Spectral Variability of Cyg X-1 at Energies Above 1 MeV M. McConnell, J. Ryan University of New Hampshire, Durham, NH W. Collmar, V. Schönfelder, H. Steinle, A. Strong Max Planck Institute (MPE), Garching, Germany H. Bloemen, W. Hermsen, L. Kuiper SRON - Utrecht, Utrecht, Netherlands K. Bennett Astrophysics Division, ESTEC, Noordwijk, Netherlands B. Phlips Universities Space Research Association, Washington, DC J. Poutanen Stockholm Observatory, Saltsjbaden, Sweden A.A. Zdziarski Copernicus Astronomical Center, Warsaw, Poland AAS-HEAD Meeting Honoloulu, HI 5-10 November 2000 1
Spectral States of Galactic Black Holes The behavior of Cygnus X-1 is much like that seen in other galactic black hole sources. high state LOW STATE breaking g-ray state low soft X-ray flux high hard X-ray flux hard X-ray spectrum low state HIGH STATE power-law g-ray state high soft X-ray flux low hard X-ray flux soft X-ray spectrum The nature of the variability at energies above 1 MeV has not been clearly established. (from Grove et al. 1998) 2
Long-Term Variability of X-Ray Flux -1.5 These data cover nearly the entire CGRO mission. -2.0-2.5-3.0-3.5 Spectral Index BATSE Spectral Index (20-100 kev) -1 ) -4.0 100 80 60 40 20 8500 BATSE Flux (20-100 kev) ASM Count Rate (cts s RSTE ASM Count Rate (2-10 kev) 9000 9500 10000 10500 11000 11500 0.5 0.4 0.3 0.2 0.1 Flux (cm -2 s -1 in 20-100 kev) Truncated Julian Day (TJD) 3
COMPTEL Observations» COMPTEL provides the best data at energies above 1 MeV.» Most COMPTEL data collected during the low X-ray state.» COMPTEL also collected data during two high state periods: CGRO Viewing Period 318.1 February 1-8, 1994. Not seen by COMPTEL. Consistent with extrapolation of hard X-ray spectrum. CGRO Viewing Period 522.5 June 14-25, 1996. Significant signal seen by COMPTEL. Consistent with extrapolation of hard X-ray spectrum. (Level of hard X-ray flux higher than that during VP 318.1.) Here we report on the results from an analysis of high state data collected during VP 522.5 and its comparison with a low state spectrum compiled from several weeks of CGRO data. 4
Low State Spectrum Contemporaneous CGRO Spectra Compiled from data collected during first three years of CGRO mission McConnell et al., in press, ApJ, v544 (20-Nov-2000) 10 1-1 ) -2 sec -1 MeV 10 0 10-1 10-2 10-3 10-4 10-5 Flux (cm 10-6 10-7 10-8 COMPTEL BATSE-EBOP OSSE EGRET thermal / non-thermal model (PS96) 1-component Comptonization model (T94) 2-component Comptonization model (T94) exponentiated power-law model sec -1 MeV-1 ) 2 2 cm -2 (MeV 10-2 10-3 Flux * E 10-4 COMPTEL BATSE-EBOP OSSE EGRET thermal / non-thermal model (PS96) 1-component Comptonization model (T94) 2-component Comptonization model (T94) exponentiated power-law model 10-9 10-10 0.1 1 10 100 Energy (MeV) 10-5 0.1 1 10 100 Energy (MeV) Flux Spectrum E 2 Flux Spectrum 5
Low State Spectrum» Contemporaneous broad-band spectrum using data from BATSE, OSSE and COMPTEL (McConnell et al., 2000).» Data selected for those periods with consistent hard X-ray flux.» Photon spectrum shows evidence for emission out to ~ 5 MeV.» Model fits originally performed in photon space. Recent analysis now incorporates full response information for both BATSE and OSSE.» Standard Comptonization models are inadequate above ~1 MeV.» A hybrid thermal / non-thermal model can provide an acceptable fit. The spectrum requires a non-thermal component at high energies. 6
Hybrid Thermal / Non-Thermal Model The XSPEC model COMPPS has been used to fit the data. Poutanen & Svensson ApJ, 470, 249 (1996) Models the data using an electron spectrum that consists of a thermal (Maxwellian) component plus a non-thermal (power-law) component. The free parameters of the model include : the electron temperature (kt e ) power-law index (p e ) of the non-thermal component low-energy cutoff (g min ) of non-thermal component (g max = 10000) optical depth of the corona (t) amplitude to fit BATSE spectrum (A BATSE ) amplitude to fit OSSE spectrum (A OSSE ) amplitude to fit COMPTEL spectrum (A COMPTEL ) 7
