1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[12pt,preprint]{emulateapj}
3: \notetoeditor{refer to: Fiamma Capitanio. Institute: IASF-INAF,Via Fosso del Cavaliere, 100, 00133 Rome ITALY. Tel: +390649934557. E-mail: fiamma.capitanio@rm.iasf.cnr.it}
4: \newcommand{\inst}{\altaffilmark}
5: \newcommand{\instM}{\altaffiltext}
6:
7: %\documentclass{aa}
8: %\documentclass[psfig,referee]{aa} % for a referee version
9: \usepackage{graphicx}
10: %\usepackage{txfonts}
11: \usepackage{natbib}
12: \shorttitle{IGR J17091--3624 and IGR J17098--3628: a multiwavelength long term campaign}
13: \shortauthors{Capitanio F. et al.}
14: \begin{document}
15: %\authorrunning{F. Capitanio et al.} \titlerunning{Spectral states of the new X-ray binary IGR J17098--3628 observed by {\it INTEGRAL} and XRT}
16: \title{The two INTEGRAL X-ray transients IGR J17091--3624 and IGR J17098--3628: a multi-wavelength long term campaign}
17: % \subtitle{}
18:
19: \author {F. Capitanio\inst{1}, M. Giroletti\inst{2}, M. Molina\inst{3},
20: A. Bazzano\inst{1}, A. Tarana\inst{1}, J. Kennea\inst{4}, A. J. Dean\inst{3},
21: A. B. Hill\inst{3}, M. Tavani\inst{1}, P. Ubertini\inst{1}}
22:
23: \altaffiltext{1}{INAF IASF-Roma, Via Fosso del Cavaliere 100, 00033 Rome,
24: Italy}
25:
26: \altaffiltext{2}{INAF Istituto di Radioastronomia, via Gobetti 101, 40129
27: Bologna, Italy}
28:
29: \altaffiltext{3}{School of Physics and Astronomy, University of Southampton,
30: Highfield Southampton, SO17 1BJ, UK}
31:
32: \altaffiltext{4}{Department of Astronomy and Astrophysics, Pennsylvania State
33: University, University Park, PA 16802, Pennsylvania, USA}
34:
35: %\altaffiltext{4}{ IASF/INAF, Bologna, Via Godetti 101, I-40129 Bologna, Italy}
36: % \date{Received; accepted }
37:
38:
39: \begin{abstract}
40:
41: IGR J17091--3624 and IGR J17098--3628 are two X-ray transients discovered by
42: {\it INTEGRAL} and classified as possible black hole candidates (BHCs). We
43: present here the results obtained from the analysis of multi-wavelength data
44: sets collected by different instruments from 2005 until the end of 2007 on both
45: sources. IGR J17098--3628 has been regularly detected by {\it INTEGRAL} and
46: {\it RXTE} over the entire period of the observational campaign; it was also
47: observed with pointed observations by {\it XMM} and {\it Swift}/XRT in 2005 and
48: 2006 and exhibited flux variations not linked with the change of any
49: particular spectral features. IGR J17091--3624 was initially in quiescence
50: (after a period of activity between 2003 April and 2004 April) and it was then
51: detected again in outburst in the XRT field of view during a {\it Swift}
52: observation of IGR J17098--3628 on 2007 July 9. The observations during
53: quiescence provide an upper limit to the 0.2-10 keV luminosity, while the
54: observations in outburst cover the transition from the hard to the soft
55: state. Moreover, we obtain a refined X-ray position for IGR J17091--3624 from
56: the {\it Swift}/XRT observations during the outburst in 2007. The new position
57: is inconsistent with the previously proposed radio counterpart. We identify in
58: VLA archive data a compact radio source consistent with the new X-ray position
59: and propose it as the radio counterpart of the X-ray transient.
60:
61: \end{abstract}
62:
63: \keywords{X-rays: binaries -- X-rays: stars -- X-rays: individual: IGR
64: J17098--3628 -- IGR J17091--3624 -- black hole candidate}
65:
66:
67: \section{Introduction}
68:
69: Among the sources included in the {\it INTEGRAL} IBIS/ISGRI survey
70: catalog~\citep{Bird}, the two X-ray transients IGR J17091--3624 and IGR
71: J17098--3628 are remarkable because of their proximity, being only 9\arcmin.6
72: away from each other. The two sources have been detected in different periods
73: of time from April 2003 until the beginning of 2008. In particular, IGR
74: J17091--3624 was detected by IBIS in 2003 and remained detectable for one year
75: and, after a period of quiescence, it was again in outburst in July 2007. IGR
76: J17098--3628 was detected for the first time by {\it INTEGRAL} in 2005 (while
77: IGR J17091--3624 was not visible) and it has then remained detectable with
78: variable flux in the soft X-ray energy range (E $<$ 20 keV) up to now. On the
79: basis of their spectral behavior, both IGR J17091--3624 and IGR J17098--3628
80: are classified as probable BHCs \citep{Luto03,Greb2007}.
81:
82: In this paper, we present the results of a long term monitoring, primarily at
83: high energy, of these two transients, and discuss the identification in
84: archival data of a radio counterpart for IGR J17091--3624.
85: The date, the exposure time, the instruments used and the sources detection of the different observations are summarized in Table~\ref{log}.
86: The paper is organized as follows: in \S\ref{intro1} and \S\ref{intro2}, we introduce the
87: two BHCs; in \S\ref{obs}, we describe our observations; in \S\ref{res}, we
88: present the results; finally, we give a discussion and present our conclusions
89: in \S\ref{disc}.
90:
91: \begin{table}
92: \begin{center}
93: \caption{IGR J17098-3628 and IGR J17091-3624 observations log from 2003 until September 2007}
94: \label{log}
95: \begin{tabular}{lcccc}
96: \hline
97: \hline
98:
99: Start date & End date & Satellite & Detected sources & Exposure \\
100: %2001-09-20 & 2001-10-01 & {\it BeppoSAX} & IGR J17091-3624 & 650 ks & in't Zand 2003 \\
101: 2003-04-12 & 2003-04-21 & {\it INTEGRAL} & IGR J17091-3624\tablenotemark{a} & 15 ks \\
102: 2003-04-20 & \nodata & {\it RXTE} & IGR J17091-3624\tablenotemark{b} & 4.5 ks \\
103: 2003-04-23 & 2003-05-09 & VLA\tablenotemark{c}& IGR J17091/IGR J17098 & 500 s \\
104: 2003-08-10 & 2004-04-02 & {\it INTEGRAL} & IGR J17091-3624\tablenotemark{a} & 154 ks \\
105: 2005-02-20 & 2005-05-01 & {\it INTEGRAL} & IGR J17098-3628\tablenotemark{d} & 300 ks \\
106: 2005-03-29 & \nodata & {\it RXTE} & IGR J17098-3628\tablenotemark{d}& 2.1 ks \\
107: 2005-05-01 & 2007-02-28 & {\it INTEGRAL} & IGR J17098-3628 & 450 ks\\
108: 2005-05-01 & 2005-09-14 & {\it Swift} & IGR J17098-3628 & 8.5 ks \\
109: 2006-08-25 & \nodata & {\it XMM} & IGR J17098-3628 & 8 ks \\
110: 2007-02-19 & \nodata & {\it XMM} & IGR J17098-3628 & 16 ks \\
111: 2007-07-19 & 2007-08-07 & {\it Swift} & IGR J17091/IGR J17098 & 11 ks \\
112: 2007-08-25 & 2007-09-30 & {\it INTEGRAL} & IGR J17091-3624 & 50 ks \\
113: \hline
114: \end{tabular}
115: \tablenotetext{a}{Capitanio et al. 2006}
116: \tablenotetext{b}{$ $Lutovinov \& Revnivtsev 2003}
117: %\tablenotetext{b}{ Lutovinov\& Revnivtsev 2003}
118: \tablenotetext{c}{Rupen et al. 2003}
119: \tablenotetext{d}{Grebenev et al. 2007}
120: \end{center}
121: \end{table}
122:
123:
124: \subsection{IGR J17091--3624} \label{intro1}
125:
126: IGR J17091--3624 was discovered by {\it INTEGRAL}/IBIS during a Galactic
127: Center observation on 2003 April 14--15~\citep{Kuulk}. Initially, the flux was
128: $\sim$20 mCrab in the 40--100 keV energy band exhibiting a hard spectrum, while
129: it was not detected in the 15--40 keV band, with an upper limit of $\sim$10
130: mCrab. During subsequent observations of the Galactic Center Deep Exposure
131: (GCDE) on 2003 April 15--16, the source flux increased to $\sim$40 mCrab in the
132: 40--100 keV band and to 25 mCrab in the 15--40 keV (the IBIS flux statistical
133: error is less then 10\%).
