An Overview of Some Recent Results in the Galactic Centre

The X-ray Universe 2014

Trinity College, Dublin

2014 June 16-19

Frederick K. Baganoff

MIT Kavli Institute for Astrophysics and Space Research

Outline

• Traces of past Sgr A* activity• Current Sgr A* activity

– Sgr A* flares– Sgr A*/G2 interaction monitoring

• Magnetar SGR J1745-2900

NuSTAR Detects Sgr A* Flares

NuSTAR Detects Sgr A* Flares

Sgr A* X-ray Visionary Project

Frederick K. Baganoff, Michael A. Nowak (MIT Kavli Institute), Sera Markoff (API, University of

Amsterdam) and Sgr A* XVP Collaboration

www.sgra-star.com

• Obtain first high-resolution X-ray spectra of SgrA* and diffuse emission in the central pc

• Spatially and spectrally resolve accretion flow within Sgr A*’s Bondi radius (~3”)

• Measure energy and width of known Fe Kalpha line(s) at ~6.6 keV in ACIS-I Spectrum

• Detect optically thin plasma emission lines, (e.g., Si, S, Fe), if present at levels predicted by RIAF models

• Measure relative abundances of emitting plasmas and absorbing column along LOS

• Obtain first high-resolution X-ray spectra of SgrA* and diffuse emission in the central pc

• Spatially and spectrally resolve accretion flow within Sgr A*’s Bondi radius (~3”)

• Measure energy and width of known Fe Kalpha line(s) at ~6.6 keV in ACIS-I Spectrum

• Detect optically thin plasma emission lines, (e.g., Si, S, Fe), if present at levels predicted by RIAF models

• Measure relative abundances of emitting plasmas and absorbing column along LOS

Science GoalsScience Goals

• Radio polarization measurements indicate most accreting matter does not reach Sgr A*’s event horizon --- dynamics and thermal structure of plasma will tell us how matter flows in and how much and where some of it flows out

• Monitor X-ray flares of Sgr A* with increased cadence and minimal pile-up

• Perform coordinated multi-wavelength monitoring of flares from radio up to gamma-rays, including mm VLBI

• Constrain 3D GRMHD simulations of Sgr A* accretion flow using MW properties of flares and 1mm VLBI imaging

• Radio polarization measurements indicate most accreting matter does not reach Sgr A*’s event horizon --- dynamics and thermal structure of plasma will tell us how matter flows in and how much and where some of it flows out

• Monitor X-ray flares of Sgr A* with increased cadence and minimal pile-up

• Perform coordinated multi-wavelength monitoring of flares from radio up to gamma-rays, including mm VLBI

• Constrain 3D GRMHD simulations of Sgr A* accretion flow using MW properties of flares and 1mm VLBI imaging

Science GoalsScience Goals

Observational Overview

• Completed 3Ms exposure of Sgr A* - 38 separate HETGS observations from 2012 February 6 to October 31

• Observing constraints very challenging:– Desired specific roll angles to minimize background: 76.4

(6), 76.6 (1), 92.2 (10), 268.7 (7), 270.7 (1) & 282.3 (13) degrees

– Desired long exposures for multi-wavelength monitoring of Sgr A* flares: > 90ks (14)

Sgr A* XVP Cumulative Light Curve

•Brightest flare ~140x quiescence •LX ~ 2 x 1035 erg/s (2-8 keV)•Nowak et al. (2012)

Brightest Flare Light CurveBrightest Flare Light Curve

Consistent w/ minimal pile-up

Bayesian Blocks finds structure at peak

Sgr A* Flare and Quiescent Spectra -- ACIS-I & HETGS

Sgr A* Flare and Quiescent Spectra -- ACIS-I & HETGS

•HETGS 1st Order

•HETG 0th Order

•ACIS-I

•HETGS 0th + 1st Order

•HETGS 1st Order

•HETGS 0th Order

Wang+2013: He-like Fe Kalpha line at 6.7 keV; No 6.4 keV line predicted by Sazonov+2012; but see Warwick talk

Consistent Flare Properties for Brightest Chandra & XMM Flares

Consistent Flare Properties for Brightest Chandra & XMM Flares

ACIS-I Quies.

