1: %%
2: %% Beginning of file 'sample.tex'
3: %%
4: %% Modified 2005 December 5
5: %%
6: %% This is a sample manuscript marked up using the
7: %% AASTeX v5.x LaTeX 2e macros.
8:
9: %% The first piece of markup in an AASTeX v5.x document
10: %% is the \documentclass command. LaTeX will ignore
11: %% any data that comes before this command.
12:
13: %% The command below calls the preprint style
14: %% which will produce a one-column, single-spaced document.
15: %% Examples of commands for other substyles follow. Use
16: %% whichever is most appropriate for your purposes.
17: %%
18: \documentclass[12pt,preprint]{aastex}
19:
20: %% manuscript produces a one-column, double-spaced document:
21:
22: %%\documentclass[manuscript]{aastex}
23:
24: %% preprint2 produces a double-column, single-spaced document:
25:
26: %% \documentclass[preprint2]{aastex}
27:
28: %% Sometimes a paper's abstract is too long to fit on the
29: %% title page in preprint2 mode. When that is the case,
30: %% use the longabstract style option.
31:
32: %% \documentclass[preprint2,longabstract]{aastex}
33:
34: %% If you want to create your own macros, you can do so
35: %% using \newcommand. Your macros should appear before
36: %% the \begin{document} command.
37: %%
38: %% If you are submitting to a journal that translates manuscripts
39: %% into SGML, you need to follow certain guidelines when preparing
40: %% your macros. See the AASTeX v5.x Author Guide
41: %% for information.
42:
43: \newcommand{\vdag}{(v)^\dagger}
44: \newcommand{\myemail}{skywalker@galaxy.far.far.away}
45:
46: %% You can insert a short comment on the title page using the command below.
47:
48: %%\slugcomment{Not to appear in Nonlearned J., 45.}
49:
50: %% If you wish, you may supply running head information, although
51: %% this information may be modified by the editorial offices.
52: %% The left head contains a list of authors,
53: %% usually a maximum of three (otherwise use et al.). The right
54: %% head is a modified title of up to roughly 44 characters.
55: %% Running heads will not print in the manuscript style.
56:
57: \shorttitle{IR observations of 1E 161348-5055}
58: \shortauthors{A. De Luca et al.}
59:
60: %% This is the end of the preamble. Indicate the beginning of the
61: %% paper itself with \begin{document}.
62:
63: \newcommand{\chan}{{\sl Chandra}}
64: \newcommand{\rosat}{{\sl ROSAT}}
65: \newcommand{\asca}{{\sl ASCA}}
66: \newcommand{\sax}{{\sl BeppoSax}}
67: \newcommand{\cxo}{{\sl CXO}}
68: \newcommand{\xmm}{{\sl XMM-Newton}}
69: \newcommand{\spitzer}{{\sl Spitzer}}
70:
71: \newcommand{\nh}{N_{\rm H}}
72:
73: \newcommand{\hst}{{\sl HST}}
74: \newcommand{\vlt}{{\sl VLT}}
75: \newcommand{\vltn}{{\sl Very Large Telescope}}
76: \newcommand{\hstn}{{\sl Hubble Space Telescope}}
77: \newcommand{\ctio}{{\sl CTIO}}
78: \newcommand{\aao}{{\sl AAO}}
79:
80: \newcommand{\wfi}{{\sl WFI}}
81: \newcommand{\wfin}{{\sl Wide Field Imager}}
82: \newcommand{\fors}{{\sl FORS1}}
83: \newcommand{\naco}{{\sl NACO}}
84: \newcommand{\isaac}{{\sl ISAAC}}
85: \newcommand{\nacon}{{\sl NAos COnica}}
86: \newcommand{\nicmos}{{\sl NICMOS}}
87: \newcommand{\ein}{{\sl Einstein}}
88:
89: \newcommand{\tmass}{{\sl 2MASS}}
90: \newcommand{\gsc}{{\sl GSC-2}}
91: \newcommand{\sex}{{\sl SExtractor}}
92: \newcommand{\dao}{{\em Daophot}}
93: \newcommand{\roma}{{\em Romafot}}
94: \newcommand{\all}{{\em Allframe}}
95:
96:
97: \begin{document}
98:
99: %% LaTeX will automatically break titles if they run longer than
100: %% one line. However, you may use \\ to force a line break if
101: %% you desire.
102:
103: \title{Deep infrared observations of
104: the puzzling central X-ray source in RCW103\altaffilmark{2,3}}
105:
106: %% Use \author, \affil, and the \and command to format
107: %% author and affiliation information.
108: %% Note that \email has replaced the old \authoremail command
109: %% from AASTeX v4.0. You can use \email to mark an email address
110: %% anywhere in the paper, not just in the front matter.
111: %% As in the title, use \\ to force line breaks.
112:
113: \author{A. De Luca\altaffilmark{1}}
114: %\altaffilmark{1,2,3} and Ivan R. King\altaffilmark{1}}
115: \affil{INAF - Istituto di Astrofisica Spaziale e Fisica Cosmica,
116: Via Bassini 15, I-20133 Milano, Italy}
117: \email{deluca@iasf-milano.inaf.it}
118: \and
119: \author{R.P. Mignani}
120: \affil{University College London, Mullard Space Science Laboratory, Holmbury St. Mary, Dorking, Surrey, RH5 6NT Unite
121: d Kingdom}
122: %\email{aastex-help@aas.org}
123: \and
124: \author{S. Zaggia}
125: \affil{INAF - Osservatorio Astronomico di Padova, Vicolo dell'Osservatorio 5,
126: I-35122, Padova, Italy}
127: \and
128: \author{G. Beccari}
129: \affil{INAF - Osservatorio Astronomico di Bologna, Via Ranzani 1, I-40127,
130: Bologna, Italy }
131: \and
132: \author{S.Mereghetti, P.A. Caraveo\altaffilmark{2}}
133: \affil{INAF - Istituto di Astrofisica Spaziale e Fisica Cosmica,
134: Via Bassini 15, I-20133 Milano, Italy}
135: \and
136: \author{G.F. Bignami\altaffilmark{1}}
137: \affil{Agenzia Spaziale Italiana, Via Liegi 26, I-00198 Roma, Italy}
138: %% Notice that each of these authors has alternate affiliations, which
139: %% are identified by the \altaffilmark after each name. Specify alternate
140: %% affiliation information with \altaffiltext, with one command per each
141: %% affiliation.
