EDP Sciences
Free Access
Issue
A&A
Volume 581, September 2015
Article Number L9
Number of page(s) 5
Section Letters
DOI https://doi.org/10.1051/0004-6361/201527043
Published online 23 September 2015

© ESO, 2015

1. Introduction

V404 Cyg (hereafter V404) is a low-mass X-ray binary (LMXB) consisting of a black hole (BH) with mass estimates ranging from ~9 to 15M, and a M K3 III companion (Casares & Charles 1994; Shahbaz et al. 1994; Khargharia et al. 2010), located at a parallax distance 2.39 ± 0.14 kpc (Miller-Jones et al. 2009). The inclination of the binary’s rotational axis is 67° (Shahbaz et al. 1994; Khargharia et al. 2010), the orbital period 6.5 d (Casares et al. 1992). This transient underwent three periods of outbursts during the twentieth century (Richter 1989), the last, in May 1989, leading to its discovery as an X-ray transient by the Ginga satellite (as GS 2023+338, Makino et al. 1989). V404 showed bright X-ray flares on short time-scales (e.g. Makino et al. 1989; Terada et al. 1994), which makes it an excellent source to study the connections between the accretion and ejection phenomena, which are the probable origin of this behaviour. V404 is one of the closest stellar mass BHs, making it a rare case where quiescence can be studied in detail. Variable remnant activity, attributed to a compact jet, was detected from radio to hard X-rays (e.g. Hynes et al. 2004; Xie et al. 2014). V404 is one of the few sources that defines the radio/X-ray correlation over a wide range of luminosities, down into quiescence (Corbel et al. 2008). The good knowledge of the quiescent state makes understanding new outburst observations paramount as they allow the mechanisms responsible for the increased activity to be probed.

thumbnail Fig. 1

Left: INTEGRAL LCs of V404 in four spectral domains (a larger version of the plot including all energy ranges is available as Fig. 4). Right: 20–40 keV count rate (top), and 3–13/20–40 keV softness ratio (SR, bottom). The dashed horizontal line corresponds to 1 Crab (20–40 keV). MJD 57 193 is 2015 June 20.

Open with DEXTER

On 2015 June 15 (MJD 57 188), V404 went into outburst again. It was first detected by Swift (BAT and XRT) (Barthelmy et al. 2015) and then with MAXI and INTEGRAL (Negoro et al. 2015; Kuulkers et al. 2015). These early alerts triggered follow-up observations at all wavelengths. Preliminary results all report the detection of the source, variations of specific spectral features, and an extreme flaring activity at all wavelengths (e.g. Mooley et al. 2015; Munoz-Darias et al. 2015; Ferrigno et al. 2015; Motta et al. 2015; Tetarenko et al. 2015a,b; Tsubono et al. 2015). We triggered our INTEGRAL ToO program to obtain quasi-continuous coverage in X/γ-rays and in the optical V band. The first detection of multiple flares exceeding 30 Crab in 20–40 keV and possible correlated flaring in the V band were reported in Rodriguez et al. (2015a) and Domingo et al. (2015).

Our observations caught the source during the most intense and variable phase of this new outburst. Here, we first give the details of the observations and data reduction (Sect. 2). We then focus on the flaring behaviour at high energies (Sect. 3), which we compare to the activity in the optical V band (Sect. 4). A preliminary phenomenological spectral characterization of different intensity intervals is presented in Sect. 5. We discuss our results and compare V404 to other microquasars in Sect. 6.

2. Observations and data reduction

Our ToO program (Fig. 1) covered MJD 57 193.66–57 198.17 (2015-06-20 15:50 UTC to 2015-06-25 4:05 UTC), that is, INTEGRAL revolutions 1555 (continuous coverage) and 1556 (two periods). The data of all the INTEGRAL instruments (see Winkler et al. 2003, and references therein for all instrumental details) were reduced with the Off line Scientific Analysis (OSA) v10.1 software suite, with the latest calibration files available at the time of writing.