134:
135: Immediately after the {\it INTEGRAL} discovery, providing the position of
136: IGR J17091--3624, an {\it RXTE\/} observation was performed and the source
137: was then searched in the X-ray catalogs. IGR J17091-3624 was found in the
138: archival data of both the TTM telescope on board the KVANT module of the
139: {\it Mir\/} orbital station \citep{Atel2}, and in the {\it BeppoSAX} WFC \citep{Atel4}.
140: A first study of the IBIS/ISGRI spectral evolution of the source~\citep{Luto03, Luto05}
141: showed a source hardening with a photon index changing from $\Gamma = 2.2 \pm 0.1$ to
142: $\Gamma=1.6\pm 0.1$ from 2003 April to 2003 August. A subsequent detailed
143: analysis of the IBIS, JEM-X and {\it RXTE}/PCA data of the entire outburst
144: duration~\citep{Cap2006} revealed an indication of an hysteresis like behavior
145: and the presence of a hot disc black body emission component during the source
146: softening.
147:
148: From the investigations reported above, IGR J17091--3624 appears as a
149: moderately bright variable transient source, with a flaring activity in 1994
150: October ({\it Mir}/KVANT/TTM), 1996 September ({\it BeppoSAX}/WFC), 2001
151: September \citep[{\it BeppoSAX}/WFC,][]{Atel4}, 2003 April \citep[{\it
152: INTEGRAL}/IBIS,][]{Kuulk}, and 2007 July. In this paper, we report on the last
153: episode of activity, as well as on limits on the quiescent state.
154:
155:
156: \subsection{IGR J17098--3628} \label{intro2}
157:
158: IGR J17098--3628 was detected for the first time with {\it INTEGRAL}/IBIS 9.4'
159: off IGR J17091--3624 \citep{Atel444} during deep Open Program observations of
160: the Galactic Center region on 2005 March 24. The average fluxes were 28.2
161: $\pm$ 1.4 and 38.7 $\pm$ 2.8 mCrab in the 18--45 and 45--80 keV bands,
162: respectively. Further analysis \citep{Atel447} reported that the source was
163: evolving in both brightness and spectral shape with an indication of
164: softening.
165:
166: From 2005 March 29 to April 4, an observational campaign was performed with {\it
167: RXTE}/PCA. The spectral shape given by {\it INTEGRAL}/IBIS and {\it RXTE}/PCA
168: varied throughout the observations and it is modelled by a soft black body
169: emission component plus a hard tail and an absorption consistent with the
170: Galactic one~\citep{Greb2007}. The spectral variations suggested that this
171: source was an X-ray nova going in outburst and probably a BHC \citep{Greb2007}.
172:
173: The source was then observed with the {\it Swift} satellite \citep{Atel476}
174: with an exposure time of 2.8 ks on 2005 May 1; it was quite bright, with
175: an estimated flux of 1.3 $\times$ $ 10^{-9}$ ergs s$^{-1}$ cm$^{-2}$ in the 0.5--10 keV
176: energy band. The analysis of the {\it Swift}/XRT data for IGR J17098--3628
177: refined the source coordinates as follows: RA 17$^h$ 09$^m$ 45.9$^s$, Dec --36$^\circ$
178: 27$\arcmin$ 57$\arcsec$ (J2000), with an uncertainty radius of about 5$\arcsec$
179: \citep{Atel476}. This position is 30$\arcsec$ from the {\it INTEGRAL} position
180: reported by \cite{Atel444}.
181:
182: % The source flux appeared to drop by ~8\% over the duration of the observation \cite{Atel476}.
183:
184:
185: Following the soft X-ray detection of IGR J17098--3628 with {\it Swift}/XRT, a
186: probable radio counterpart has also been found \citep{Atel490}. In particular,
187: the first data set of four consecutive Very Large Array (VLA) radio
188: observations, made on 2005 March 31, April 5, April 12, and May 4 at 4.86 GHz,
189: showed only one significant radio source within the 2$\arcmin$ {\it INTEGRAL}
190: error circle, located at RA 17$^h$ 09$^m$ 45.934$^s$ $\pm$ 0.011s, Dec --36$^\circ$
191: 27$\arcmin$ 57.30$\arcsec~\pm$ 0.55$\arcsec$ (J2000) \citep{Atel490}.
192:
193: Thanks to the radio observations, a probable optical/infrared identification
194: has been found within the 2MASS All-Sky Catalog and the SuperCOSMOS Sky
195: Survey~\citep{Atel478,Atel479,Atel490,Atel494}. On the base of the optical
196: identification, \citet{Greb2007} estimated the source distance as ${\it d}$=
197: 10.5 kpc and an upper limit on the inclination angle: ${\it }\geq$
198: 77$^{\circ}$.
199:
200:
201: \section{The multi-wavelength observation campaigns} \label{obs}
202:
203: We report here on the analyzed data collected by {\it INTEGRAL}, {\it Swift}
204: and {\it XMM} from 2005 till 2007. To illustrate the need for a multi-wavelength
205: campaign with good positional accuracy, we show in Figure~\ref{2003-2005IBIS} two
206: {\it INTEGRAL} mosaic images of the region of the two BHCs. In the left panel,
207: we show the region during the IGR J17091--3624 outburst in 2003 (IGR
208: J17098--3628 was switched off), while in the right panel we show the same zone
209: when only IGR J17098--3628 was visible.