HETGS Flare

XMM 2002 F

XMM 2007 F

Photon index ~ 2 and N~ 2 and NHH ~ 15 x 10 ~ 15 x 102222 cm cm-2-2

Properties of Brightest X-ray Flare

Properties of Brightest X-ray Flare

Good agreement with two brightest XMM flares(Nowak et al. 2012, ApJ, 759, 95)

39 X-ray Flares in 3Ms Exposure: 1.12 +/- 0.18 flares/day

Neilsen+2013

Sgr A* Flare Science Topics

• X-ray flares are non-thermal but mechanism is still undetermined: synchrotron with cooling break (SB), external Compton (EC) or synchrotron self-Compton (SSC)?

• X-ray & NIR flares appear related (opt thin process)

• What causes variable X-ray-to-NIR flux ratio?

• mm/radio events are weaker & longer timescale (~opt thick process)

• Are mm/radio events correlated with X-ray/NIR?

– If so, do they lead, lag or sometimes lead other times lag?

– Need sufficient sample of MWL flares to establish or refute correlation with X-ray/NIR flares

Sgr A* Flare Science Topics

• Origin(s) of flares undetermined:• magnetic reconnection?• shock in jet or inner accretion flow?• stochastic acceleration?• magnetic excitation by infalling asteroids?

• What determines flare duty cycle and energy budgets?• Need broad-band flare spectra and time evolution

observations to understand flares• What do flares tell us about innermost environment of Sgr A*’s

accretion flow?• Does this phenomenon scale to other LLAGN?• Will G2 or similar events change Sgr A*’s rate of flaring and its

accretion state?

NuSTAR Detects Sgr A* Flare!

Barriere+2014NuSTAR PI: F. Harrison (Caltech)

NuSTAR Detects Sgr A* Flares

NuSTAR Detects Sgr A* Flares

NuSTAR Detects Sgr A* Flares

NuSTAR Spectra

Flare Model Fits To NuSTAR Spectrum

Where are Sgr A* Flares Located?

3Ms Chandra HETGS Image of Sgr A*

• Events projected to common tangent plane

• Events outside ~250 pixels rotated to align grating arms

• Moderate smearing of co-aligned spectra

•S2/S3 chip gap

•Fe XXV•Ar XVII /S XVI

•S XV

•S XVI

•Ca XIX

Fe XXV 1.8505 Angstroms

No Detectable lines

No Detectable lines

Ca XIX 3.177 & 3.211 Angstroms

Ar XVII 3.95 & Ar XVII/S XVI 3.995 Angstroms

Ar XVII 3.95 & Ar XVII/S XVI 3.995 Angstroms

S XVI 4.727 & 4.733 S XV 5.039 Angstroms

S XV 5.039 & 5.101 Angstroms

No Detectable Lines

XVP Progress Summary

• Number of X-ray flares tripled and essentially pile-up free: flare distributions may hint at multiple mechanisms

• 0th-order quiescent spectrum (Fe Kalpha) disagrees with Sazonov et al. (2012) model for origin of extended emission vs RIAF

• But see talk by Warwick: Low-luminosity X-ray sources and the Galactic ridge X-ray emission!

• Brightest flare properties consistent with two brightest XMM flares: photon index ~ 2 and NH ~ 15 x 1022 cm-2

• Modeling background to produce cleanest gratings spectrum of Sgr A* in quiescence

• Believe that we have good detection of He-like Fe line; working to reduce background to detect additional lines => possible relative abundance measurements

Is G2 a Gas Cloud on Its Way Towards the Supermassive Black

Hole at the Galactic Centre?