Images and 100 s binned light curves (LC) from the Joint European X-ray Monitors (JEM-X) and the Imager on Board the INTEGRAL Satellite (IBIS) were produced in two bands (3–13, and 13–30  keV) for JEM-X unit 1, and in four bands (20–40, 40–80, 80–150, and 150–300 keV) for the IBIS Soft Gamma-Ray Imager (ISGRI).

The event data of the Spectrometer on INTEGRAL (SPI) were fitted with models for the celestial sources and instrumental background following standard reduction processes. The 20–100 keV LC of V404  as well as the other sources in the field were obtained in bins of 400 s. Background models were built based on the pre-flaring data of a representative empty sky region, adjusting the normalization coefficient per hour (see, e.g., Strong et al. 2005, for a more general description of the method).

As source intensity and hardness vary strongly on short time-scales, we extracted luminosity/hardness dependent JEM-X, ISGRI, and SPI spectra over specific time intervals of clean data. The spectra from the same time intervals were jointly fitted within XSPEC v12.8.2. Since the instruments’ responses are possibly different for the high intensities observed, only phenomenological spectral fits are presented, and the fit results should be viewed with some caution.

The INTEGRAL/Optical Monitoring Camera (OMC) fluxes and magnitudes were derived from a photometric aperture of 3 × 3 pixels (1 pix. = 17.504), slightly circularized, that is, removing one quarter pixel from each corner (standard output from OSA). The photometric aperture was centred on the source coordinates (default centroid algorithm) and did not include any significant contribution from other objects. We removed measurements with a severe problem flag, and, to restrict the noise, only measurements of 50 and 200 s duration were considered.

3. Model-independent description of the flaring

Multi-wavelength LCs of V404  from the V band up to γ-rays are highly structured with several large flares separated by calmer periods seen in all bands (Fig. 1, and see also Fig. 4 for a plot with all energy ranges). In the following, count rates (CR) are given in the ISGRI 20–40 keV range. When the source CR increased above ~150–200 cts/s, an intense X-ray flare systematically followed. In the following, we thus set 1 Crab1 as the typical limit between the off-flare and flaring intervals. We identified 18 main events, that is, peaks that reached at least 6 Crab (labelled with Roman numerals2 in Fig. 1, their main characteristics are given in Table 1), with 11 exceeding 20 Crab during our observations. Flares IV, XI, and XIII are the brightest we observed, reaching 43 Crab. The flares occurred isolated (e.g. III, IV, VI) as well as in groups with peak-to-peak intervals as short as 22 min (Va, Vb) The flares lasted 0.4–2.4 h, except for peaks IV and XIII. The former shows a rather broad profile and has multiple peaks. This event lasted 4.8 h in total and is the longest flare of our observation. The latter reached about 40 Crab. The peak itself lasted about 1.5 h, but was preceded by a ~3 h long, 3 Crab plateau seen only above 13 keV. It was followed by flares XIV and XV, which show decreasing peak values.

The 3–13/20–40 keV softness ratio (SR, Fig. 1, right) shows that the strong variability of the source is associated with variations of SR from ~0.03 to ~1.3 corresponding to Γ = 0.1–2.5 in simulated JEM-X/ISGRI power-law spectra. SR ~ 0.6 corresponds to Γ = 2.0. Strong spectral variations are visible in the off-flare intervals (Fig. 1, right). All the flares are hard, and all have SR < 0.4 (Γ< 1.8).

thumbnail Fig. 2

V-band (green), 3–13 keV (red), and 20–40 keV (blue) LCs around flares I a) and X b). The inserts show the cross-correlation functions of the 3–13 keV (red) and 20–40 keV (blue) vs. the optical LCs over the same time intervals as the LCs. The dashed vertical lines represent the 0 lag level.