210:
211: \begin{figure}
212: % \centering \epsscale{0.9}\plotone{fig1.eps}
213: \plotone{fig1.eps}
214: \caption{Left: 2003 IBIS/{\it INTEGRAL} 80 ks mosaic image (20-40 keV) of IGR J17091--3624 and IGR J17098--3628 zone: only IGR J17091--3624 is visible. Right: 2005 IBIS/{\it INTEGRAL} 120 ks mosaic image (20-40 keV) of IGR J17091--3624 and IGR J17098--3628 zone: only IGR J17098--3628 is visible}
215: \label{2003-2005IBIS}
216: \end{figure}
217:
218: The two sources are too close to each other to be resolved by X-ray and
219: $\gamma$-ray monitors such us {\it RXTE}/ASM and {\it Swift}/BAT. Therefore, it
220: is not possible to measure their flux separately and it is rather difficult to
221: follow their individual activities. Observations with better positional
222: accuracy are needed to assign each flux variation to one of the two sources
223: with confidence. As an example, Figure~\ref{2007ASM} shows the contaminated
224: {\it RXTE}/ASM IGR J17098--3628 plus IGR J17091--3624 light curve between
225: 1.5--12 keV from 2001 until the beginning of 2008. The black and gray arrows
226: represent the periods of {\it BeppoSAX}, {\it Swift}/XRT and {\it XMM} observation campaigns that
227: verified the activity of respectively IGR J17091--3624 and IGR J17098--3628.
228:
229: \begin{figure}
230: % \centering \epsscale{0.8}\plotone{fig2.eps}
231: \plotone{fig2.eps}
232: \caption{IGR J17098--3628 {\it RXTE}/ASM light curve contaminated by IGR J17091--3624 outbursts from 2001 till January 2008. The black triangles and gray arrows represent the periods of {\it BeppoSAX}, {\it Swift}/XRT and {\it XMM} observation campaigns that verified the activity of respectively IGR J17091--3624 and IGR J17098--3628. The time is expressed in Modified Julian Date (52000 (MJD)= 2001-04-01; 53500 (MJD)= 2005-05-10; 54500 (MJD)= 2008-02-04).}
233: \label{2007ASM}
234: \end{figure}
235:
236:
237: The X-ray Telescope (XRT) on board {\it Swift} satellite is an imaging
238: instrument operating in the 0.2--10~keV energy range with a single photon point
239: spread function of 18$\arcsec$ (half-power diameter) and a spectral resolution
240: of 140~eV at 5.9~keV. We have collected all the XRT data available from the
241: NASA HEASARC public
242: archive\footnote{http://heasarc.gsfc.nasa.gov/docs/archive.html}. The XRT data
243: were processed using the most recently available standard {\it Swift} tools:
244: XRT software version 0.11.4, FTOOLS version 6.3.1, and XSPEC version 11.3.2.
245: The ancillary response files were generated with the {\it xrtpipeline} task
246: {\it xrtmkarf}. The channels of each spectrum were re-binned in order to achieve
247: a minimum of 100 counts per bin. The {\it Swift}/BAT transient monitor light
248: curve was provided by the {\it Swift}/BAT
249: team\footnote{http://swift.gsfc.nasa.gov/docs/swift/results/transients/index.html}.
250:
251: We also analyzed two {\it XMM} pointed observations performed on 2006 August
252: 25 and 2007 February 19. {\it XMM} data were reprocessed using the {\it
253: XMM}-Newton Standard Analysis Software (SAS) version 7.0, employing the latest
254: available calibration files. Only single X-ray events (PATTERN=0) were taken
255: into account for the PN;
256: % Only patterns corresponding to single, double, triple and quadruple X-ray events for the two MOS cameras were selected (PATTERN $\leq 12$), while for the PN only single events (PATTERN=0) were taken into account;
257: the standard selection filter FLAG=0 was applied. Exposures have been filtered
258: for periods of high background. Since pile-up was present in both observations,
259: source counts were extracted from annular regions of typically 90$\arcsec$
260: external radius centered on the source, and the central 7$\arcsec$ of the PSF
261: have been excised; background spectra were extracted from circular regions
262: close to the source or from source-free regions of typically 20$\arcsec$ of
263: radius. The ancillary response matrices (ARFs) and the detector response
264: matrices (RMs) were generated using the {\it XMM}-SAS tasks {\it arfgen} and
265: {\it rmfgen}; spectral channels were re-binned in order to achieve a minimum of
266: 100 counts per each bin.
267:
268: The analyzed {\it INTEGRAL} data set consists of all Key Program and public
269: observations from 2005 May until the end of 2007. The {\it INTEGRAL} data
270: reduction of both the X-ray monitor JEM-X~\citep{Lund} and the $\gamma$-ray
271: telescope IBIS~\citep{Ube} were performed using the latest release of the
272: standard Offline Scientific Analysis (OSA)~\citep{Gold} version 7.
273:
274: We have also searched in the radio archives for data containing IGR J17091--3624
275: and IGR J17098--3628 observations. Very Large Array (VLA)\footnote{The National
276: Radio Astronomy Observatory is operated by Associated Universities, Inc., under
277: cooperative agreement with the National Science Foundation.} observations were
278: taken for a few epochs in 2003, following the outburst of IGR J17091--3624, and
279: in 2005, after the IGR J17098--3628 detection. We have obtained and calibrated
280: the data in AIPS with the standard procedures.
281:
282:
283: \section{Results} \label{res}
284: \subsection{Swift 2005 Observations}
285:
286: The 2005 {\it Swift} observations were performed from May 1 until September 24,
287: after the peak of the IGR J17098--3628 outburst. A detailed analysis of this
288: outburst peak has been reported by \citet{Greb2007}. We report here the
289: spectral analysis of 6 pointings (about 2 ks each) that were performed in
290: Window Timing (WT). Figure~\ref{2005Swift} shows the image of one XRT pointing
291: observation performed in photon counting mode in order to detect the refined
292: position of the source; as the Figure shows, IGR J17098--3628 is quite bright
293: while IGR J17091--3624 is not visible.
294:
295: \begin{figure}
296: % \centering \epsscale{0.7}\plotone{fig3.eps}
297: \plotone{fig3.eps}
298: \caption{350 s {\it Swift}/XRT row image during 2005. The observation was pointed on IGR J17091--3624, which is not detected. IGR J17098--3628 is quite bright and affected by pileup.}
299: \label{2005Swift}
300: \end{figure}
301:
302: During all the {\it Swift} observations, IGR J17098--3628 showed a slightly
303: variable flux of about 3\% with a constant spectral shape. The best fit for
304: each individual observation is a multicolor disc black body component [{\it
305: diskbb} in XSPEC ~\cite{diskbb}] with an internal temperature of about 0.9 keV;
306: the equivalent hydrogen column [$N_H=(0.88 \pm 0.02) \times 10^{22}$~cm$^{-2}$] is
307: consistent with the galactic one. Table~\ref{2005_par} shows the spectral
308: parameters of the fit for each observation.
309:
310: The {\it INTEGRAL}/IBIS analysis of the data collected in the same period of
311: the {\it Swift} observation campaign did not reveal any source emission above
312: 15 keV up to 3 $\sigma$ level. The ASM monitor detected the source with an average
313: flux of about 30 mCrab, that is a value slightly above the detection limit of
314: JEM-X monitor (about 20-25 mCrab). Moreover, the source has been observed only
315: in the JEM-X partially coded field of view because of dithering strategy
316: used for {\it INTEGRAL} observations. For these reasons, the X-ray monitor
317: JEM-X did not detect the source in each single science window at the 3 $\sigma$
318: level. However, adding the JEM-X science windows together yields a total
319: exposure time of 450 ks which then provides a 36$\sigma$ level detection of
320: the source in the mosaic.
