observing black holes with 1m-class telescopes charles bailyn yale university with thanks to: j....

Observing Black Holes With 1m-Class Telescopes

Charles Bailyn

Yale University

With thanks to: J. McClintock (CfA), R. Remillard (MIT), J. Orosz (SDSU), Yale students and data aides C. Baldner, A. Cantrell, B. Cobb, S. Curry, Z. Dugan, M. Dwyer, F. Edelman, L. Ferrara, J. Greene, B. Heflin, R. Jain, L. Jeanty, R. Kennedy-Shaffer, D. Maitra, E. Neil, K. Whitman, CTIO/SMARTS staff M. Buxton, D. Gonzalez, J. Espinoza, A. Miranda, J. Nelan, S. Tourtellotte, R. Winnick

Observing Black Holes With 1m-Class Telescopes

Introduction to Black Hole X-ray TransientsIntroduction to Black Hole X-ray Transients 2 ½ Strong-Field Relativistic Effects2 ½ Strong-Field Relativistic Effects Recent results on BHXTs with SMARTSRecent results on BHXTs with SMARTS

Transient X-ray Binaries

Accreting compact objectAccreting compact object Eddington-limited outbursts (rise of days; Eddington-limited outbursts (rise of days;

duration of weeks; recurrence of decades)duration of weeks; recurrence of decades) Superluminal jetsSuperluminal jets Quiescent light dominated by companionQuiescent light dominated by companion

Scientist’s conception of GRO J1655-40 in outburst from Rob Hynes

The Mass Function

1)1(

sin2 2

12

31

3

)( MMfM

M

iMG

PK

Bailyn et al. 1995 GRO J1655-40

measurable only in quiescence!

PK

Mass Limit of Neutron Stars R > Schwarzschild only if M < 3 Msun R > Schwarzschild only if M < 3 Msun

(modest change for spinning n.s.)(modest change for spinning n.s.) Equation of State agrees with experiment Equation of State agrees with experiment Extrapolation is causalExtrapolation is causal

1.5 < M/Msun < 2.2 1.5 < M/Msun < 2.2 Limit for Plausible Equations of StateLimit for Plausible Equations of State

Higher limits require eitherHigher limits require either Non-standard gravityNon-standard gravity

Non-baryonic starNon-baryonic star

“Proof” of a Black Hole

L=10^4 Lsun, T=10^3 Tsun L=10^4 Lsun, T=10^3 Tsun => R =10^-4 Rsun=> R =10^-4 Rsun

Millisecond time variability Millisecond time variability => R<10^-3 Rsun=> R<10^-3 Rsun

P, K => M > 3MsunP, K => M > 3Msun => compact object too massive to be a => compact object too massive to be a

neutron starneutron star Requires black hole, non-baryonic star, or Requires black hole, non-baryonic star, or

non-GR gravity non-GR gravity

Finding a Black Hole

Wait for new X-ray transientWait for new X-ray transient Identify optical counterpartIdentify optical counterpart Wait for quiescenceWait for quiescence Measure mass functionMeasure mass function If f(M)>3, you win!If f(M)>3, you win!

(RXTE era discovery rate ~1/yr)(RXTE era discovery rate ~1/yr)

Determining Black Hole Mass

Mass ratio measurable from line broadening Mass ratio measurable from line broadening (usually a small effect)(usually a small effect)

Inclination from ellipsoidal variabilityInclination from ellipsoidal variability 2ndary non-spherical (Roche lobe filling)2ndary non-spherical (Roche lobe filling) Two maxima and two minima per orbitTwo maxima and two minima per orbit Amplitude depends on inclinationAmplitude depends on inclination

1)1(

sin2 2

12

31

3

)( MMfM

M

iMG

PK

Ellipsoidal Variations

Problems with Ellipsoidal Variability

Residual Disk Light (degenerate with Residual Disk Light (degenerate with inclination to first order)inclination to first order)

Other light sources (hot spots etc)Other light sources (hot spots etc) Star spots (especially on late-type 2ndaries)Star spots (especially on late-type 2ndaries) Eclipses (star of disk, disk of star)Eclipses (star of disk, disk of star)

NOTE: different temperature dependencesNOTE: different temperature dependences

GRO J1655-40 (Greene. Bailyn & Orosz 2001)

P = 2.62192(20) days

f(M) = 2.73 +/- 0.09

Inclination 70 +/- 2

M_1 = 6.3 +/- 0.5

M_2 = 2.6 +/- 0.3

Dynamically Confirmed Black Hole Candidates

Extragalactic BHXBs: M33 X-7 (Orosz et al. 2007)

