Sgr A* in Context: Daily Flares as a Probe of the Fundamental X-ray Emission Process in Accreting Black Holes

Our central Galactic supermassive black hole, Sgr A*, exists mostly in a very stable, extremely low-luminosity (∼ 10LEdd), thermal quiescent state, which is interrupted roughly daily by a brief, nonthermal X-ray flare. Because they are not accompanied by significant changes in the radio wavelengths, the flares make Sgr A* unusual in the context of black holes accreting at slightly higher rates. Those sources display a radio/X-ray luminosity correlation whose normalization scales with central mass, and that holds over orders of magnitude in accretion power. There is significant scatter in this correlation, due in part to measurement uncertainties and intrinsic variability. By studying the correlation in sources bracketing Sgr A* in radio luminosity and whose physical parameters are well measured, we can derive a statistical measure of this local scatter. We find that Sgr A* in quiescence and the lower intensity flares fall well below the correlation in X-ray luminosity. The brightest flares are consistent within the scatter, which may indicate an upper bound on the X-ray luminosity. This trend is suggestive of a state transition at the extreme low end of accretion activity, only above which the radio/X-ray correlation is tracked. This scenario is easily testable because it must fulfill three unique observational predictions: 1) As long as Sgr A* remains at its current radio luminosity, no X-ray flare will be seen which statistically exceeds the prediction of the correlation, 2) no source already on the correlation will be seen to flare in the X-rays similar to Sgr A* (i.e., without corresponding increases in the radio luminosity), and 3) sources below a critical accretion rate or luminosity will show similar flares as Sgr A*, on timescales appropriate to their masses. Subject headings: black hole physics—Galaxy: center—radiation mechanisms: non-thermal—accretion, accretion disks—X-rays: general 1. CORRELATIONS IN SGR A* VS. OTHER BLACK HOLES The anomalously low luminosity of Sgr A* (∼ 10LEdd; Melia & Falcke 2001) has puzzled researchers for over two decades, raising questions about its relationship to other, more typical, active nuclei. It seems unlikely that Sgr A* is the only one of its kind; if it simply represents the lowest end of the luminosity scale, its behavior should map onto trends we detect in other accreting black hole sources. In this Letter, we propose three observationally verifiable predictions to probe Sgr A*’s relationship to more canonical black hole sources. Sgr A*’s proximity (8 kpc; Reid 1993; Eisenhauer et al. 2003) has resulted in the constraining of its physical parameters better than almost any other galactic nucleus, with the exception of NGC 4258 (e.g., Herrnstein 1997). Studies of Sgr A*’s orbiting central cluster stars reveal a 4× 10M⊙ mass (Schödel et al. 2003; Ghez et al. 2003), which along with the similarly well-constrained distance, allows us to assess Sgr A*’s relationship to other sources with known parameters. Despite Sgr A*’s extremely weak high-energy activity, its radio characteristics are typical of other lowluminosity AGN (LLAGN), M81* in particular (e.g., Ho 1999; Brunthaler et al. 2001; Bower et al. 2002b). Its steady X-ray spectrum (LX ∼ 2× 1033 erg/s) is soft (Γ ∼ 2.7) and extended, arguing for a thermal origin (Baganoff et al. 2003). Recent theoretical models developed to explain its behavior have focused variously on inflow scenarios (e.g., Liu & Melia 2002; Yuan et al. 2003), outflow scenarios (Falcke & Markoff 2000) and combinations of the two (Yuan et al. 2002). Most of these models have been applied to other low-luminosity Electronic address: [email protected] 1 NSF Astronomy & Astrophysics Postdoctoral Fellow sources with some success, indicating that Sgr A* shares many characteristics with its brighter cousins, but may just represent the most underluminous extreme. Sgr A* showed the first signs of AGN-like activity in the second cycle of Chandra observations, with a dramatic (∼50x increase) nonthermal, hard flare on timescales of tens of minutes (Γ ∼ 1.3; Baganoff et al. 2001). Further observations have established that flaring occurs about once a day, with typical increases of 5–10x in flux (Baganoff 2003). The cmradio emission, however, has not yet been seen to vary by more than a factor of a few (Melia & Falcke 2001). While the multiwavelength variability characteristics have not yet been fully determined, the “submm bump” in Sgr A*’s spectrum, which includes the IR band, is clearly related to the flaring X-ray component (e.g. Eckart et al. 2004). The physical origin is still being debated (see articles in, e.g., Cotera et al. 2004), however most quiescent-state models for Sgr A* can be adapted to explain the flares. A consensus has formed that the submm/IR variability is due to synchrotron emission from mildly relativistic, quasi-thermal electrons very close to the central object, while the X-ray flares are due to either a continuation of this synchrotron emission due to a hard tail in the distribution, synchrotron self-Comptonized emission (SSC) or combinations of the two (e.g., Markoff et al. 2001; Liu & Melia 2002; Yuan et al. 2004). These magnetic mechanisms dominate in what seems to be the absence of a canonical (Shakura & Sunyaev 1973) thin disk (Falcke & Melia 1997; Liu et al. 2004). Often low-luminosity, accreting X-ray binaries (XRBs) in their low/hard state (LHS; see McClintock & Remillard 2003) are associated with compact jets and share a general morphology with LLAGN. A correlation between the radio and X-ray