142: \altaffiltext{1}{Istituto Universitario di Studi Superiori di Pavia, V.le
143: Lungo Ticino Sforza 56, 27100
144: Pavia, Italy}
145: %%\altaffiltext{2}{Universit\'a degli Studi di Pavia,
146:
147: \altaffiltext{2}{Based on observations
148: collected at the European Southern Observatory, Paranal, Chile under
149: programme ID 67.D-0198(A),
150: 077.D-0764(A)}
151: \altaffiltext{3}{Based on observations with the NASA/ESA
152: Hubble Space Telescope, obtained at the Space Telescope Science
153: Institute, which is operated by AURA, Inc. under contract No.
154: NAS 5-26555.}
155:
156:
157: %% Notice that each of these authors has alternate affiliations, which
158: %% are identified by the \altaffilmark after each name. Specify alternate
159: %% affiliation information with \altaffiltext, with one command per each
160: %% affiliation.
161:
162:
163: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
164: %% style, abstract will output a Received/Accepted line after the
165: %% title and affiliation information. No date will appear since the author
166: %% does not have this information. The dates will be filled in by the
167: %% editorial office after submission.
168:
169: \begin{abstract}
170: 1E 161348-5055 (1E 1613) is a point-like, soft X-ray source
171: originally identified as a radio-quiet, isolated neutron star,
172: shining at the center of
173: the 2000 yr old supernova remnant RCW103.
174: 1E 1613 features a puzzling 6.67 hour periodicity as well as a
175: dramatic variability
176: %on the few year time scale.
177: over a time scale of few years.
178: %in flux,
179: % (by a factor $\sim100$),
180: %spectral shape, pulse shape and pulsed fraction.
181: Such a temporal behavior,
182: coupled to the young age and to the lack of an obvious optical counterpart,
183: makes 1E 1613 a unique source among all compact objects associated to SNRs.
184: It could either be the first low-mass X-ray binary system discovered
185: %immediately after its formation,
186: inside a SNR,
187: or a peculiar isolated magnetar with an extremely slow
188: spin period.
189: Analysis of archival IR observations, performed in 2001 with the VLT/ISAAC
190: instrument, and in 2002 with the NICMOS camera onboard HST unveils
191: a very crowded field. A few sources are positionally consistent with the
192: refined X-ray error region that we derived from the analysis of 13 Chandra observations.
193: To shed light on the nature of 1E 1613, we have performed
194: deep IR observations of the field with the NACO instrument at the
195: ESO/VLT,
196: %aimed at looking for a counterpart by
197: searching for variability.
198: %We have also
199: None of the candidates, however, shows a clear modulation
200: at 6.67 hours, nor has a significant long term variability.
201: Moreover, none of the candidates stands out for peculiar
202: colors with respect to the bulk of the sources detected in the field.
203: We find no compelling reasons to associate any of the candidates to 1E 1613.
204: On one side, within the frame of the binary system model
205: for the X-ray source, it is very unlikely that one of the candidates
206: be a low-mass companion star to 1E 1613. On the other side,
207: if the X-ray source is an isolated magnetar surrounded by
208: a fallback disc, we cannot exclude that the IR counterpart be hidden
209: among the candidates.
210: If none of the potential counterparts is linked to the X-ray source,
211: 1E 1613 would remain undetected in the IR down to
212: Ks$>$22.1.
213: Such an upper limit is consistent only with an extremely low-mass star
214: (an M6-M8 dwarf) at the position of 1E 1613, and makes rather problematic the
215: interpretation of 1E 1613 as an accreting binary system.
216: \end{abstract}
217:
218: %% Keywords should appear after the \end{abstract} command. The uncommented
219: %% example has been keyed in ApJ style. See the instructions to authors
220: %% for the journal to which you are submitting your paper to determine
221: %% what keyword punctuation is appropriate.
222:
223: \keywords{stars: neutron -- stars: individual (1E 161348-5055)}
224:
225: %% From the front matter, we move on to the body of the paper.
226: %% In the first two sections, notice the use of the natbib \citep
227: %% and \citet commands to identify citations. The citations are
228: %% tied to the reference list via symbolic KEYs. The KEY corresponds
229: %% to the KEY in the \bibitem in the reference list below. We have
230: %% chosen the first three characters of the first author's name plus
231: %% the last two numeral of the year of publication as our KEY for
232: %% each reference.
233:
234:
235: %% Authors who wish to have the most important objects in their paper
236: %% linked in the electronic edition to a data center may do so by tagging
237: %% their objects with \objectname{} or \object{}. Each macro takes the
238: %% object name as its required argument. The optional, square-bracket
239: %% argument should be used in cases where the data center identification
240: %% differs from what is to be printed in the paper. The text appearing
241: %% in curly braces is what will appear in print in the published paper.
242: %% If the object name is recognized by the data centers, it will be linked
243: %% in the electronic edition to the object data available at the data centers
244: %%
245: %% Note that for sources with brackets in their names, e.g. [WEG2004] 14h-090,
246: %% the brackets must be escaped with backslashes when used in the first
247: %% square-bracket argument, for instance, \object[\[WEG2004\] 14h-090]{90}).
248: %% Otherwise, LaTeX will issue an error.
249:
250: \section{Introduction}
251:
252: The X-ray point source 1E 161348$-$5055 (1E 1613 hereinafter) was
253: discovered with the \ein\ observatory close to the geometrical center
254: of the very young ($\sim2000$ years) shell-type supernova remnant
255: (SNR) RCW103 \citep{tuohy80}. The association of the point source
256: with the SNR is very robust on the basis of the good positional
257: coincidence, with 1E 1613 lying within $\sim20''$ of the SNR
258: center. Furthermore, HI observations of this region \citep{reynoso04}
259: pointed to a spatial correlation of the two objects in view of their
260: similar distance ($\sim3.3$ kpc). Historically, 1E 1613 was the first
261: radio-quiet, isolated neutron star candidate,
262: %found in a young SNR,
263: %showing only a blackbody
264: with thermal
265: X-ray spectrum, no counterparts at other
266: wavelengths, no pulsations, no non-thermal extended emission. Since
267: then, a handful of similar enigmatic sources, all characterized by a
268: %soft
269: thermal
270: X-ray spectrum, lack of standard pulsar activity,
271: high X-ray to optical flux ratio
272: (F$_X$/F$_{opt}>10^3$), and general lack of pulsations, have been
273: discovered inside young SNRs. Such sources, possibly the youngest
274: members of the family of radio-quiet, isolated neutron stars
275: (including also the Anomalous X-ray Pulsars and Soft Gamma Repeaters),
276: were dubbed, as a class, ``Central Compact Objects''
277: \citep[CCOs,][]{pavlov02} -- see \citet{deluca07} for a recent review.
