2GM*

R*

●

(Konigl 1991; Calvet & Gullbring 1998; Gunther et al. 2007; Cranmer 2008)

Testing the Accretion Shock Model

Vff = (1 – R*/rt )1/2

~ 510 km/s

Te = 3.4 MK

Macc = f A* ρpre vff

2GM*

R*

●

(Konigl 1991; Calvet & Gullbring 1998; Gunther et al. 2007; Cranmer 2008)

“Settling”

Te and Ne from Ne IX agree with the shock model.

Model predicts Ne at O VII 7 times larger than observed.

•Consider a new 2-Region model: Region 1 = the shock front Region 2 = the post-shock region•Each region: Ne , Te, NH, V, and M•Predict r, i, and f for He-like ions•V2 = 300 x V1 => M2 = 30 x M1

Te and Ne from Ne IX agree with the shock model.

Model predicts Ne at O VII 7 times larger than observed.

•Consider a new 2-Region model: Region 1 = the shock front Region 2 = the post-shock region•Each region: Ne , Te, NH, V, and M•Predict r, i, and f for He-like ions•V2 = 300 x V1 => M2 = 30 x M1

Definitely not “settling”

Soft X-ray Excess (OVII) Ubiquitous

Gudel & Telleschi 2007also see Robrade & Schmitt 2007

Courtesy A. Szentgyorgyi

• Ne IX diagnostics for 3 observation segments ~150 ksec each give different Te and NH

(Ne does not vary).

• Variable Te implies changing rin, and thus Mdot

• Observed diagnostics constrain model Mdot , B, and rout

• Te varies from 3.1 to 1.9 K

• Mdot varies a factor of ~5

• Filling factor varies a factor of ~7

• rin varies from 1.75 R* to 3.52 R*

• B varies from 800 to 500 G

Accretion Variability

Brickhouse et al. 2010, in progress

Conclusions• High S/N HETG spectrum derives from 3 regions: a hot 10 MK corona, an accretion shock, and a cool

post-shock region.

• Diagnostics show excellent agreement with simple models of the shock itself.

• Diagnostics show that standard, one-dimensional models of the post-shock cooling plasma don’t work.

• Te and NH vary (Ne does not), implying variability in Mdot, B, and rtrunc

• Our values for Mdot are in good agreement with optical and UV methods.