Tests of General Relativity

Solar system: very high precision, weak Solar system: very high precision, weak field limitfield limit

Pulsars: very high precision, 1Pulsars: very high precision, 1stst and 2 and 2ndnd order fieldsorder fields

Gravitational waves: strong field, multi-Gravitational waves: strong field, multi-parameter parameter not yet observednot yet observed

Accreting black holes: strong field, multi-Accreting black holes: strong field, multi-parameter, many constraintsparameter, many constraints

Reversing the Question

IF General Relativity applies and stars are IF General Relativity applies and stars are baryonic, THEN these are Black Holes;baryonic, THEN these are Black Holes;

Reversing the Question

IF General Relativity applies and stars are IF General Relativity applies and stars are baryonic, THEN these are Black Holes;baryonic, THEN these are Black Holes;

IF these are not Black Holes, THEN either IF these are not Black Holes, THEN either General Relativity does not apply or there are non-General Relativity does not apply or there are non-baryonic stars.baryonic stars.

So, search for consequences of strong-field So, search for consequences of strong-field relativity, such as the event horizon, the inner-relativity, such as the event horizon, the inner-most stable circular orbit, possibly jet formation most stable circular orbit, possibly jet formation

Do Event Horizons Exist?

Black Hole

Infalling Gas

Boundary layer X-rays

Neutron Star

Disks vs. ADAFs

Keplerian Accretion Disk: half energy dissipated Keplerian Accretion Disk: half energy dissipated in disk, half at boundary layerin disk, half at boundary layer

Advection Dominated Accretion FlowsAdvection Dominated Accretion Flows Two T plasma: ions hot, electrons coolTwo T plasma: ions hot, electrons cool Thermal, kinetic energy advected inwardsThermal, kinetic energy advected inwards >99% of energy dissipated at boundary layer>99% of energy dissipated at boundary layer Requires low mass accretion rateRequires low mass accretion rate

Quiescent transients fit outer disk + ADAFQuiescent transients fit outer disk + ADAF

Evolution of Spectral States

Esin et al. 1997

(see also McClintock & Remillard 2006)

A0620-00 in Quiescence

Narayan, McClintock & Yi 1996

Black Holes vs. Neutron Stars

Garcia et al. 2001

Problems and Uncertainties

Dependence on orbital period – binary Dependence on orbital period – binary evolutionevolution

Changing nature of time-dependant Changing nature of time-dependant accretion flow (disk, corona, ADAF etc)accretion flow (disk, corona, ADAF etc)

What about outflows (ADIOS, jets)?What about outflows (ADIOS, jets)?

Innermost Stable Circular Orbit Relativity predicts an ISCO at a radius Relativity predicts an ISCO at a radius

determined by M determined by M and and J of Black HoleJ of Black Hole In “high-soft” state, X-ray spectrum fits In “high-soft” state, X-ray spectrum fits

superposition of black bodies from disksuperposition of black bodies from disk ISCO represents hottest contributing black ISCO represents hottest contributing black

bodybody Spectral modelling can measure ISCOSpectral modelling can measure ISCO With known mass and geometry, one can With known mass and geometry, one can

determine Jdetermine J

Recent ISCO Determinations

McClintock et al. 2006

Superluminal Jets?

Well-known special relativistic effect in quasarsWell-known special relativistic effect in quasars Now observed in several BHXNs (radio Now observed in several BHXNs (radio andand X-ray X-ray

observations)observations) Associated with “low-hard” (non-thermal) Associated with “low-hard” (non-thermal)

emission statesemission states Collimation and energy mechanisms unclear – Collimation and energy mechanisms unclear –

frame-dragging frame-dragging may may be importantbe important Correlation of jet strength with J would be Correlation of jet strength with J would be

importantimportant Amount of mass ejected crucial to understand Amount of mass ejected crucial to understand

ADAFsADAFs

Small and Moderate Aperture Research Telescope System (SMARTS)

Operates 4 1m-class Operates 4 1m-class telescopes at CTIOtelescopes at CTIO

Variety of instruments Variety of instruments and operating modesand operating modes

Over a dozen Over a dozen participating institutions participating institutions (now including NExScI)(now including NExScI)

~25% of time available ~25% of time available through NOAOthrough NOAO

Current SMARTS Capabilities

1.5m + low and high resolution 1.5m + low and high resolution spectrographs (queue observing)spectrographs (queue observing)