278:
279: What makes 1E 1613 unique among CCOs is its very peculiar and puzzling
280: temporal behaviour. A factor 10 variability on the time
281: scale of few years was already evident within the historical \ein/\rosat/\asca\
282: dataset \citep{gotthelf99}. This was confirmed by \chan/ACIS
283: observations which showed that the source brightened by a factor
284: $\sim$60 between September 1999
285: (when the source's flux was $\sim8 \times 10^{-13}$ erg cm$^{-2}$
286: s$^{-1}$ in the 0.5-8 keV energy range) and February 2000
287: (with a record flux of $\sim5 \times 10^{-11}$ erg
288: cm$^{-2}$ s$^{-1}$),
289: %to fade down
290: decreasing
291: to $\sim10^{-11}$ erg cm$^{-2}$
292: s$^{-1}$ \citep{sanwal02} afterwards. The analysis of \chan\
293: monitoring observations showed that the source flux has been
294: continuously fading since then \citep{deluca06}. The 1999 16 ks
295: \chan\ observation, performed when the source was
296: in a ``low state'', hinted a possible $\sim$6.4 hours ($\sim$ 23 ksec)
297: periodicity \citep{garmire00}. However, subsequent \chan\ and \xmm\
298: observations found the source in ``active state''
299: with a remarkably complex
300: light curve including dips, with an overall $\sim20\%$ modulation
301: \citep{sanwal02,becker02},
302: and could not
303: ultimately confirm its periodicity. The breakthrough came with a long, 90 ks
304: \xmm\ observation \citep{deluca06}, performed when the source was in a
305: ``low state'' ($\sim2\times10^{-12}$ erg cm$^{-2}$ s$^{-1}$), which
306: clearly showed a 6.67 hour periodicity with a strong ($\sim50\%$), almost
307: sinusoidal modulation.
308: % The periodical modulation was found almost
309: %perfectly sinusoidal, with a striking flux variation between minimum
310: %($\sim1.2\times10^{-12}$ erg cm$^{-2}$ s$^{-1}$) and maximum
311: %($\sim3\times10^{-12}$ erg cm$^{-2}$ s$^{-1}$).
312:
313:
314: On the optical side, \vlt\ observations of the crowded field of 1E 1613
315: performed with \fors\ and \isaac\ identified a possible counterpart in
316: a very red object (I$>25$, J$\sim22.3$, H$\sim19.6$ and Ks$\sim18.5$)
317: located within the \chan\ error circle \citep{sanwal02}. The
318: existence of this object was also confirmed by \hst/\nicmos\ follow-up
319: observations \citep{mignani04}.
320: %However, the proposed counterpart
321: %was later discarded by \citet{pavlov04} on the basis of
322: %updated astrometry,
323: %its
324: %refined \chan\ position,
325: %with no other object within the X-ray error
326: %circle detected down to H$\sim22-23$.
327: A search for a counterpart
328: in the far IR with the Spitzer telescope was also performed,
329: with negative results \citep{wang07}.
330:
331: As discussed by
332: \citet{deluca06}, the peculiar combination of long-term variability,
333: 6.67 hours periodicity, young age and underluminous optical/IR
334: counterpart settle the case for a unique phenomenology. 1E 1613 could
335: be a very young binary system, composed of a recently born compact
336: object and of a low-mass star in an eccentric orbit, powered by an
337: unusual ``double'' (wind+disc) accretion mechanism.
338: Very recently, a different binary scenario for 1E 1613 has been proposed
339: by \citet{pizzolato08}.
340: Alternatively, 1E 1613
341: could be a peculiar isolated object, e.g. a magnetar, dramatically
342: slowed-down, possibly by interaction with a debris disc
343: (De Luca et al., 2006; see also Li, 2007, for an update of the isolated magnetar
344: model for 1E 1613).
345: % \citep[see also][for an update of the isolated magnetar model for 1E 1613]{li07}.
346: Both
347: the binary system and the isolated object
348: scenario are highly unusual and require non-standard assumptions
349: about the formation and evolution of compact objects in supernova
350: explosions.
351:
352: In order to shed light on the nature of the puzzling source 1E 1613 ,
353: we have performed new, deep IR observations of the field with the
354: \vlt, with the main aim of identifying the source counterpart by means
355: of a sensitive search for modulation at the expected 6.67 hour
356: periodicity. We have also re-analysed the archived \vlt\ and \hst\
357: observations \citep{sanwal02,mignani04} in order to search for
358: long-term variability of a possible counterpart. We presented a first
359: account of our VLT results in \citet{mignani07}.
360:
361: Since a precise position is of paramount importance for our counterpart search,
362: we have first reassessed 1E 1613 X-ray position using a set of Chandra archival
363: data together with a very recent deep Chandra observation performed by our group
364: (\S 2). Using our new position, we have analysed our new IR VLT data, as
365: well as the archival ones (\S 3 and 4). Results are discussed in \S 5.
366: %The paper is organized as follows: observations and data
367: %analysis are described in \S 2 and 3, respectively, while the results
368: %are discussed in \S4.
369:
370: \section{The X-ray position of 1E 1613}
371: %\subsection{\chan\ astrometry}
372:
373: In order to maximize the identification chances,
374: an improved X-ray position of the source is
375: required. Although 1E 1613 has been extensively observed with \chan\,
376: almost all data
377: %Several observations of 1E 1613 have been performed with \chan\
378: %since the observatory launch. However, almost all data
379: collected
380: before 2002 are of little use to derive the source
381: position owing to the presence of offsets in the astrometry, as
382: computed by the {\em Aspect tool}
383: \footnote{http://cxc.harvard.edu/cal/ASPECT/fix\_offset/fix\_offset.cgi},
384: and/or to source pile-up. Indeed,
385: we found discrepancies among the reconstructed target coordinates,
386: much larger than the expected astrometric accuracy. Thus, we decided
387: to rely only on Chandra data collected after 2002, for which no
388: astrometric offsets are reported. Such data include 12 short (4-5 ks)
389: observations with ACIS/S \citep{sanwal02,deluca06} as well as a very
390: recent, deep (80 ks) observation performed with HRC/S by our group.
391: The short observations were performed with different roll
392: angles, with the source at a large ($\sim8$ arcmin)
393: offaxis angle in order to reduce pile-up effects.
