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: From White Dwarfs to Quasars Boston, MA 14 July 2010

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

Accretion shock models → T e and N e for given mass accretion rate Kastner et al. (2002) find high N e 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=7 o (pole-on) M = 0.8 M Sun R = 0.7 R Sun 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 T e Light curve Emission measure distribution and variability allow us to isolate the accretion shock. Brickhouse et al. 2010, ApJ, 710, 1835

He-like Line Ratio Diagnostics He-like Energy Levels N e and T e Diagnostic Ratios (Smith et al. 2009)

Chen et al Accurate Atomic Theory: Ne IX G-ratio Diagnostic G-ratio vs T e Smith et al. 2009

He-like Ions in TW Hya: O VII, Ne IX, and Mg XI Diagnostics for T e and N e

X-Ray Line Ratio Diagnostics for Density and Temperature N e = 6 x cm -3 Mg XI 3 x Ne IX 6 x O VII T e = 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: N H = 4.1 x cm -2 Ne IX: N H = 1.8 x cm -2 Series lines rule out resonance scattering: Tau ~ g f λ, for a given ion

Testing the Accretion Shock Model V ff = (1 – R*/r t ) 1/2 ~ 510 km/s T e = 3.4 MK M acc = f A * ρ pre v ff 2GM * R * ● (Konigl 1991; Calvet & Gullbring 1998; Gunther et al. 2007; Cranmer 2008)

Testing the Accretion Shock Model V ff = (1 – R*/r t ) 1/2 ~ 510 km/s T e = 3.4 MK M acc = f A * ρ pre v ff 2GM * R * ● (Konigl 1991; Calvet & Gullbring 1998; Gunther et al. 2007; Cranmer 2008) “Settling”

T e and N e from Ne IX agree with the shock model. Model predicts N e 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, N H, V, and M Predict r, i, and f for He-like ions V2 = 300 x V1 => M2 = 30 x M1

T e and N e from Ne IX agree with the shock model. Model predicts N e 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, N H, 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 2007 also see Robrade & Schmitt 2007

Courtesy A. Szentgyorgyi

Ne IX diagnostics for 3 observation segments ~150 ksec each give different T e and N H (N e does not vary). Variable T e implies changing r in, and thus Mdot Observed diagnostics constrain model Mdot, B, and r out T e varies from 3.1 to 1.9 K Mdot varies a factor of ~5 Filling factor varies a factor of ~7 r in 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. T e and N H vary (N e does not), implying variability in Mdot, B, and r trunc Our values for Mdot are in good agreement with optical and UV methods.