CGRO Viewing Period 522.5 (Target-of-Opportunity high X-ray state)» Soft X-ray increase began on 10 May 1996 (RXTE, 2-12 kev).» Soft X-ray peak flux at 2 Crab on 19 May 1996 (pre-flare ~ 0.5 Crab)» Correlated decrease in hard X-rays (BATSE, 20-200 kev).» CGRO declared a target-of-opportunity (ToO) on June 13.» CGRO pointing (OSSE, COMPTEL, EGRET) began on June 14.» CGRO Z-axis pointed 5 from Cygnus X-1.» ToO observation (CGRO viewing period 522.5) lasted 11 days. This high state period is clearly seen in the X-ray time history (panel 3) between TJD 10200 and TJD 10350. 8
COMPTEL Imaging - VP 522.5 The 1-3 MeV COMPTEL image exhibited an unusually strong signal. No signal was visible at lower energies (0.75-1 MeV). This fact alone suggested that something unusual was taking place. 10 8 6 4 2 0-2 Galactic Latitude (degs) PSR 1951+32 u u Cyg X-1 Likelihood Map -4 78 76 74 72 70 68 66 64 Galactic Longitude (degs) Location Contour Map (note different scale) 9
High State Spectrum - VP 522.5 10 1 Flux Spectrum -2 sec -1 MeV -1 ) Flux (cm 10 0 10-1 10-2 10-3 10-4 COMPTEL OSSE BATSE High State Low State 10-5 10-6 CGRO VP 522.5 COMPPS Model fit with 100 kev threshold free relative normalizations 10-7 0.01 0.1 1 10 Energy (MeV) 10
High State Spectrum - VP 522.5 E 2 Flux Spectrum 10-2 COMPTEL BATSE OSSE -1 ) 2-2 -1 cm sec (MeV MeV 10-3 High State Low State 2 10-4 Flux * E CGRO VP 522.5 COMPPS Model fit with 100 kev threshold free relative normalizations 10-5 0.1 1 10 Energy (MeV) 11
High State Spectrum» A power-law with index of -2.6 provides a good fit to the data, with the power-law extending to at least 10 MeV.» The data is also well fit with a hybrid thermal / non-thermal model.» Both model fits, unlike that for the low state, require a large (factor of ~1/3) normalization of the COMPTEL data.» The reason for this large normalization is not entirely understood. Here we assume this to represent a response normalization. The high energy power-law is inconsistent with emission from bulk motion Comptonization, which predicts a cutoff near 500 kev. 12
Low State vs.. High State hybrid model fits to data > 100 kev Parameter Low State High State kt 64 ± 4 kev 66 ± 9 kev p e 5.0 ± 0.3 3.8 ± 0.3 g min 1.6 1.8 t 2.6 ± 0.6 0.6 ± 0.4 A BATSE 1.02 4.62 A OSSE 0.87 3.79 A COMPTEL 1.08 10.21 Errors represent estimated 90% confidence levels. These data are preliminary. A more complete analysis is in progress. In the context of the hybrid model, the high state spectrum shows: 1) smaller optical depth 2) harder non-thermal electron component 3) same electron temperature 13
Physical Interpretation Poutanen & Coppi (1998) HIGH X-RAY STATE HIGH STATE smaller optical depth smaller thermal component non-thermal component dominates LOW X-RAY STATE LOW STATE larger optical depth larger thermal component thermal component dominates 14
Summary» Composite CGRO spectra for both the low and high X-ray states.» The spectra exhibit bimodal spectral behavior, as seen in other galactic black hole candidates, with pivot point near 1 MeV.» Power-law spectrum of high state spectrum extends to at least 10 MeV, with no evidence for any cutoff.» This is inconsistent with models that describe the emission in terms of bulk motion Comptonization, which predict a cutoff near 500 kev.» A hybrid thermal/non-thermal model can describe the data.» The results are consistent with the model of Poutanen & Coppi (1998), although a predicted decrease in kt e for the high state is not seen. 15