321:
322: %the source has an average flux with substantially the same spectrum found by the XRT data analysis (!!! da controllare !!!). %Figure (PIPPO) shows the XRT/ JEM-X simultaneous spectrum... %\\ %Aggiungere figura PIPPO\\ %\\
323:
324: \begin{table}
325: \begin{center}
326: \caption{XRT fit parameters of the 2005 IGR J17098--3628 campaign (errors are at 90\% confidence level): $T_{in}$= inner temperature of the accretion disk, $N_{disk}$= normalization constant of the {\it diskbb} model, Flux$_{0.5-10}$= unabsorbed model flux between 0.5-10 keV, ${\chi}^{2}_{red}$= reduced $\chi$ square}
327: \label{2005_par}
328: \begin{tabular}{lccccc}
329: \hline
330: \hline
331: pointing & Esposure & $T_{in}$ & $N_{disk}$ & Flux$_{0.5-10}$ & ${\chi}^{2}_{red}$ \\
332: Date & ks & keV & - &ergs cm$^{-2}$ s$^{-1}$ & -\\
333: 2005-05-01 &0.4 & 1.09$^{0.01}_{-0.01} $ & 76$^{4}_{-4}$ & 2.4$\times$10$^{-9}$ & 1.08 \\
334: 2005-05-10 &1.0 & 1.06$^{0.02}_{-0.02} $ & 80$^{6}_{-6}$ & 1.2$\times$10$^{-9}$ & 0.98 \\
335: 2005-05-21 &1.0 &0.98$^{0.02}_{-0.02}$ &87$^{8}_{-7}$ & 1.6$\times$10$^{-9}$ & 0.99\\
336: 2005-08-13&1.5 &0.82$^{0.02}_{-0.02}$&95$^{11}_{-10}$ & 9.6$\times$10$^{-10}$ & 1.28 \\
337: 2005-09-05& 2.1&0.93$^{0.01}_{-0.01}$&68$^{5}_{-5}$ &1.1$\times$10$^{-9}$& 1.23 \\
338: 2005-09-14&3.2 &0.93$^{0.01}_{-0.01}$&73$^{4}_{-4}$ &1.5$\times$10$^{-9}$& 1.20 \\
339: \hline
340: \end{tabular}
341: \end{center}
342: %\begin{list}{}{} \item[$^{\mathrm{3}}$]{Normalization constant in XSPEC units}
343: %\end{list}
344:
345: \end{table}
346:
347:
348:
349:
350: \subsection{XMM 2006-2007 Observations}
351:
352: {\it XMM} observations of the field containing both sources were performed in
353: 2006 August 29 and in 2007 February 19. These observations revealed that IGR
354: J17091--3624 still did not show any detectable emission, while IGR J17098--3628
355: was in a relatively bright state. {\it INTEGRAL} observed the field containing
356: the two sources at the same time as {\it XMM}, with no detections from either
357: sources; the {\it INTEGRAL} upper limit flux for a single pointing (2400 s) is
358: about 20 mCrab in the range 20--100 keV.
359:
360: \begin{figure}
361: % \centering \epsscale{0.5}\plotone{fig4.eps}
362: \plotone{fig4.eps}
363: \caption{{\it XMM}/EPN image (16 ks) of the 2007-02-19 pointing observation. IGR J17091-3624 is not visible while IGR J17098-3628 is bright and affected by pile up (The dark bar in the image is due to the gap between two adjoining CCDs).}
364: \label{2007XMM}
365: \end{figure}
366:
367: \subsubsection{IGR J17091--3624}
368:
369: No detectable emission was revealed from IGR J17091--3624. Since the distance
370: of this source is still unknown, we assumed IGR J17091--3624 as located at the
371: Galactic Center (8 kpc). The corresponding best upper limit luminosity, derived
372: for the quiescent state of this source and based on the {\it XMM} observations (exposure time 7 ks),
373: is $L \leq 7 \times 10^{32}$~ergs$^{-1}$ \citep{XMM_s}.
374:
375: \subsubsection{IGR J17098--3628}
376:
377: During August 2006 and February 2007, the {\it XMM} spectral shape of the
378: source did not show, once more, any important variation. The only noticeable
379: difference was the flux decrease of about 20\% between the two pointings. Both
380: spectra are characterized by an absorbed multicolor disc black body emission
381: with an inner temperature that varied from about 1 keV in the first pointing to
382: about 0.9 keV in the second pointing (see Table~\ref{XMMtable}).
383: The equivalent hydrogen column, $N_{H}=0.8 \times 10^{22}$~cm$^{-2}$ is in good agreement with
384: that measured by XRT in 2005 and with the galactic value. Table~\ref{XMMtable}
385: shows the fit parameters of both pointings. There is no evidence in the
386: spectral shape of a reflection component or Fe emission line according with the
387: 2005 XRT data set analysis. Figure~\ref{XMMspec} shows the unfolded spectrum
388: of the two {\it XMM}/PN pointings.
389:
390: \begin{figure}
391: \centering
392: \includegraphics[angle=-90, scale=0.3]{fig5a.eps}
393: \includegraphics[angle=-90, scale=0.3]{fig5b.eps}
394: %\plottwo{fig5.eps}{fig6.eps}
395:
396: \caption{{\it XMM}/PN unfolded spectra of IGR J17098--3628. left panel: 2006-08-25 spectrum. Right panel: 2007-02-19 spectrum. The data are fitted with {\it diskbb} model. Fit parameter values are reported in Table~\ref{XMMtable}}
397: \label{XMMspec}
398: \end{figure}
399:
400:
401: \begin{table}
402: \begin{center}
403: \caption{PN fit parameters of the two {\it XMM} pointings (errors are at 90\% confidence level): $T_{in}$= inner temperature of the accretion disk, $N_{disk}$= normalization constant of the {\it diskbb} model, Flux$_{0.5-10}$= unabsorbed model flux
404: between 0.5-10 keV, ${\chi}^{2}_{red}$= reduced $\chi$ square
405: }
406: \label{XMMtable}
407: \begin{tabular}{lccccc}
408: \hline
409: \hline
410: pointing & Exposure &$T_{in}$ & $N_{disk}$ & $F_{0.5-10}$ & ${\chi}^{2}_{red}$ \\
411: Date & ks & keV & \nodata &ergs~cm$^{-2}$ s$^{-1}$ & \nodata \\
412: 2006-08-25&8 & 1.04$^{+0.01}_{-0.01}$ & 47$^{+3}_{-3}$& 1.2$\times$10$^{-9}$ & 0.85 \\
413: 2007-02-19& 16 &0.87$^{+0.01}_{-0.01}$ &65$^{+2}_{-2}$ & 8$\times$10$^{-10}$ & 0.97 \\
414: \hline
415: \end{tabular}
416: \end{center}
417: %\begin{list}{}{} \item[$^{\mathrm{3}}$]{Normalization constant in XSPEC units}
418: %\end{list}
419: \end{table}
420:
421:
422: This behavior is confirmed by the {\it RXTE}/ASM light curve from 2005 May 1 to
423: 2007 April 10. In this period, IGR J17091--3624 was in quiescence and the ASM
424: light curve is only due to IGR J17098--3628 flux emission. As
425: Figure~\ref{2005Camp} shows, the {\it RXTE}/ASM hardness-intensity diagram
426: confirms that the source varied only the flux with a constant hardness ratio.