394: Data were retrieved from the Chandra X-ray Centre (CXC)
395: Science Archive. Calibrated (``level 2'') data were produced using
396: the Chandra Interactive Analysis of Observations software (CIAO
397: v.3.3). The target position was computed for each ACIS/S dataset by
398: performing a source detection in the 0.5--10 keV range using the {\em
399: wavdetect} task. After averaging the target coordinates computed from
400: each dataset, we obtained $\alpha (J2000)$=16$^h$ 17$^m$ 36.228$^s$,
401: $\delta (J2000)$= -51$^\circ$ 02' 24$\farcs7$ with a r.m.s. of
402: 0\farcs5 and 0\farcs25 in right ascension and declination,
403: respectively. The Chandra/ACIS astrometric accuracy for sources at
404: offaxis angles larger than 3 arcmin is degraded with respect to the on-axis case
405: because of PSF blurring, but no systematic studies of
406: such an effect are available. Thus, we assume the observed r.m.s. on the
407: source coordinates as the 1$\sigma$ uncertainty on the X-ray position.
408: % [QUANTO E' RELIABLE L'ERRORE ASTROMETRICO DI OGNI
409: %SINGOLA OSSERVAZIONE ? DOBBIAMO DIRE SE SI E' TENTATO DI FARE LA
410: %BORESIGHT?]
411: The target position in the deep HRC observation was
412: computed in the same way, yielding $\alpha(J2000)$=16$^h$ 17$^m$
413: 36.232$^s$, $\delta (J2000)$= -51$^\circ$ 02' 24$\farcs6$ with a
414: nominal $1\sigma$ radial uncertainty of $0\farcs41$, according
415: to the CXC calibration
416: team\footnote{http://cxc.harvard.edu/cal/ASPECT/celmon/}.
417: The HRC coordinates are
418: perfectly consistent with those computed using the 12 short ACIS/S
419: observations. Combining the two measurements, we computed the best
420: estimate of the source coordinates, i.e. $\alpha (J2000)$=16$^h$
421: 17$^m$ 36.23$^s$, $\delta (J2000)$= -51$^\circ$ 02' 24$\farcs6$ with a
422: $1\sigma$ uncertainty of $0\farcs285$ and $0\farcs185$
423: in right ascension and declination, respectively.
424:
425:
426: \section{IR Observations and data reduction}
427:
428: %In the following sections observations are described, grouped by
429: %telescope/instrument and sorted according to the observing epochs.
430:
431: \subsection{The 2006 VLT/NACO observations}
432:
433: Our new set of IR observations was performed in visitor mode on May
434: 23rd and 24th 2006
435: %(programme 077.D-0764(A), PI. A. De Luca)
436: with \nacon\ (\naco), an adaptive optics (AO) imager and spectrometer
437: mounted at the fourth Unit Telescope (UT4) of the ESO \vltn\ (\vlt) at
438: the Paranal Observatory, Chile. In order to
439: provide the best combination of angular resolution and
440: sensitivity, \naco\ was operated with the S27 camera
441: % with a corresponding
442: giving a field of view of $28''\times28''$ and a pixel scale of
443: 0\farcs027. The only suitable reference star for the adaptive optics
444: correction was the \gsc\ star S230213317483 ($V=15.2$), located
445: 21\farcs1 away from our target. For this reason, the resulting image
446: quality was not optimal and appeared to be
447: very sensitive to the atmospheric
448: conditions. The Visual ($VIS$) dichroic element and wavefront sensor
449: ($4500-10000 \: \AA$) were used. Observations were performed in the
450: $K_s$ ($\lambda=2.18 \: \mu$ ; $\Delta \lambda= 0.35 \: \mu$) filter.
451:
452: In order to monitor continuously the potential counterpart
453: candidates within the \chan\ error
454: circle covering at least two 6.67 hour cycles,
455: we obtained a total of 21 consecutive observations in the two
456: nights (see Table~\ref{NACOdata}).
457: Each observation lasted about 2300 s and was split in
458: sequences of short randomly dithered exposures with Detector
459: Integration Times (DIT) of 60 s. The
460: airmass was mostly below 1.3, while the seeing was rarely below $\sim
461: 0\farcs8$, affecting the performance of the AO. Sky
462: conditions were photometric in both nights. Night (twilight flat
463: fields) and day time calibration frames (darks, lamp flat fields) were
464: taken daily as part of the \naco\ calibration plan. Standard stars
465: from the Persson et al. (1998) fields were observed in both nights
466: for photometric calibration. The data have been processed using the
467: ESO \naco\
468: pipeline\footnote{www.eso.org/observing/dfo/quality/NACO/pipeline} and
469: the science images coadded.
470:
471:
472:
473:
474: \subsection{VLT/ISAAC archival data}
475:
476: IR observations of 1E 1613 were performed in service mode
477: %(programme 67.D-0198(A), PI. G. Garmire)
478: between April and July 2001
479: using the NIR spectro-imager \isaac\ mounted at the First Unit
480: Telescope (UT1) of \vlt.
481: The Short Wavelenght (SW) camera was used, equipped with a
482: Rockwell Hawaii 1024$\times$1024 pixel Hg:Cd:Te array, which has a
483: projected pixel size of 0\farcs148 and a field of view of
484: 152$''\times$152$''$. Observations were performed through the $J$
485: ($\lambda= 1.25 \mu$; $\Delta \lambda= 0.29 \mu$), $H$($\lambda= 1.65
486: \mu$; $\Delta \lambda= 0.30 \mu$) and $K_s$ ($\lambda= 2.16 \mu$;
487: $\Delta \lambda= 0.27 \mu$) band filters. A total of 13 observations
488: were performed in the $H$-band with the aim of pinpointing the CCO
489: counterpart through the detection of a flux modulation, as suggested
490: by the possible periodicity of the X-ray source hinted by the early
491: Chandra observations available at that epoch \citep{garmire00}.
492: %at the $\sim$ 6
493: %hours X-ray period identified by \cite{garmire00}.
494: Additional
495: pointings in the $J$ and $K_s$-bands were performed to study the
496: colors of the candidate counterpart (see Table \ref{ISAACdata}).
497:
498: %% Unfortunately, due to the Paranal scheduling constraints the
499: %% observations were spread over a few nights. and it was not possible to
500: %% cover completely at least one cycle of the X-ray period and to obtain
501: %% a complete phase coverage over different cycles.
502: To allow for
503: subtraction of the variable IR sky background each observation was
504: split in sequences of shorther dithered exposures (DIT=20 s in the
505: $H$-band and 40 s in the others).