Gillessen+2012

• ~3 earth-mass gas cloud approaching Sagittarius A* on a nearly radial orbit

• Pericenter ~3100 RS, ~0.03” or ~36 light hr at 2013.5

• Cloud has begun to disrupt over past 3 yr, probably due to tidal shearing

• Dynamical evolution and radiation of cloud will probe properties of accretion flow and feeding processes of Sgr A*

• keV emission of Sgr A* may brighten significantly at closest approach

• Hydrodynamic simulation predicts increased feeding of Sgr A* in a few years

Infalling Dusty Gas Cloud in Galactic Center

• Cloud detected at M and L’, not Ks or H

• Dusty cloud Td ~ 550 K

• Proper motion ~42 mas/yr or 1670 km/s

• Br g radial velocity ~1650 km/s

• e ~ 0.94 bound orbit• Orbital period ~137 (11) yr

• Panel d shows orbits of cloud and star S2

Velocity Shear in Gas Cloud

• Integrated Br g maps vs stellar PSF ~21 mas E-W

• Position-velocity maps of Br g emission show head-tail structure; ~62 mas for head

• Tail spread ~200 mas downstream of head

• Velocity gradient ~2 km/s/mas

• 89 (30) km/s in 2003 increased to 350 (40) km/s in 2011

Cloud Properties

• LIR ~ 5 Lsun; LBr g ~ 2 x 10-3 Lsun

• Case B recombination: ne ~ 2.6 x 105 fv-½cm-3

• Specific angular momentum ~50x less than other clouds

• Current density ~300fv-½x greater than surrounding hot

accretion flow; decrease to ~60fv-½x at peribothron

X-ray Emission as Probe of Accretion Flow Profile & BH Feeding

• Cloud remains cold until just before peribothron

• Post-shock Tc ~ 6-10 x 106 K

• Lx <~ 1034 erg/s (2-8 keV); possibly variable

• Stronger X-ray emission for steeper radial profiles of accretion flow density & temperature and higher fv

• May release up to 1048 erg over decade => <Lx> ~1039-40 erg/s

• Sufficient to produce Fe Ka reflection features seen in Galactic Center (e.g., Sgr B2) => possible light echoes from previous clouds accreting onto Sgr A*?

Sgr A*/G2 Cloud Predictions

• Gillessen+2012: RJ & KH instabilities compress & heat G2 near pericenter causing sustained increase in NIR & X-ray fluxes on ~4-month dynamical timescale

• Sadowski+2013: MHD simulations predict radio bow shock visible 7-9 months prior to CM pericenter

• Shcherbakov 2013: magnetically arrested cloud model resists compression near pericenter but predicts radio bow shock and X-ray enhancement ~18 months prior to CM pericenter (i.e., 2012); original model ruled out by Chandra Sgr A* XVP monitoring – no X-ray enhancement detected

• Shcherbakov 2014: revised best-fit model to data predict enhancements below quiescent Sgr A* level => could see nothing if G2 is gas cloud

Alternative Interpretation – G2 a Star

• Phifer+2013: G2 is dusty, windy young star similar to other features with similar NIR colors in central 2’

• Report Br_gamma line position offset from L’ position => suggest VLT slit pulling in two separate red clumps to produce apparently tidally sheared feature in position-velocity diagram

• MPE group disagrees; no consensus between groups yet• If G2 a star => will remain intact through pericenter passage and

move away from Sgr A* with constant Br_gamma flux• However some wind and envelope material may fall onto Sgr A*

accretion flow over next few years => keep monitoring

Chandra View of Central Parsec – Pre Magnetar

Chandra Monitoring of Sgr A*/G2 & SGR J1745-2900 in 2014

Chandra Avg Quiescent Count Rates of Sgr A*/G2 & SGR J1745-2900

Exponential decay ~ 146 d

Chandra Light Curve of Sgr A*/G2

Chandra Monitoring

• NIR-derived orbit => pericenter ~120 AU 2014 late March +/- 3 weeks (Gillessen+2013; Meyer+2014)

• Chandra/ACIS-S3 monitored Sgr A*/G2 6 times between 2014 Feb 21 & June 3

• SGR J1745-2900 located 2.4” (~0.1 pc) in projection from Sgr A*• Extract light curves & spectra within 2” radius of magnetar and