427:
428: \begin{figure}
429: % \centering \epsscale{0.6}\plotone{fig7.eps}
430: \plotone{fig6.eps}
431: \caption{{\it RXTE}/ASM hardness-intensity diagram of the period from 2005 May 1 until 2007 April 10. In this period IGR J17091--3624 was in quiescence and the ASM flux emission is due only to IGR J17098--3628. The hardness ratio is defined as: $HR=({flux_{B}/flux_{A}})$ where A and B are respectively the 1.5--3 keV and 3--5 keV energy bands. The plot clearly shows that the hardness value have not changed all over the observation period.}
432: \label{2005Camp}
433: \end{figure}
434:
435:
436:
437: \subsection{The July 2007 {\it Swift}/XRT monitoring: the new IGR J17091--3624 outburst}\label{2007sec}
438:
439: On 2007 July 9, as a consequence of an increase in the {\it Swift}/BAT and {\it
440: RXTE}/ASM light curves, a {\it Swift}/XRT 2.5 ks ToO was granted. The
441: observation was performed in window timing mode in order to avoid the pileup
442: and revealed the presence of two sources in the XRT field of view, both of
443: which were causing the flux increase in the RXTE/PCA light curve. Another 500 s
444: pointing observation, performed in photon counting mode in order to locate the
445: position of the two sources, revealed the presence of both IGR J17098--3628
446: and IGR J17091--3624, which was newly in outburst. Both sources were in
447: relatively bright state as Figure~\ref{2007XRT} shows.
448:
449: \begin{figure}
450: % \centering \epsscale{0.6}\plotone{fig8.eps}
451: \plotone{fig7.eps}
452: \caption{{\it Swift}/XRT 0.1-10 keV image of the July 2007 ToO {\it Swift} 500s observation. The observation has been performed in photon counting mode in order to locate the position of the sources. IGR J17098--3628 is affected by pile up.}%The rings in the image are due to the presence of a strong source (GX349+2) at the limit of the field of view}
453: \label{2007XRT}
454: \end{figure}
455:
456: A {\it Swift} monitoring campaign was then requested for a total of 3 more
457: observations, 2 ks each (one per week between 2007 July 24 and 2007 August 7)
458: pointed on IGR J17091--3624. The key outcomes of this observation campaign for
459: each source are summarized in the following sections.
460:
461: %The XRT observations were made in window timing mode except a short one (500s) performed in photon counting mode in order to better locate the position of the two sources(Figure~\ref{2007XRT}).
462:
463: \subsubsection{IGR J17098--3628}
464:
465: IGR J17098--3628 has been observed during only three of the pointings associated with the observational campaign, for a total exposure of $\sim$ 8 ks. The source flux only showed flux variations of about 30\% in the range 0.5-10 keV as was inferred from the light curve shown in Figure~\ref{lc0982007}.
466: The source spectrum revealed only minor changes and the best overall fit is represented by a {\it diskbb} component with an inner temperature of about 0.9-1 keV.
467:
468: \begin{figure}
469: \centering
470: \includegraphics[angle=-90, scale=0.4]{fig8.eps}
471: \caption{{\it Swift}/XRT 0.1-10 keV light curve of IGR J17098--3628 from 2007-07-09 to 2007-07-17 (bin time: 1000 s). The gap in the light curve is due to the time interval between the second and the third XRT pointing}
472: \label{lc0982007}
473: \end{figure}
474:
475:
476:
477: \subsubsection{IGR J17091--3624}
478:
479: This XRT ToO was a unique opportunity to catch the new outburst of IGR
480: J17091--3624 from the very beginning and also provided the refined position of
481: the source \citep{KenCap}: RA 17$^h$ 09$^m$ 07.6$^s$, Dec --36$^\circ$ 24$\arcmin$ 24.9$\arcsec$
482: (J2000), with an estimated uncertainty radius of 3.6$\arcsec$ (90\%
483: confidence), consistent with the {\it INTEGRAL} position. However, this X-ray
484: position rules out the tentative radio counterpart previously proposed for IGR
485: J17091--3624 \citep{Atel490,Pand}, which lies outside the XRT error circle,
486: approximately 86$\arcsec$ away. This location also rules out the optical
487: counterpart suggested by Negueruela \& Schurch (2006) which was also based on
488: the radio counterpart. During the {\it Swift} observations the 0.5-10 keV flux
489: of IGR J17091--3624 increased from $5 \times 10^{-10}$ (ergs~cm$^{-2}$
490: s$^{-1}$) to $2 \times 10^{-9}$ (ergs~cm$^{-2}$
491: s$^{-1}$). Figure~\ref{lcurve_091} shows the 0.1--10 keV light curve of the
492: 2007 XRT monitoring campaign.
493:
494: \begin{figure}
495: \centering
496: \includegraphics[angle=-90, scale=0.4]{fig9.eps}
497: \caption{{\it Swift}/XRT (0.1-10 keV) light curve of IGR J17091--3624 from 2007-07-09 to 2007-08-07 (bin time: 1000 s). }
498: \label{lcurve_091}
499: \end{figure}
500:
501: In Figure ~\ref{2007ASM_BAT}, we show respectively the {\it Swift}/BAT 15--50
502: keV (top panel) and {\it RXTE}/ASM 2--10 keV (bottom panel) light curves at the
503: time of the 2007 outburst. The luminosity peak present in both light curves is
504: considered to be mainly due to the IGR J17091--3634 outburst, since the
505: contamination from the IGR J17098--3628 flux should not imply such a strong
506: variation of the flux. In fact, as discussed before, IGR J17098--3628 showed a
507: soft spectrum and a flux variation clearly lower than the one reported by the
508: two light curves of Figure~\ref{2007ASM_BAT}.
509:
510: However, the BAT and ASM light curve show a behavior typical of that expected
511: at the beginning of a black hole binary outburst, with the hard luminosity peak
512: preceding the soft one. This behavior is also confirmed by the spectral
513: analysis: a power law component with a photon index of $1.4 \pm 0.1$ represents
514: the best fit of the first two observations (9 July and 16 July 2007).
515: For the spectra of the subsequent observations, it is necessary to add a
516: multicolor disc black body component to the power law. Table~\ref{2007_results}
517: summarizes the spectral parameter values of the XRT 2007 observation campaign.