1.3m + ANDICAM - dual channel O/IR 1.3m + ANDICAM - dual channel O/IR (monitoring queue observing ONLY)(monitoring queue observing ONLY)

1.0m + 4K CCD (user runs) 1.0m + 4K CCD (user runs) 0.9m + 2K CCD (user/service alternate)0.9m + 2K CCD (user/service alternate)

Current SMARTS Capabilities

1.5m + spectrograph/IR imager (service and 1.5m + spectrograph/IR imager (service and queue observing)queue observing)

1.3m + ANDICAM - dual channel O/IR 1.3m + ANDICAM - dual channel O/IR (monitoring queue observing ONLY)(monitoring queue observing ONLY)

1.0m + 4K CCD (user runs) 1.0m + 4K CCD (user runs) 0.9m + 2K CCD (user/service alternate)0.9m + 2K CCD (user/service alternate)

Yale/SMARTS BHXN Program

Observe ~12 sources per night in O/IRObserve ~12 sources per night in O/IR Quiescence: build up long-term ellipsoidal Quiescence: build up long-term ellipsoidal

lightcurveslightcurves New outbursts – trigger X-ray observations New outbursts – trigger X-ray observations

Outburst monitoring – state changes, multi-Outburst monitoring – state changes, multi-

wavelength correlationswavelength correlations

Expectations for Optical/IR During Outburst Cycle

Disk Instabilities lead to Fast Rise and Disk Instabilities lead to Fast Rise and Exponential Decay (FRED)Exponential Decay (FRED)

Optical precedes X-rays and lasts longerOptical precedes X-rays and lasts longer Same sequence of states in rise and fallSame sequence of states in rise and fall O/IR is a superposition of thermal spectraO/IR is a superposition of thermal spectra

Aquila X-1

Neutron star transient (displays bursts)Neutron star transient (displays bursts) Shortest recurrence time (~ 1 year)Shortest recurrence time (~ 1 year) Orbital period ~ 18 hoursOrbital period ~ 18 hours Nearby neighbor ~ 2 mags brighter in Nearby neighbor ~ 2 mags brighter in

quiescencequiescence Declination ~ 0: everyone can play!Declination ~ 0: everyone can play! SMARTS lightcurve in Maitra & Bailyn SMARTS lightcurve in Maitra & Bailyn

20082008

O/IR vs X-rays in Aquila X-1


F.R.E.D.s


F.R.E.D.s

L.I.S.s


F.R.E.D.s

L.I.S.s

Mini-outbursts

Aquila X-1: 2000 Outburst


Maitra & Bailyn, 2004

8 Years of Aquila X-1

F.R.E.D.s – similar to expectationsF.R.E.D.s – similar to expectations L.I.S.s – variable flux, low/hard X-rays, L.I.S.s – variable flux, low/hard X-rays,

also seen in other neutron star transientsalso seen in other neutron star transients Mini-outbursts – no X-ray response in ASMMini-outbursts – no X-ray response in ASM Optical precedes X-ray, as expectedOptical precedes X-ray, as expected Hysteresis of X-ray states, unexpected, also Hysteresis of X-ray states, unexpected, also

seen in black hole candidatesseen in black hole candidates

4U1543-47

Soft X-ray transient with ~ 10 year Soft X-ray transient with ~ 10 year recurrence timescalerecurrence timescale

Low mass function and low inclination Low mass function and low inclination > black hole system (Orosz et al. 2001) > black hole system (Orosz et al. 2001)

A-star companion in ~ 1 day orbitA-star companion in ~ 1 day orbit OUTBURST IN SUMMER 2002!OUTBURST IN SUMMER 2002!

4U1543-47 in 2002Buxton &

Bailyn 2004

4U1543-47 in 2002Buxton & Bailyn, 2004

4U1543-47 in 2002

Recent SMARTS data of GX 339-4

Expectations for Optical/IR During Outburst Cycle

Disk Instabilities lead to Fast Rise and Disk Instabilities lead to Fast Rise and Exponential Decay (FRED)Exponential Decay (FRED)

Optical precedes X-rays and lasts longerOptical precedes X-rays and lasts longer Same sequence of states in rise and fallSame sequence of states in rise and fall O/IR is a superposition of thermal spectraO/IR is a superposition of thermal spectra

“Typical” Quiescent Data(Greene. Bailyn & Orosz 2001)

P = 2.62192(20) days

f(M) = 2.73 +/- 0.09

Inclination 70 +/- 2

M_1 = 6.3 +/- 0.5

M_2 = 2.6 +/- 0.3


Cantrell et al. 2008

V4641 Sgr in Quiescence

Cantrell et al. in prep.