506: The total integration times per observation were 2000 s ($J$
507: and $K_s$ bands) and 1000 s ($H$-band). All observations,
508: with the exception of the July 23rd one, were taken
509: under photometric conditions,
510: with a seeing often better than 1\farcs0 and airmass below 1.2.
511: Twilight flat fields, dark frames, as well as images of standard stars
512: from the Persson et al. (1998) fields, were taken daily as part of
513: the \isaac\ calibration plan. The data were reduced and calibrated
514: using the ESO \isaac\
515: pipeline\footnote{http://www.eso.org/observing/dfo/quality/ISAAC/pipeline}. For
516: each exposure sequence, single frames were registered and coadded to
517: produce a background subtracted and cosmic-rays free image.
518:
519:
520: \subsection{HST/NICMOS archival data}
521:
522: IR observations of the field of 1E 1613 were performed on
523: August 15th and October 8th 2002 with the \hst\
524: %(Programme 9467, PI. D.Sanwal).
525: Observations were performed with \nicmos\ using the NIC2
526: camera ($19\farcs2 \times 19\farcs2$ field of view, 0\farcs075 pixel
527: size) with the $110W$ ($\lambda = 1.128 \: \mu$, $\Delta \lambda =
528: 0.16 \: \mu$), $160W$ ($\lambda = 1.606 \: \mu$, $\Delta \lambda =
529: 0.11 \: \mu$) and $205W$ ($\lambda = 2.071 \: \mu$, $\Delta \lambda =
530: 0.18 \: \mu$) filters. To cope with visit scheduling constraints the
531: target had to be observed for 10 spacecraft orbits distributed over
532: two different visits.
533: In each visit, a sequence
534: of six exposures (2590 s each) was performed in the $160W$ filter to
535: search for variability of the originally proposed candidate
536: counterpart \citep{sanwal02}, and two exposures in the $110W$ (935 s)
537: and $205W$ (1007 s) filters to derive color information (see Table
538: \ref{NICMOSdata}).
539:
540: %% Unfortunately, this allowed to cover only
541: %% partially two cycles of the $\sim$6 hours period of the X-ray source
542: %% and to obtain only a random phase sampling.
543:
544: To decrease the instrumental overheads,
545: observations were taken in MULTIACCUM mode, split in sequences of 9
546: and 18 sub-exposures in the $110W$ filter and in the others,
547: respectively. The data were downloaded from the European \hst\
548: science archive\footnote{http://www.stecf.org/archive/} after
549: on-the-fly recalibration with the best reference files available,
550: frame coaddition and cosmic ray filtering.
551:
552:
553: \subsection{\vlt\ and \hst\ astrometry}
554:
555: In order to precisely register the \chan\ position on our IR images we
556: have refined the default image astrometry. We have computed the
557: astrometric solution on the \isaac\ images by fitting the positions
558: and coordinates of 60 reference stars selected from the \tmass\
559: catalogue. The reference stars positions have been computed by a
560: two-dimensional gaussian fitting procedure with accuracies of few
561: hundredths of pixels. The astrometric fit was performed using the
562: STARLINK package ASTROM with a six order polynomial to account for the
563: detector distorsions and yielded a rms of $\sim 0\farcs090$ in both
564: right ascension and declination. Since very few \tmass\ stars fall in
565: both the narrow \nicmos\ ($19\farcs2 \times 19\farcs2$) and \naco\
566: ($28''\times28''$) fields of view they do not provide an adequate
567: primary reference grid. For this reason, in both cases we have
568: computed the astrometric solution by using a secondary reference grid
569: made up of 26 secondary stars identified in common with the \isaac\
570: $Ks$-band image. The astrometric solutions computed for the \nicmos\
571: and \naco\ images thus yielded a rms of $0\farcs042$ and $0\farcs040$
572: per coordinate, respectively. By adding in quadrature the rms of the
573: astrometric solution of the \isaac\ image ($\sim 0\farcs090$ per
574: coordinate), we thus end up with an overall uncertainty of $\approx$
575: 0\farcs1 per coordinate on both the \nicmos\ and \naco\ images
576: astrometry. In all cases, we have accounted for the intrinsic
577: 0\farcs2 absolute astrometric accuracy of \tmass
578: \footnote{http://spider.ipac.caltech.edu/staff/hlm/2mass/overv/overv.html}.
579:
580: Fig. ~\ref{charts} shows the deepest \vlt\ and \hst\ images of the field
581: of 1E 1613 with the computed \chan\ position overlayed
582: (see also the caption to Fig. ~\ref{charts}). Seven
583: objects (labelled in the figure) are detected in the vicinity of the
584: X-ray position; three of them are consistent with the 99\% c.l. error region.
585: %%within or close ($< 3
586: %%\sigma$) to the revised \chan\ X-ray error region.
587: In all cases, the
588: objects' profiles are point-like and consistent with the instruments
589: PSFs. We note that the originally proposed counterpart object \#1
590: \citep{sanwal02} now is only marginally consistent (at $\sim 3 \sigma$
591: c.l.) with the position of
592: %revised \chan\
593: 1E 1613.
594: Due to the field crowding, only the two brightest
595: objects (\#1 and 2) are detected in the lower resolution \vlt/\isaac\
596: image (Fig. ~\ref{charts}, left), while the faintest ones (\#3-7) are
597: detected in the higher resolution \hst/\nicmos\ and \vlt/\naco\
598: images. Objects \#1 and 2 are detected also in the \vlt/\isaac\ $J$
599: and $K_s$-band images. All objects are detected in the \hst/\nicmos\
600: $160W$ and $205W$-bands, while only objects \#1 and 2 are also
601: detected in the $110W$ band, although the former is only marginally
602: detected. This is likely due to the short exposure time (see Table
603: \ref{NICMOSdata}) and to the fact that the $110W$-band is slightly bluer
604: than the \isaac\ $J$-band. The fact that objects \#3-7 are all
605: detected in the $160W$ and $205W$ bands but not in the $110W$ one
606: suggests that they are very red and heavily absorbed.
607: To be conservative, we will consider all of the seven sources in our
608: investigation.
609:
610:
611: \section{Data analysis}
612:
613: Due to the field crowding and to the faintess of most candidates, for
614: all observations magnitudes were computed through PSF photometry
615: which, in this case, yields more accurate results than
616: standard aperture photometry. We note that for the \naco\
617: observations the accuracy of the PSF photometry is affected by the
618: quality of the AO correction, which was not optimal due to the
619: relatively large offset of the guide star and to the varying
620: atmospheric conditions (\S3.1). Since the \naco\ PSF is largely
621: oversampled, to increase the signal--to--noise ratio \naco\ images
622: have been resampled with a $3\times3$ pixels window using the {\em
623: swarp} program (Bertin E., Terapix Project). For the PSF photometry
624: we used the Stetson (1992, 1994) suite of programs \dao, following the
625: procedures described in Zaggia et al. (1997).