518:
519: \begin{figure}
520: % \centering \epsscale{0.8}\plotone{fig11.eps}
521: \plotone{fig10.eps}
522: \caption{Top panel: IGR J17091--3624 {\it Swift}/BAT (15--50 keV) light curve. Bottom panel: IGR J17091--3624 {\it RXTE}/ASM (2--10 keV). Both curves are contaminated by the IGR J17098--3628 flux emission. The time is expressed in Modified Julian Date (54000 (MJD)= 2006-09-22;54200 (MJD)= 2007-04-10; 54400 (MJD)= 2007-10-27)}
523: \label{2007ASM_BAT}
524: \end{figure}
525:
526:
527: \begin{table}
528: \begin{center}
529: \caption{The fit parameters of the observed spectra of IGR J17091--3624 during 2007 {\it Swift}/XRT monitor campaign (errors are at 90\% confidence level): $T_{in}$= disk inner temperature, $N_{disk}$= {\it diskbb} model normalization constant, $\Gamma$= power law photon index, $N_{pow}$= power law model normalization costant, $Flux_{0.5-10}$= unabsorbed model flux between 0.5-10 keV, ${\chi}^{2}_{red}$= reduced $\chi$ square.}
530: \label{2007_results}
531: \begin{tabular}{lcccccccc}
532: \hline
533: \hline
534: pointing & Exposure & $T_{in}$ & $N_{disk}$ & $\Gamma$ & $N_{pow}$ & $Flux_{0.5-10}$ & ${\chi}^{2}_{red}$. \\
535: Date & ks & keV & \nodata & \nodata &pho~keV$^{-1}$~cm$^{-2}$~s$^{-1}$ &ergs~cm$^{-2}$~s$^{-1}$ & \nodata \\
536: 2007-07-09& 2.4 & \nodata & \nodata & 1.4 $^{0.1}_{-0.1}$& 0.05$^{0.01}_{-0.01}$ & 4.7$\times$10$^{-10}$ & 0.96 \\
537: 2007-07-16& 3.4 & \nodata & \nodata &1.3$^{0.1}_{-0.1}$&0.028$^{0.003}_{-0.003}$ & 3.0$\times$10$^{-10}$ & 1.10 \\
538: 2007-07-24& 2.6 &0.86$^{0.03}_{-0.03}$ &69$^{13}_{-13}$ &1.6$^{0.1}_{-0.1}$&0.12$^{0.02}_{-0.02}$ & 1.8$\times$10$^{-9}$ & 1.20 \\
539: 2007-08-07& 1.9 & 0.75$^{0.05}_{-0.07}$ &57$^{30}_{-23}$ &2.0$^{0.1}_{-0.1}$&0.23$^{0.04}_{-0.03}$ & 2.1$\times$10$^{-9}$ & 1.08\\
540: \hline
541: \end{tabular}
542: \end{center}
543:
544: %\end{list}
545: \end{table}
546:
547: {\it INTEGRAL} detected this source only at the end of the outburst because of visibility constrains. Unfortunately the source was outside the field of view of the {\it INTEGRAL} X-ray monitor JEM-X, thus only the IBIS high energy data were available.
548: Figure~\ref{091_2007} shows the 20-100 keV IBIS image of the field of view which includes the two sources between August and September 2007 (for a total exposure time of $\sim$50 ks). As Figure~\ref{091_2007} shows IBIS detected above 20 keV only IGR J17091--3624.
549: \begin{figure}
550: % \centering \epsscale{0.5}\plotone{fig12.eps}
551: \plotone{fig11.eps}
552: \caption{IGR J17091--3624 {\it INTEGRAL}/IBIS 20-100 keV image between August and September 2007 (total exposure time of $\sim$ 50 ks).}
553: \label{091_2007}
554: \end{figure}
555:
556:
557: The first {\it INTEGRAL} IGR J17091--3624 observation of the two sources field (2007-08-25) was performed 18 days after the last XRT observation (2007-08-07). However, as Figure~\ref{2007ibis_xrt} shows, XRT and IBIS spectra are in good agreement. The {\it INTEGRAL} observations, after the end of the XRT campaign, revealed that the spectrum of the source continued its softening with a photon index that varied from 2 to $\simeq$ 3 at the end of September 2007.
558:
559: \begin{figure}
560: \centering
561: \includegraphics[angle=-90, scale=0.4]{fig12.eps}
562: \caption{IGR J17091--3624 2007 July outburst: XRT-IBIS combined spectrum: the XRT observation (2 ks) had been performed 18 days before that of IBIS (14 ks). The best fit model is represented by a multicolor disc black body (red dashed line) plus a power law component (green dashed line). The total model is represented in the figure by the solid green line. The spectral parameters are: $T_{in}=0.75 \pm 0.1 $keV, $N_{disk}=57 \pm 25$, $\Gamma =2 \pm 0.1$. The flux between 0.5-100 keV is : 9$\times$10$^{-9}$ ergs cm$^{-2}$ s$^{-1}$}
563: \label{2007ibis_xrt}
564: \end{figure}
565:
566:
567:
568:
569: \subsection{Radio Counterparts}
570:
571: A plot of the NVSS data in the region of IGR J17091--3624 and IGR J17098--3628 is
572: shown in Figure~\ref{f.nvss} together with the position and associated
573: uncertainties of the {\it INTEGRAL} and {\it Swift}/XRT measurements (marked by
574: vertical and diagonal crosses, respectively) and the tentative radio
575: counterparts from VLA literature data \citep{Atel152,Atel490}. In the region
576: of IGR J17098--3628, there is clearly an excellent agreement between the radio
577: and X-ray positions, which gives a high confidence identification. The NVSS
578: contours themselves show some emission in this region, although it is quite
579: extended and only significant at the 2.5 $\sigma$ level. In the region of IGR
580: 17091--3624, the refined {\it Swift}/XRT position is clearly not consistent
581: with the proposed radio counterpart. No signal is detected in the NVSS image in
582: that region.
583:
584:
585:
586: \begin{figure}
587: \plotone{fig13.eps}
588: \caption{NVSS image of the region of IGR J17091--3624 and IGR
589: 17098--3628, contours traced at ($-1, 1, 2, 4, ...) \times 1$ mJy
590: beam$^{-1}$ (thin grey lines). Darker bold symbols represent IBIS
591: positions (big plus signs), {\it Swift}/XRT positions (circles, 90\%
592: confidence), and previously proposed radio counterparts (crosses). The
593: key in the bottom right corner shows the radio beam size ($45\arcsec
594: \times 45\arcsec$ HPBM).}
595: \label{f.nvss}
596: \end{figure}
597:
598: We have therefore searched the VLA archive for higher resolution images of
599: these regions, primarily to find a radio counterpart to IGR J17091--3624 but
600: also in order to study the time and the spectral evolution of the sources. We
601: have found archival data from 2003 in the region of IGR J17091--3624 and from
602: 2005 in the region of IGR J17098--3628. The data from 2003 were taken in the
603: compact D configuration on Apr 23, Apr 26, May 06, and May 09, with exposures
604: of a few minutes. We show an 8.4 GHz image of the IGR J17091--3624 region in
605: Figure~\ref{f.091}, superimposed to the {\it Swift}/XRT position. A radio source
606: is clearly revealed in a position consistent with the high energy error
607: circle. This component is present in the three epochs at 8.4 GHz, and shows an
608: increase in total flux. Data at 5 GHz are also available at three epochs; the
609: source is marginally detected in the first one (3$\sigma$, 5 GHz data only),
610: and clearly revealed in the next two. For the epochs where simultaneous data
611: are available, a comparison to the 8.4 GHz flux reveals an inverted
612: spectrum. Table~\ref{t.091} reports on the measured flux densities; the
613: position of the radio source is RA 17$^h$ 09$^m$ 07.5$^s$ $\pm$ 0.5$^s$, Dec
614: --36$^\circ$ 24$\arcmin$ 24$\arcsec$ $\pm$ 3$\arcsec$.