Quiescent Behavior

Optical lightcurves vary with timeOptical lightcurves vary with time Disk contribution both important and Disk contribution both important and

variablevariable Long term data sets modelled with Long term data sets modelled with

consistent consistent orbital parameters are crucialorbital parameters are crucial Caution needed in comparing quiescent Caution needed in comparing quiescent

ADAF-associated X-ray emission!ADAF-associated X-ray emission!

Black Hole X-ray Transients in 2008

22 “Dynamically Confirmed Black Hole 22 “Dynamically Confirmed Black Hole Candidates”Candidates”

Most have masses well above neutron starsMost have masses well above neutron stars Strong field relativistic effects apparently Strong field relativistic effects apparently

manifestedmanifested Complexity of outburst cycle and accretion flow Complexity of outburst cycle and accretion flow

beginning to be exploredbeginning to be explored

Future Work New sources from RXTE/ASM, Swift, including New sources from RXTE/ASM, Swift, including

Local Group targetsLocal Group targets Daily SMARTS data for ~ dozen sources and for Daily SMARTS data for ~ dozen sources and for

new sourcesnew sources Exploration of state changes and careful binary Exploration of state changes and careful binary

parameter measurements necessary to interpret X-parameter measurements necessary to interpret X-ray dataray data

IR-dominated non-thermal component may probe IR-dominated non-thermal component may probe inner inner accretion flow (is there mid-IR emission in accretion flow (is there mid-IR emission in quiescence?)quiescence?)

Goal: a full time-dependent description of the mass Goal: a full time-dependent description of the mass flowflow

Outburst of 1550-564 in 2000Jain et al. 2001

IR Dominated Flares

Significant O/IR emission from central Significant O/IR emission from central sourcesource

Peaks in mid-IR (Spitzer ToO not yet Peaks in mid-IR (Spitzer ToO not yet activated)activated)

Cannot be thermal and in binary systemCannot be thermal and in binary system Associated with transition to low state, Associated with transition to low state,

QPOs and radio emissionQPOs and radio emission Similar to other “optical plateaus”?Similar to other “optical plateaus”? Energetic synchrotron source from jet??Energetic synchrotron source from jet??

V4641 Sgr Quiescent Lightcurves

X-ray lightcurves 2003

Maitra & Bailyn 2006

Conclusions from Outbursts of V4641 Sgr

Optical/X-ray delay suggests optical is Optical/X-ray delay suggests optical is dominated by reprocessingdominated by reprocessing

If so, variability NOT from Doppler If so, variability NOT from Doppler boosting, but intrinsic (why??)boosting, but intrinsic (why??)

Outburst cycle not like other sources – short Outburst cycle not like other sources – short duration, short recurrence time, no thermal duration, short recurrence time, no thermal statestate

V4641 Sgr = SAX1819.3-2525

Strong black hole Strong black hole candidate: f(m)=6 Mocandidate: f(m)=6 Mo

Most massive, hottest Most massive, hottest 2ndary star (R ~ 13.5 2ndary star (R ~ 13.5 in quiescence!)in quiescence!)

Short, violently Short, violently variable outbursts: variable outbursts: microblazar??microblazar??

Simultaneous X-ray/Optical Observations of V4641 Sgr

Bailyn et al. (2007)

Accretion Disk Instabilities

T

=

Outburst Physics II: X-ray Irradiation

Importance of O/IR Data of (Transient) X-ray Binaries

In quiescence, observe companion stars In quiescence, observe companion stars > binary parameters > binary parameters

In outburst, observe outer parts of disk In outburst, observe outer parts of disk > boundary condition for inner parts of > boundary condition for inner parts of flowflow

Optical and Infrared Lightcurves of Soft X-ray Transients

Theoretical ExpectationsTheoretical Expectations Observational CapabilitiesObservational Capabilities Recent data 1 – Aql X-1Recent data 1 – Aql X-1 Recent data 2 – 4U1543-47Recent data 2 – 4U1543-47 Conclusions: outburst physics, triggers, Conclusions: outburst physics, triggers,

future projectsfuture projects


Hysteresis in outburst morphology (also in Hysteresis in outburst morphology (also in 1999 outburst – Maccarone & Coppi)1999 outburst – Maccarone & Coppi)

Slightly softer high/soft state in decline – no Slightly softer high/soft state in decline – no equivalent in outburstequivalent in outburst

Are we seeing the heated neutron star Are we seeing the heated neutron star surface at the end of the outburst??surface at the end of the outburst??