626: %For the \nicmos\ images, where the shape of the PSF is
627: %more complex, we have used the package \roma\ which allows to
628: %interactively fit the PSF.
629: To improve and maximize the object detection we used as a reference
630: the co-added and deeper \vlt/\isaac\ $H$, \hst/\nicmos\ $160W$, and
631: \vlt/\naco\ $K_s$-band images (see Fig. ~\ref{charts}) for each
632: telescope/instrument dataset to create a master list of objects which
633: we registered on the single images and used as a mask for object
634: detection. For each single image, the model PSF was then computed by
635: fitting the profile of a number of bright, but non-saturated, reference
636: objects. Such model PSF was used to measure object fluxes at the
637: reference positions. Photometry calibration was applied using the
638: zeropoints provided by the \vlt\ and \hst\ data reduction pipelines.
639: Since the \isaac\ and \naco\ zeropoints are by default computed
640: through aperture photometry, the aperture correction was applied to
641: the \dao\ magnitudes. For the \vlt\ observation, airmass correction
642: was applied using the atmospheric extinction coefficients measured for
643: the Paranal Observatory \citep{patat04}.
644:
645: \subsection{Short term variability}
646:
647: To search for short-term variability from the candidate counterparts
648: we started from the \vlt/\naco\ dataset which is the only one to
649: provide both a complete coverage and an accurate phase sampling of the
650: 6.67 hours period of the X-ray source over at least two cycles
651: (see Table \ref{NACOdata}). Using as a mask the master list created
652: from the co-added $K_s$-band image we have run \dao\ to compute the
653: PSF photometry on the single images.
654: %As a
655: %reference, we used the sixth image of the esposure sequence obtained on
656: %the first night (see Table \ref{NACOdata}), which was taken under the
657: %best seeing and airmass conditions, and for which the AO correction
658: %was better.
659: The derived single object catalogues were then matched and compared
660: using the \dao\ routine \all. To avoid systematic offsets induced by
661: night-to-night zeropoint fluctuations the photometry of the second
662: night was renormalized to the first one.
663:
664: %The relative light curves of all the candidate counterparts for the
665: %two consecutive nights (May 23rd and 24th) are plotted in
666: %Fig.~\ref{naco_lc} (left panel), where magnitude difference were computed with
667: %respect to the average.
668: Magnitude differences with respect to the average value are plotted in
669: Fig.~\ref{naco_lc} (left panel) for all candidate counterparts
670: for the two consecutive nights (May 23rd and 24th).
671: Object \#7 is not included since it is not
672: detected in the single images but only in the co-added one (see
673: Fig. ~\ref{charts}).
674: %% As seen, only for one candidate (e.g object \#1)
675: %% a hint of a possible periodic modulation can be recognized in both the
676: %% first and the second night while, e.g. for objects \#3 and 4 the
677: %% variations are obviously inconsistent and for the other candidates are
678: %% completely erratic.
679: Next, for each measurement, we computed the corresponding phase with respect
680: to the 6.67 hours X-ray period, assuming phase 0 to be at MJD 53879.0
681: and we folded the second night on
682: the first one. The folded light curves are plotted in
683: Fig.~\ref{naco_lc} (right panel).
684:
685: A large scatter of the flux measurements is apparent in both panels of
686: Fig.~\ref{naco_lc}, where error bars account
687: for statistical errors only.
688: Such a variability is generally erratic and not correlated with phase
689: (Fig.~\ref{naco_lc}). Indeed, visual inspection of the light curves of object \#1
690: and \#3 would suggest a nearly sinusoidal variation, which, however,
691: does not pass statistical tests.
692: Considering source \#1, a simple constant fails to reproduce the folded
693: light curve ($\chi^2_{\nu}=3.9$, 18 d.o.f.). Adding a $sin$ function does not yield a better
694: description ($\chi^2_{\nu}=3.6$, 16 d.o.f.; such improvement has a 20\%
695: chance occurrence probability); adding a second harmonic does not improve
696: the situation ($\chi^2_{\nu}=4.0$, 14 d.o.f.).
697: Focusing on source \#3, a constant fit yields $\chi^2_{\nu}=1.5$ (15 d.o.f.),
698: while adding a $sin$ function yields $\chi^2_{\nu}=1.0$ (13 d.o.f.).
699: Such an improvement has a chance occurrence probability of 3\%, which is definitely
700: too high to claim for evidence of modulation.
701:
702: %Models including one or two harmonics
703: %yield a slightly better fit than a simple constant,
704: %but the chance occurrence probability of such an improvement is
705: %as large as 20\%.
706: %Thus, no clear evidences for periodic modulation are found for any object.
707: Thus, none of the candidate counterparts show evidence for periodic modulation.
708:
709: In order to estimate upper limits on short-term variability,
710: a careful discussion of errors is required.
711: Indeed, the apparent flux variations exceed the expected poisson fluctuations
712: among different measurements. A steady flux model is formally
713: not consistent with most of the observed light curves
714: (as seen above, e.g., for object 1 it yields $\chi^2 _{\nu}=3.9$, 18 d.o.f.).
715:
716: %%However, we note that these apparent variations
717: %%are always consistent with the r.m.s of the distribution, i.e. they
718: %%are not statistically significant. As a further check, for each
719: %%measurement we also computed the corresponding phase with respect to
720: %%the 6.67 hours X-ray period, assuming phase 0 to be at MJD DDDD.DD
721: %%according to the ephemeris of (...), and we folded the second night on
722: %%the first one. The relative ``folded'' light curves are plotted in
723: %%Fig.~\ref{folding}.
724: %%As expected, also this check does not yield any
725: %%statistically convincing evidence of periodicity.
726:
727: Such a large scatter suggests that our relative photometry
728: measurements are contaminated by random errors induced e.g.,
729: by fluctuations in the atmospheric conditions, sky background and
730: seeing. Other sources of errors are the variations in the AO
731: correction to the image PSF, which depend both on the objects
732: positions in the instrument field of view and on the correct guiding
733: of the reference star. Furthermore, errors are also induced by
734: the PSF fitting and background subtraction procedures. Other errors
735: induced by the data reduction process and/or by glitches in the
736: detector performance should also to be taken into account.