615:
616:
617: \begin{figure}
618: \plotone{fig14.eps}
619: \caption{Very Large Array 8.4 GHz image of the region of IGR
620: 17091--3624, taken on 2003 May 09. Contours are traced at $(-0.3, 0.3,
621: 0.6, 1.2)$ mJy beam$^{-1}$; the gray scale range is 0.2 to 1.1 mJy
622: beam$^{-1}$. The circle shows the refined {\it Swift} position (90\%
623: confidence), and the cross shows the previously proposed radio
624: counterpart. The key in the bottom left corner shows the radio beam size
625: ($40.4\arcsec \times 7.3\arcsec$ HPBM in PA $30^\circ$, measured north
626: to east).}
627: \label{f.091}
628: \end{figure}
629: % \begin{deluxetable}{lccc}[hb!]
630: % \tablecaption{Flux density of the radio counterpart of IGR 17091--3624 \label{t.091}}
631: % \tablehead{
632: % \colhead{Date} & \colhead{$S_{\mathrm{5\, GHz}}$} & \colhead{$S_{\mathrm{8.4\, GHz}}$} &
633: % \colhead{Spectral index} \\
634: % \colhead{} & \colhead{(mJy)} & \colhead{(mJy)} & \colhead{}
635: % }
636: % \startdata
637: % 2003 Apr 23 & 0.32 & \nodata & \nodata \\
638: % 2003 Apr 26 & 0.48 & 0.56 & $-0.3$ \\
639: % 2003 May 06 & 0.42 & 0.67 & $-0.8$ \\
640: % 2003 May 09 & \nodata & 0.73 & \nodata \\
641: % \enddata
642: % \tablecomments{The observation of 2003 Apr 23 was at 5 GHz only and is only a $3 \sigma$ detection; 2003 May 09 is only at 8.4 GHz. Spectral index is defined according to $S(\nu) \propto \nu^{-\alpha}$.}
643: % \end{deluxetable}
644:
645: \begin{table}
646: \begin{center}
647: \caption{Flux density of the radio counterpart of IGR J17091--3624. The observation of 2003 Apr 23 was at 5 GHz only and is a 3 $\sigma$ detection; 2003 May 09 is only at 8.4 GHz. Spectral index is defined according to $S(\nu) \propto \nu^{-\alpha}$.}
648: \label{t.091}
649: \begin{tabular}{lcccc}
650: \hline
651: \hline
652: Date & $S_{\mathrm{5\, GHz}}$ & $S_{\mathrm{8.4\, GHz}}$ & Spectral index \\
653: - & (mJy) & (mJy) & -\\
654: 2003 Apr 23 & 0.32 & \nodata & \nodata \\
655: 2003 Apr 26 & 0.48 & 0.56 & $-0.3$ \\
656: 2003 May 06 & 0.42 & 0.67 & $-0.8$ \\
657: 2003 May 09 & \nodata & 0.73 & \nodata \\
658:
659: \hline
660: \end{tabular}
661: \end{center}
662:
663: %\end{list}
664: \end{table}
665:
666: \section{Discussion and Conclusions} \label{disc}
667: The huge amount of data collected and the broad energy coverage allowed us to perform a detailed spectral analysis of these two sources, following their evolution during more than two years and fixing upper limits when one of the sources was not visible. In the following a detailed summary of the results for both sources can be found.
668:
669: \subsection{IGR J17098--3628}
670: Our analysis, spanning from May 2005 to September 2007, proves that the source
671: spectrum shows a soft black body component with an internal temperature of
672: about $\sim$ 1 keV and an internal radius comparable to the last stable orbit
673: of the accretion disc. The luminosity due to soft component \citep[assuming
674: $d=10.5$ kpc,][]{Greb2007} varied during the two years of our observational
675: campaign from $\sim 2 \times 10^{37}$ to $\sim 5 \times 10^{36}$
676: ergs~s$^{-1}$. The flux variation range and the spectral parameters are
677: comparable with the ones reported by \citet{Greb2007} during the first phases
678: of the outburst on April 2005. In our data set each flux variation seems not
679: to be correlated with any relevant spectral features, as confirmed by the lack
680: of evident variation of the {\it RXTE}/ASM hardness-intensity diagram
681: (Figure~\ref{2005Camp}). However respect to the previous April 2005
682: observations, we did not detect any power law emission at higher {\it
683: INTEGRAL}/IBIS energy band. This indicates that the spectral shape was evolved
684: from the first phase of the outburst being dominated only by a thermal disc
685: component coming from the accretion disc around the compact object.
686:
687: We can conclude that this source seems to have spent 2.5 years in the high soft state with a disc black body component substantially identical to the one previously observed at the beginning of the outburst~\citep{Greb2007}. Curiously the only difference is the lack of any high energy emission. In fact the hard component fell below the detection limit of IBIS after about three months from the beginning of the outburst. Hence the geometry and the temperature of the accretion disc have not shown any significant variation up to now, on the other end the power law emission quenched probably because of the electron temperature of the corona fallen below the disc seed photons temperature making the inverse Compton scattering processes inefficient.
688:
689:
690: \subsection{IGR J17091--3624}
691:
692: Previous studies have demonstrated that the X-ray luminosity of the BHCs in quiescence is lower than that of neutron star X-ray binaries and falls below $\sim$ 10$^{32}$-10$^{33}$ ergs~s$^{-1}$ \citep{Campana}. In the quiescent state, neutron star X-ray binaries are normally detected and hence the {\it XMM} upper limit of IGR J17091--3624 is a good indication that the source is a BHC.
693:
694: The {\it Swift} ToO caught the source at the beginning of its outburst when it was still in a low/hard state. The best fit of the first observation (2007-07-09/16) is a power law spectrum with $\Gamma$=1.4$\pm$ 0.1. The source spectrum substantially softened and the black body component became dominant with an increasing temperature and a steeper power law.
695: {\it INTEGRAL} continued to monitor the source with IBIS after the end of the XRT observational campaign, up to the end of September 2007. These observations confirm the softening of the source. A power law component without an energy cutoff provided the best fit of the IBIS spectra between 20 and 80 keV, probably due to jets or corona reprocessing of disc seed photons. The power law photon index varied during two month period from about 1.4 (beginning of July) to about 3 (end of September). Unfortunately the relatively low source flux at high energies (7 $\times$ 10$^{-10}$ ergs~cm$^{-2}$ s$^{-1}$ between 20-100 keV) and the variation of the spectral shape did not provide sufficient statistics to extend the spectra up to 80 keV and to verify the presence or not of an high energy cutoff in the spectrum.
696:
697: Thanks to the XRT refined position, we found the radio counterpart
698: of the source. As soon as 9 days after the detection by IBIS in 2003, a radio
699: source at the sub-mJy level was detected at 5 GHz. The source has been detected
700: also at 8.4 GHz, and it showed an increase in flux over the subsequent two
701: weeks. The spectrum is inverted, characteristic of self-absorbed synchrotron
702: radiation from a compact jet. This behavior is typical of BHC in low hard
703: states. Although it is difficult to guess the spectral shape at higher
704: frequencies, it is reasonable to estimate that the total radiative luminosity of
705: the compact jet is of the order of 10$^{31}$ erg sec$^{-1}$, assuming the distance to
706: the Galactic Center. No information is available after the 2007 outburst, and it
707: will be clearly valuable to obtain new radio observations after future episodes
708: of activity.