SgrA* products within 1.25” radius; Sgr A* contaminated ~0.6-0.9% by magnetar flux

• Avg unabsorbed quiescent 2-8 keV flux of Sgr A* in 3Ms XVP project 4.5e-13 erg cm2 s-1 (Nowak+2012, Nielsen+2013)

• Measured fluxes corrected for magnetar contamination in range (3.5-4.1)e-13 erg cm2 s-1 with uncertainties of +/-0.6e-13 erg cm2 s-1 (90% confidence)

• Conclusion: Chandra finds no evidence for sustained enhanced emission at Sgr A* so far

• See ATel #6242 (Haggard+2014) for details

NIR & Radio Monitoring

• ATel #6110 - May 2 - Keck NIRC2 detects G2 at 3.8 um with de-reddened flux density 1.7 ± 0.2 mJy (or equivalently an observed L’ magnitude of 14.1 ± 0.2), consistent with 2002-2013 measurements (Ghez+2014)

• Conclude G2 currently undergoing closest approach & still intact• ATel #5969 – March 10 – NRAO VLA Service monitoring at multiple

freqencies & dates find no enhancements (Chandler+2014)• ATel #6083 - April 21- JVN monitoring at 22 GHz finds avg fluxes

consistent with previous levels (Tsuboi+2014)

SGR 1745-2900 : a magnetar within grasp of our SMBH

The angular separation from Sgr A* is 2.4"+/-0.3" (within the PSF of all X-ray instruments beside Chandra).

(Rea+2013)

SGR 1745-2900: Timing propertiesP = 3.7635537(2) s (epoch TJD 16424.5509871)871),Pdot =6.61(4)x10-12 s/sBdip =1.6x1014 G

Edot = 5x1033 erg/stauc = 9 kyr

Large timing noise! Pdot changes or glitches…

(Mori et al. 2013; Kennea et al. 2013; Rea et al. 2013, Kaspi et al. 2014, Coti Zelati et al. 2014, in prep)

SGR 1745-2900: Spectral evolution

•(Coti-Zelati, Rea, Papitto et al. 2014, in prep)

Is SGR 1745-2900 bound to Sgr A*?

SGR 1745-2900 has, on average, a 90% probability of being bound to the SMBH with an orbital period between 500 yr to

several kyrs.(Rea+2013)

Is SGR 1745-2900 so close to SgrA* ?

Using the Cordes & Lazio (2002) H distribution we estimate that a DM=1750 pc cm-3 results in a distance of 8.3 kpc (same as SgrA*). If we assume this distance, a 2.4" projected distance translates in:

d = 0.09+/-0.02 pc (90% error).

Distance determination from DM of the radio pulsar:

(Rea+2013)

Chance coincidence probability

- neutron star density in the Galactic disc 10-4 pc-3 we expect ~0.3 neutron stars in a narrow cone of aperture 2.4", regardless of the age.

- with an age < 10 kyr, the probability of SGR J1745–2900 being a neutron star wandering across the line of sight reduces to <3x10-6.

No foreground/background object.

- it has been estimated that in the 1pc around the Galactic center there are ~80 pulsar with an average age of 107 yr, and we expect one or two pulsars with < 10kyr. (see e.g. Freitag et al. 2006; Wharton et al. 2012)

Why is the young pulsar close to Sgr A* a magnetar?

(Schoedel+2009; Lu+2009; Rea+2013)

We know of about a hundred young radio pulsars, but only 12 young magnetars. Why ts the only young PSR expected within 1pc from Sgr A* a magnetar?

1. The birth rate of magnetars is comparable with that of "normal" pulsars"

2. The Galactic center is more inclined to form magnetars than normal pulsars.

...probably both are true!

•2.4”

•SgrA*

•SGR J1745-2900

VLT-NACO

•SGR J1745-2900

Summary

Our normal Milky Way Galaxy has an active Galactic Center

Weak AGN activity currently

Possibly much stronger in the past century (see Ponti talk)

X-ray Transients

Magnetars & Pulsars in central pc => one day find aPSR close enough to Sgr A* to test GR?