737: %Thus, systematic errors
738: %dominate over the purely statistical errors.
739: %attached to the single flux measurements.
740: It is also interesting to note that the brightest source
741: candidate (object \#2, $K_s$=15.5) shows a much smaller r.m.s. with
742: respect to the other, much fainter ones ($K_s> 18$), implying
743: that such errors are larger for sources with a lower signal to noise ratio.
744: %weight more or less depending on the
745: %vary with the object's brightness.
746: Since it would be extremely difficult to formally
747: quantify all the above effects on the photometry of each single
748: object, we decided to use an empirical approach. The results are
749: shown in Fig.~\ref{fig:rms} where, for all objects detected in the field,
750: we have plotted the flux variation r.m.s. as a function of
751: the average object's flux. As already hinted in
752: Fig.~\ref{naco_lc}, flux measurements for fainter
753: sources are more scattered. In particular, the flux variation r.m.s
754: of our candidate counterparts are fully consistent with those of field
755: objects of comparable brightness. Thus, we conclude that none of the
756: candidate counterparts shows evidence for a significant short-term
757: variability, nor of a significant 6.67 hour modulation.
758: %along the 6.67 hour X-ray period. For each of them, we assumed
759: Indeed, the
760: measured flux variation r.m.s. can be interpreted as a
761: $1\sigma$ upper limit to any
762: possible variability. Such upper limits, together with the
763: time-averaged magnitudes, are summarized in Table \ref{tab:rms}.
764:
765: For completeness, we have analyzed the other available datasets which
766: span different epochs and sample different source states. We have
767: repeated our analysis using the \vlt/\isaac\ $H$-band dataset (see
768: Table \ref{ISAACdata}) which, unfortunately, provides repeated flux
769: measurements only for the two brightest candidates (objects \#1 and 2)
770: with non-continuous coverage and non-unifourm
771: sampling of the 6.67 hours X-ray period. We found that the apparent flux
772: variations are compatible with the measured r.m.s., suggesting
773: that, also in this case, random errors induced by fluctuations in the
774: atmospheric conditions, sky background and seeing, dominate.
775: %our
776: %measurements. These effects are obviously enhanced by the fact that
777: %the \isaac\ observations were spread over a few nights.
778: As before, we
779: assumed the measured flux variation r.m.s. as the $1\sigma$
780: variability upper limit (see Table \ref{tab:rms}).
781:
782: Finally, we have repeated our analysis using the \hst/\nicmos\ $160W$
783: dataset which provides repeated,
784: atmosphere-free flux measurements
785: % for all our candidates
786: although with non-continuous coverage
787: and non-uniform sampling of the 6.67 hours X-ray period (see Table
788: \ref{NICMOSdata}).
789: %However, with respect to the \vlt\ ones, the \hst\
790: %observations were obviously not affected by the errors induced
791: %by the atmosphere. In principle, this should allow to set more
792: %stringent upper limits on the candidates variability.
793: Since the \hst\
794: observations have been split in two different orbits separated by
795: about 60 days, we first searched for variability
796: %in the measurements obtained from each of the two
797: in each visit.
798: %s separately.
799: Once more, we could
800: not find statistically significant evidence for variability. In
801: particular, we found none for object \#7, i.e. the faintest of the
802: candidates and the only one for which the repeated \vlt/\naco\
803: observations did not yield time variability information.
804: While the uncertainty on the source period prevents
805: folding of the light curves derived from different visits,
806: by comparing the photometry of the two visits we found
807: that the fluxes measured from the second one are systematically
808: fainter by $\sim 0.1-0.2$ magnitudes. Such an effect is most probably due
809: to the use of calibration frames non contemporaneous to the observations
810: (unfortunately, a single set of dark and flatfield frames is available
811: for the two visits) and, possibly, to a
812: non-optimal correction for the ``pedestal'' (see NICMOS Instrument
813: Handbook).
814: % We have investigated this
815: %effect and we have found that the observations performed in the two
816: %visits were reduced using the same set of dark and flatfield frames.
817: %Unfortunately, no closest in time calibration frames were available
818: %for on-the-fly recalibration. The observed offset may be thus likely
819: %due to the use of non optimal calibration frames and, possibly, to a
820: %non-optimal correction for the ``pedestal'' (see NICMOS Instrument
821: %Handbook).
822: The derived variability upper limits (see Table \ref{tab:rms})
823: are thus less constraining than those derived from the single visits,
824: although, as expected, somewhat tighter than those derived from the
825: analysis of the \vlt/\naco\ and \isaac\ datasets.
826:
827:
828: \subsection{Long term variability}
829:
830: We have used the whole dataset to search for indications of long term,
831: IR variability on time scale of years, possibly associated to the evolution
832: of the X-ray source. As a reference, we have
833: used the flux measurements in the closest passbands, i.e. those
834: obtained from the \vlt/\isaac\ $K_s$-band, \hst/\nicmos\ $205W$ and
835: \vlt/\naco\ $K_s$-band observations.
836: Since we did not find any evidence for variability within the dataset of
837: each instrument, we have generated, for each dataset, time-averaged images.
838: %Since measurements obtained days
839: %and weeks apart are consistent with a steady flux (see previous
840: %section) we have used, for each dataset, time-averaged images.
841: % and we have assumed as a reference the corresponding average epoch.
842: %The main problem in using
843: %the available datasets to search for a long term variability is
844: %obviously due to the fact that the observations were taken with
845: %different instrument/bands, which makes a direct comparison more
846: %difficult. For instance, one should take into account the different
847: %bandpasses of the \nicmos\ $205W$ filter and of the \isaac\ $K_s$ one,
848: %which yield different magnitudes for objects with different spectra.
849: Unfortunately, a direct comparison is made difficult by the use of different
850: instruments and filters.
851: Passband transformations between the \hst\ filters and the Johnson's
852: %ones are available in the literature (e.g. ..) and
853: ones can be
854: computed
855: %on-line
856: using the {\em synphot}\footnote{http://stsdas.stsci.edu/cgi-bin/gethelp.cgi?synphot.sys}
857: package of the Space Telescope Science Data
858: Analysis Software (STSDAS).
859: %\nicmos\ Units Conversion
860: %tool\footnote{http://www.stsci.edu/hst/nicmos/tools/}.
861: However, the
862: results are affected by the uncertainty on the objects spectral type
863: which is a free parameter of the transformation equation. Since
864: our spectral classification of the candidate counterparts
865: relies mostly on one colour, with only objects \#1 and 2 also
866: detected in the \vlt/\isaac\ $J$-band and in the \hst/\nicmos\ $110W$,
867: we estimate that a straight passband tranformation will introduce an
868: unknown uncertainty in our flux estimates.