Open with DEXTER

4. Optical vs. X-ray behaviour

The comparison of the optical (OMC) and X-ray (JEM-X1 and ISGRI) LCs shows a non-trivial relationship. Significant flaring activity is evident in the V band LC (Fig. 1, left), with at least 12 clear flares. The optical flare typically lasted 0.24–2.5 h. While some events occurred in simultaneity with X-ray flares, the optical emission was delayed with respect to the X-rays in other cases. Figure 2 shows typical examples of these different behaviours. The cross-correlation function (ccf) between the X-ray and optical emission confirms the absence of lags for some of the flares (e.g. flare I, which causes the peak at 0 in the ccf of Fig. 2a), and delayed optical emission from 1.5 min to 20–30 min is seen in others. The ccf of Fig. 2b shows an example of a ~3 min lag, while the ccf of Fig. 2a (in addition to showing the simultaneity of peak I and its optical counterpart) shows lag at 20–30 min, representing the delay between the small X-ray flares preceding peak I (around MJD 53 193.9) and the subsequent optical flare (around MJD 53 193.92). Because of the time resolution of OMC, however, lags shorter than 1  min cannot be measured with our data, and additional lags of the order of seconds are not excluded.

While most of the flares show a fast rise similar to the flares observed in X-rays, the two optical events occurring close to X-ray peaks IV and XIII seem to be exceptional. All other flares show fast rises (1 h), but these two events have slower rises (about 10 and 4 h, respectively), and are both coincident with hard plateaus that precede the X-ray peaks.

5. Spectral analysis

We accumulated spectra from the brightest flares (CR > 1000 cts/s) and the off-flare intervals (CR < 165 cts/s). In the latter case we also only retained the hard intervals (SR < 0.6) in order to exclude the softening visible after MJD 57 197.9 (Fig. 1). The resulting νfν spectra are plotted in Fig. 3.

The off-flare spectrum is well fitted ( = 1.2, 66 degrees of freedom, d.o.f.) by a model consisting of a power law with a high-energy cut-off dominating at 10–100 keV plus an additional power law dominating above 100 keV. The former component has Γ = 1.0, keV, Efold = 23 ± 5 keV, the latter has Γ = 1.9 (Γ is the photon index defined as N(E) ∝ E− Γ.) Normalization constants were included to account for potential cross-calibration issues or differences in the effective exposures (deadtime corrections, telemetry drop out). When set to 1 for ISGRI, we obtain ~1.9 for SPI and ~0.6 for JEM-X1. The 20–400 keV (ISGRI) flux is ~10-8  erg/cm2/s, and the above model leads to an extrapolated 0.1–105 keV flux ~3.8 × 10-8 erg/cm2/s, that is, about 2%  LEdd for a 9 M BH.

The flare spectrum is well represented (~ 0.9, 78 d.o.f.) by a single cut-off power law with Γ = 1.54 ± 0.06, Ecut = 14.0 keV, keV. An extra power-law component is not statistically required according to an F-test. The normalization constants are both close to 1.1 The 20–400 keV (ISGRI) flux is ~10-7 erg/cm2/s, which leads to an extrapolated 0.1–105 keV flux ~3 10-7 erg/cm2/s, or about 20% LEdd for a 9 M BH.

6. Discussion

Over the four days covered by our INTEGRAL ToO, V404 showed a high level of emission with sporadic flares with a maximum 20–40 keV dynamical range of 940 (flare XVI). During its flares, V404 became the brightest X-ray object in the sky. In the hard off-flare state the spectral analysis shows two spectral components: a cut-off power law typically attributed to thermal Comptonization and an extra power law at energies beyond 100 keV. Hard tails have now been seen in a large number of systems (e.g. GRS 1915+105, Swift J1753.50127, GX 3394, or Cyg X-1, Rodriguez et al. 2008b, 2015b; Tomsick et al. 2015; Joinet et al. 2007), and their origin is still highly debated, although a compact jet origin is favoured in the case of Cyg X-1 (e.g. Russell & Shahbaz 2014; Rodriguez et al. 2015b). It is interesting that the flaring activity seems primarily due only to spectral variations of the cut-off power law. We estimate an integrated luminosity L ~ 0.2 LEdd for the >6 Crab flares. This is a lower limit to the maximum luminosity reached at the peak of the brightest flares. First because we averaged the data, without isolating the brightest portions of the flares. Moreover, we did not consider all the contributions below 10 keV (disk, jet) that can provide a significant fraction of the bolometric luminosity. Assuming a simple scaling between the CR and LEdd with a constant shape for the variable component between the off-flare hard state and the flares, we conclude that all peaks with CR  3000 cts/s (~18 Crab) (flares III, IV, Va,b, VIII, X, XI, XIIa,b, XIII and XIV, Fig. 1) reached ~LEdd.

thumbnail Fig. 3

νfν spectra obtained from the fits to the off-flare and flare intervals. The lines represent the spectral components used for the fits.