709:
710: Chaty et al. 2008, on the basis of ESO NTT observations, report on two possible infrared counterparts of IGR J17091--3624 consistent with the {\it Swift}/XRT error box. Our refined position, based on the radio observations, could support the identification of the real infrared counterpart of this source.
711:
712: Little is known about nature, duration and recurrence of outbursts in transient X-ray binaries. It is probable that the outbursts are due to randomly acting factors such as the mass transfer variations or truncation of the inner disc radius. However for IGR J17091--3624 five outbursts are known from 1994 to 2007. Thus, it appears that this source goes in outburst every three or five years. Generally the flux varies from 5 mCrab to about 20 mCrab in the range 2-10 keV over a period of few months. The 2003 outburst was the first detected at higher energy range (20-150 keV) for one year period. Comparing the two hard states of respectively the 2003 and 2007 outbursts, the last one seems to be the hardest. In fact it was possible to determine the high energy power law cutoff of the 2003 hard state \citep[$\sim$ 49 keV,][]{Cap2006} while in the 2007 hard state the cutoff is not detectable up to 80 keV. Concerning the soft state the 2003 outburst has an higher temperature black body emission~\citep{Cap2006} that does not appear to be present in the 2007 outburst. However the XRT monitoring campaign observed the source only at the beginning of its transition to the soft state and unfortunately in the subsequent {\it INTEGRAL} monitoring the source was outside the JEM-X field of view and so it was not possible to follow the entire evolution of the black body temperature.
713: %Looking at the entire evolution of the 2003 and 2007 outburst we can conclude that either in hard and in soft state the last outburst seems to be fainter than the 2003 one but with and harder spectrum
714:
715: %\section{Conclusions}
716: \begin{acknowledgements}
717: This work has been supported by the Italian Space Agency through grants I/008/07/0 and I/088/06/0.
718: We acknowledge the use of public data from the {\it Swift} data archive and all the {\it Swift} team for its support and Memmo Federici for supervising the {\it {\it INTEGRAL}\/} data analysis.
719: Particular thanks are due to Erick Kuulkeers for making his private data available for analysis.
720: \end{acknowledgements}
721:
722: \begin{thebibliography}{}
723: \bibitem[Bird et al.(2007)]{Bird}
724: Bird, A. J., Malizia, A., Bazzano, A. et al., 2007, \apjs, 170, 175,
725: \bibitem[Blustin et al.(2005)]{Atel479}
726: Blustin, A. J., Kong, A. K. H., Wu, K. et al., Atel 479, 2005
727: \bibitem[Campana et al.(2001)]{Campana}
728: Campana, S., A. N., Parmar and Stella, L. A\&A, 372, 241, 2001
729: \bibitem[Capitanio et al.(2006)]{Cap2006}
730: Capitanio, F., Bazzano, A., Ubertini, P. et al. ApJ, 643, 376, 2006
731: \bibitem[Chaty et al.(2008)]{Chaty}
732: Chaty, S., Rahoui, F., Foellmi, F. et al. A\&A in press 2008 (astroph: 0802.1774)
733: \bibitem[Goldwrum et al.(2003)]{Gold}
734: Goldwrum, A., Daid, P., Foschini, L. et al. A\&A, 411, l223, 2003
735: \bibitem[Grebenev et al.(2007)]{Greb2007}
736: Grebnev, S., A., Molkov, S., V., Revnivtsev, M., G. \& Sunyaev, 2007, ESA SP, 622, 373 (astro-ph: 0709.2313)
737: \bibitem[Grebenev et al.(2005)]{Atel444}
738: Grebenev, S. A., Molkov, S. V., Sunyaev, R. A. 2005, Atel 444
739: \bibitem[Grebenev et al.(2005b)]{Atel447}
740: Grebenev, S. A., Molkov, S. V., Revnivtsev, M. G. \& Sunyaev, R. A., 2005 Atel 447
741: \bibitem[Kennea \& Capitanio(2007)]{KenCap}
742: Kennea, J. A. \& Capitanio, F. 2007, Atel 1140
743: \bibitem[Kennea et al.(2005)]{Atel476}
744: Kennea, J. A., Burrows, D. N., Nousek, J. A. et al., 2005, Atel 476.
745: \bibitem[Kuulkers et al.(2003)]{Kuulk}
746: Kuulkers, E., Lutovinov, A., Parmar, a., Capitanio, F. et al., 2003 Atel 149
747: \bibitem[Kong(2005)]{Atel477}
748: Kong, A. K. H. 2005, Atel 477
749: \bibitem[Lund et al.(2003)]{Lund}
750: Lund, N., Budz-Joergensen, C., Westgaard, N. J., 2003 A\&A, 411, L231
751: \bibitem[Lutovinov \& Revnivtsev(2003)]{Luto03}
752: Lutovinov, A. \& Revnivtsev, M. 2003, Astr. Lett., 29, 719
753: \bibitem[Lutovinov \& Revnivtsev(2005)]{Luto05}
754: Lutovinov, A., Revnivtsev, M., Molkov, S. and Sunyaev, R. 2005, A.\&A., 430, 997
755: \bibitem[Mitsuda et al.(1984)]{diskbb}
756: Mitsuda, K., Inoue, H., Koyama et al., 1984, PASJ, 36, 741
757: \bibitem[Mowlavi et al.(2005)]{Mol2005}
758: Mowlavi, N., Kuulkers, E., Rodriguez et al., 2005, Atel 453
759: \bibitem[Negueruela \& Schurch(2006)]{Nerg}
760: Negueruela, I. \& Schurch, M. P. E., 2006, A\&A, 461, 631
761: \bibitem[in 't Zand et al.(2003)]{Atel4}
762: in 't Zand, J. J. M., Heise, J., Lowes, P. \& Ubertini, P. 2003, ATel, 160
763: \bibitem[Pandey et al.(2006)]{Pand}
764: Pandey, M., Manchanda, R. K., Rao, A. P., Durouchoux, P. \& Ishwara-Chandra 2006
765: \bibitem[Revnivtsev et al.(2003)]{Atel2}
766: Revnivtsev, M., Gilfanov, M., Churazov, E., \& Sunyaev R. 2003, ATel, 150
767: \bibitem[Rupen et al.(2003)]{Atel152}
768: Rupen, M. P., Mioduszewski, A. J., \& Dhawan, V. 2003, ATel, 152
769: \bibitem[Rupen et al.(2005)]{Atel490}
770: Rupen, M. P., Mioduszewski, A. J., Dhawan, V., 2005, Atel 490
771: \bibitem[Steeghs et al.(2005)]{Atel494}
772: Steeghs, D., Torres, M. A. P., Jonker, P. G. et al., 2005, Atel 494
773: \bibitem[Steeghs et al.(2005b)]{Atel478}
774: Steeghs, D., Torres, M. A. P., Jonker et al., 2005, Atel 478
775: \bibitem[Ubertini et al.(2003)]{Ube}
776: Ubertini, P., Lebrun, F., Di Cocco, G. et al., A\&A, 411, L131, 2003
777: \bibitem[Watson et al. (2001)]{XMM_s}
778: Watson, M. G., Auguères, J.-L., Ballet, J. et al., A\&A 364, L51, 2001
779: % \bibitem[]{}
780: % \bibitem[]{}
781: % \bibitem[]{}
782: \end{thebibliography}
783: \end{document}