Conclusions I: Outburst Mechanisms

D.I.M. + irradiation + 2-part flow works for D.I.M. + irradiation + 2-part flow works for some outburstssome outbursts

L.I.S. and mini-outbursts in Aql X-1 L.I.S. and mini-outbursts in Aql X-1 Hysteresis in X-ray states in Aql X-1Hysteresis in X-ray states in Aql X-1 IR-strong reflares in 1543-47 and 1550-564IR-strong reflares in 1543-47 and 1550-564

MORE PHYSICS REQUIRED!MORE PHYSICS REQUIRED!

Conclusions II: Triggers

Optical triggers for new outbursts Optical triggers for new outbursts lead time: 1 week lead time: 1 week especially useful for repeating outbursts especially useful for repeating outbursts should help to get rise as well as fall should help to get rise as well as fall

IR triggers for reflares IR triggers for reflares requires real-time reduction of IR data requires real-time reduction of IR data detailed radio/X-ray response not yet detailed radio/X-ray response not yet knownknown

Conclusions III: Future Work

Daily SMARTS data for Daily SMARTS data for Aql X-1, GX339-4, GRS1915+105, Aql X-1, GX339-4, GRS1915+105, Cen X-4, A0620-00, GS1124-68, Cen X-4, A0620-00, GS1124-68, GRO 1655-40, XTE 1550-564, GRO 1655-40, XTE 1550-564, 4U1543-47 4U1543-47 NEW SOURCES!NEW SOURCES!

Monitoring spectroscopy would be nice!Monitoring spectroscopy would be nice! As would short timescale photometry As would short timescale photometry

A Technical Difficulty

Light dominated by accretion flow

Transient Systems

OutburstsOutbursts Rise time: days Rise time: days Decay time: monthsDecay time: months Recurrence time: decadesRecurrence time: decades Peak luminosity: brightest X-ray sourcesPeak luminosity: brightest X-ray sources

Quiescence barely detectable in X-raysQuiescence barely detectable in X-rays

Black Hole Mass Distribution(update to Bailyn et al. 1998)

Eleven between 5-10 solar massesEleven between 5-10 solar masses Two >10 solar massesTwo >10 solar masses One <5 solar massesOne <5 solar masses ALL neutron stars <2 solar massesALL neutron stars <2 solar masses Selection effects unlikelySelection effects unlikely Mass gap from supernova events?Mass gap from supernova events?

Disks vs. ADAFs

Keplerian Accretion Disk: half energy Keplerian Accretion Disk: half energy dissipated in disk, half at boundary layerdissipated in disk, half at boundary layer

Advection Dominated Accretion FlowsAdvection Dominated Accretion Flows Two T plasma: ions hot, electrons coolTwo T plasma: ions hot, electrons cool Thermal, kinetic energy advected inwardsThermal, kinetic energy advected inwards Requires low mass accretion rateRequires low mass accretion rate

Quiescent transients fit outer disk + ADAFQuiescent transients fit outer disk + ADAF


Narayan, McClintock & Yi 1996

X-ray Binaries

GRO J1655-40 (picture courtesy Rob Hynes)

QPOs in Black Holes

Remillard et al. 2002

The Innermost Stable Orbit

Holds promise of measuring black hole spin!Holds promise of measuring black hole spin! Thermal Disk Models: superposition of Thermal Disk Models: superposition of

blackbodies – Rin a parameter of the fitblackbodies – Rin a parameter of the fit Quasi-Periodic Oscillations: direct measure Quasi-Periodic Oscillations: direct measure

of frequency of inner disk?of frequency of inner disk?

Both methods model dependent!Both methods model dependent!

Black Hole Mass Distribution

Lessons for Moderate Aperture Telescopes (2-5m) in the GMT Era

Should have Should have fewerfewer instruments (but better) instruments (but better) Should have Should have fewer fewer projects (but bigger)projects (but bigger) To maintain department/consortium access To maintain department/consortium access

to a full range of capabilities, we will have to a full range of capabilities, we will have to to trade timetrade time across mountains across mountains

observing black holes with 1m-class telescopes charles bailyn yale university with thanks to: j....

Documents

black holewait

black holel

black holesreversing

black holesif

observedaccreting black

winnickobserving black

strong field

field relativity