Open with DEXTER

The optical activity is also highly variable and shows flares (Figs. 1 and 2). Some optical flares occur in conjunction with the X-rays, while other activity periods show delays (Fig. 2). The first specific length scale of this system is the separation between the BH and the companion: using the system parameters from Sect. 1 and a 9 M BH, we estimate ~2.2 × 1012 cm or ~75 light seconds. Hence, when no delay between the optical and the X-rays is observed (e.g. flare I in Fig. 2a), the mechanism producing the optical emission could be related to X-ray reprocessing, either by an accretion disk or by the companion. The maximum delays expected would be around 60 s (outer disk), and 150 s (companion located at superior conjunction).

Optical lags 10 min could be related to variable jet properties, either as their intrinsic synchrotron emission, or from their interaction with the surrounding medium. Radio and millimeter (mm) flaring activity ascribed to discrete ejections has been reported during this outburst (e.g. Sivakoff et al. 2015; Mooley et al. 2015; Tetarenko et al. 2015b). Delays between X-rays and longer wavelengths are expected in the case of adiabatically expanding ejecta (van der Laan 1966; Mirabel et al. 1998), and Tetarenko et al. (2015b) reported a mm peak on MJD 57 195.548, about 26 min after the MJD 57 195.53 X-ray event (a <6 Crab flare that occurred slightly after flare XII). Interestingly, this delay also corresponds to the time scale of the mm flare increase (Tetarenko et al. 2015b), which renders the mm compatible with being causally related to the X-ray flare. As in other well-known sources, e.g. XTE J1550−564, GRO J1655−40, GRS 1915+105 (Fender 2006), discrete ejections may be causally connected to the X-ray activity in V404.

V404 reached maximum absolute peak intensities that are rather usual during outbursts of microquasars. If V404 was at 10 kpc, the maximum peak intensity would have been 2.5 Crab, a value similar to the aforementioned other microquasars during their brightest outburst(s) (e.g. Remillard & McClintock 2006). Short recurrent (multi-wavelength) flares have been seen only in GRS 1915+105 (e.g. Greiner et al. 1996), however, and in V404 the flaring activity similarly recurs on time scales as short as 22 min. Some of the V404 optical flares lag 20 min behind the X-rays (Fig. 2a), and similar lags are also seen at mm-radio wavelengths. This may resemble the correlated X-ray/infrared/radio oscillations also referred to as 30-min cycles of GRS 1915+105  (e.g. Fender & Pooley 1998; Mirabel et al. 1998; Rodriguez et al. 2008a). In GRS 1915+105, however, these events are associated with hard X-ray dips preceding the flares and a clear softening at the X-ray peak, marking the disappearance of the Comptonization component (Rodriguez et al. 2008b). V404, in contrast, remains hard even in the flaring states (Fig. 1), indicating that a different mechanism is responsible for the X-ray flaring (similar results were obtained from the 1989 outburst, e.g. Życki et al. 1999). One tempting possibility would be that the high-energy flares are direct boosted emission from a jet (blazar-like configuration). This would imply a jet axis not perpendicular to the orbital plane. Misaligned jets have been seen in GRO J165540 and V4641 Sgr (Maccarone 2002,and references therein). In the former, a rather modest Lorentz factor γ ~ 2.5 implies weak relativistic boosting (Hjellming & Rupen 1995). In the latter, γ ranges from 10 up to 17, and the angle between the jet axis and the orbital plane normal is as high as 50° (Maccarone 2002,and references therein). Significant Doppler boosting is expected in this case. Życki et al. (1999) argued that the spectral and intensity variability seen with Ginga in 1989 could be due to the evolution of a heavily absorbing medium. However, even with NH ≳ 1024 cm-2 (Życki et al. 1999), the activity above 20 keV is not affected by absorption, and hence the absorber alone cannot be responsible for the strong variability we observe. The

high-energy flares could be due to the shock of the relativistic jets with the dense ambient medium. Then optically thin synchrotron emission would be expected at X-ray energies, while our analysis favours thermal Comptonization. More simultaneous multi-wavelength observations will help distinguish these different possibilities.


1

The ISGRI/20–40 keV CR of the Crab is 165 cts/s F20−40 keV = 8 × 10-9 erg/cm2/s for a power-law spectrum with Γ = 2.1 and a normalization of 10 ph/cm-2/s at 1 keV.

2

V and XII contain two distinct events that are hardly distinguishable in Fig. 1. They appear under the same label in Fig. 1 (to keep it clear), and are named with a/b sub-labels in the text and Table 1.

Acknowledgments

We warmly thank the referee for the useful report that helped us to improve the quality of this paper. We also thank the INTEGRAL teams and planners for their prompt reaction and the scheduling of these observations. J.R., M.C., S.C. acknowledge funding support from the French Research National Agency: CHAOS project ANR-12-BS05-0009 (http://www.chaos-project.fr), and from the UnivEarthS Labex program of Sorbonne Paris Cité (ANR-10-LABX-0023 and ANR-11-IDEX-0005-02). XLZ acknowledges funding through DLR 50 OG 1101. M.G.H.K. was supported by the Deutsche Forschungsgemeinschaft under DFG project number PR 569/10-1 in the context of the Priority Program 1573 Physics of the Interstellar Medium. This work was supported by NASA through the Smithsonian Astrophysical Observatory (SAO) contract SV3-73016 for the Chandra X-Ray Center and Science Instruments. R.D. and X.L.Z. acknowledge support through the Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR) 50 OG 1101. O.M.C. activities are supported by Spanish MINECO grant ESP2014-59789-P. This study is based on observations made with INTEGRAL, an ESA project with instruments and science data centre funded by ESA member states, Poland and with the participation of Russia and the USA.

References

Online material

Table 1

List of the >6 Crab flares and their main properties.

thumbnail Fig. 4

INTEGRAL LCs of V404 over our ~4-day-long observations. All spectral domains considered for the LC extraction are shown here. The dashed line in the 20–40 keV panel represents the approximate level of LEdd we estimated. MJD 57 193 is 2015 June 20.

Open with DEXTER

All Tables

Table 1

List of the >6 Crab flares and their main properties.

All Figures

thumbnail Fig. 1

Left: INTEGRAL LCs of V404 in four spectral domains (a larger version of the plot including all energy ranges is available as Fig. 4). Right: 20–40 keV count rate (top), and 3–13/20–40 keV softness ratio (SR, bottom). The dashed horizontal line corresponds to 1 Crab (20–40 keV). MJD 57 193 is 2015 June 20.

Open with DEXTER
In the text
thumbnail Fig. 2

V-band (green), 3–13 keV (red), and 20–40 keV (blue) LCs around flares I a) and X b). The inserts show the cross-correlation functions of the 3–13 keV (red) and 20–40 keV (blue) vs. the optical LCs over the same time intervals as the LCs. The dashed vertical lines represent the 0 lag level.

Open with DEXTER
In the text
thumbnail Fig. 3

νfν spectra obtained from the fits to the off-flare and flare intervals. The lines represent the spectral components used for the fits.

Open with DEXTER
In the text
thumbnail Fig. 4

INTEGRAL LCs of V404 over our ~4-day-long observations. All spectral domains considered for the LC extraction are shown here. The dashed line in the 20–40 keV panel represents the approximate level of LEdd we estimated. MJD 57 193 is 2015 June 20.

Open with DEXTER
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.