869:
870: To solve the problem, as well as to account for other sources
871: of systematics, we cross-matched the object catalogues obtained from
872: the \hst/\nicmos\ $205W$-band and the \vlt/\isaac\ $K_s$-band
873: observations (202 objects in common) and we computed the correlation
874: between the magnitudes measured in the two filters. The r.m.s of the
875: fit is $\sim$ 0.05 magnitudes (using 105 sources with $H<19.5$ after
876: 2$\sigma$ clipping), i.e. of the same order of our statystical
877: photometric errors.
878: We repeated the same procedure to compute the
879: transformation between the \vlt/\naco\ and \isaac\ $K_s$-bands and
880: again we obtained an r.m.s. of $\sim$ 0.05 magnitudes
881: (using 47 sources with $K<17$, after $2.5\sigma$ clipping). For all our
882: candidates we then applied such empyrical passband
883: transformations in order to have all flux measurements consistently
884: referred to the $K_s$-band.
885:
886: %The \hst/\nicmos\ to \isaac\ and \vlt/\naco\ to \isaac\ correlations
887: %were studied and characterized on a sample of
888: %bright (Ks$\sim...$) sources in the field.
889: We also checked the correlation for the bulk of fainter
890: sources, resolved only in the sharp \hst/\nicmos\ and \vlt/\naco\
891: images. The scatter is found to be somewhat higher at faint fluxes.
892: The observed r.m.s. increases from 0.08 mag (for 19 sources in the $K_s$ 13-15.5
893: range) to $\sim0.4$ mag (for 243 sources in the $K_s$ 18-20.5 mag range).
894: The larger r.m.s. for fainter objects
895: %may be ascribed, at least in part,
896: %to genuine variability of the sources, but also
897: points to effects
898: related e.g. to passband transformation, airmass corrections, as well as to
899: the effects discussed in \S 4.1.
900: %Contamination of fainter objects' fluxes by
901: %the complex PSF wings of nearby brighter sources in the NICMOS images may
902: %also play some role.
903:
904: Fig.~\ref{multi_lc} shows the derived long term $K_s$-band lightcurve
905: for all our candidates compared with the \chan/\xmm\ X-ray lightcurve
906: obtained over the same time span. Note that, since only objects \#1
907: and 2 were detected in the \vlt/\isaac\ observations, for the fainter
908: candidates the lightcurve is based on the \hst/\nicmos\ and \vlt/\naco\
909: points only.
910: While the X-ray source was
911: continuously fading, for most candidates there is no indication of
912: IR variability on the year time scale.
913: Object \#6
914: %there is an
915: %apparently clear evidence of a $\sim 1$ magnitude decrease between
916: apparently decreases by $\sim 1$ magnitude between
917: August 2002 and May 2006. However, we note that this object falls in
918: the PSF wings of the much brighter object \#2 (see Fig.
919: ~\ref{charts}), which might have affected our photometry.
920: %We have verified [SIMONE].
921: A possible
922: %significant ($\sim 3 \sigma$)
923: $\approx0.7$ magnitudes flux decrease is also observed for objects
924: \#4 and 5 (with the latter being the only candidate to
925: fall within the 68\% c.l. X-ray error circle).
926: However, in view of
927: the larger uncertainties for fainter sources --
928: as apparent from the scatter in the
929: \hst/\nicmos\ - to - \vlt/\naco\ correlation -- such evidences
930: for variability should be taken with caution.
931:
932: %The X-ray source in the time interval sampled by the IR observation
933: %continued to fade, recovering from the 1999-2000 outburst. As apparent
934: %from Fig.~\ref{multi_lc}, the flux of 1E decreased by a factor $\sim7$
935: %between the epochs of the \vlt/\isaac\ and \vlt/\naco\ observations,
936: %and by a factor $\sim3.5$ between the epochs of the \hst/\nicmos\ and \vlt/\naco\
937: %observations.
938:
939: %Indeed, if we assume that for the \vlt/\naco\ $K_s$-band
940: %magnitudes the flux variation r.m.s (see Table \ref{rms}) is a more
941: %realistic estimate of the overall photometric error, it turns out that
942: %object \#6 is the only possible variable candidate ($\sim 3 \sigma$).
943: %In the time interval sampled by our IR observations the flux of the
944: %X-ray source.
945:
946:
947: \subsection{Color analysis}
948:
949: We used the available multi-band information to derive clues on the
950: nature of the candidate counterparts. The single band, single epoch
951: photometry catalogues derived with \dao\ were matched by \all\ to
952: create the multi-band catalogues for the \vlt/\isaac\ and
953: \hst/\nicmos\ datasets. Single epoch multi-band catalogs were finally
954: merged and magnitudes averaged. We note that since no time
955: variability has been found in the \vlt/\isaac\ and \hst/\nicmos\
956: observations (see \S 4.1), the use of average magnitudes for each
957: dataset does not affect our color analysis. To make the comparison
958: between the derived \vlt/\isaac\ and \hst/\nicmos\ color-magnitude
959: diagrams (CMDs) consistent, \nicmos\ $110W$ and $160W$ magnitudes have
960: been transformed to the \isaac\ $J$ and $H$ bandpasses using the same
961: approach applied in the previous section. The available multi-band
962: photometry of our candidates is summarized in Table \ref{multiphot}.
963: The results of our color analysis are shown in Fig. ~\ref{cmd} where
964: we plot the candidates photometry on the \vlt/\isaac\ and the
965: \hst/\nicmos\ ($J,J-K_s$) and ($H,H-K_s$) CMDs built from the
966: photometry of the field stars. As seen, nearly all candidates have
967: colors consistent with the bulk of the field stellar population. The
968: only possible exceptions are object \#7, whose colour determination is
969: however affected by the large photometry errors, and object \#6 which
970: seems to be slightly redder with respect to the CMD sequence which has
971: an average $H-K_s \sim 2$. However, this apparently peculiar deviation
972: could be partly due to the lower statistics of the \hst/\nicmos\
973: CMD, where the redder part is poorly sampled. Indeed, objects with
974: extreme $H-K_s$ appear less unusual in the much denser \vlt/\isaac\ CMD.
975: Furthermore, we warn here that object \#6, as well as perhaps \#4 and
976: 5, might be variable on the long time scale (see \S4.2). Thus, its
977: location in the CMD might not be fully representative.
978:
979:
980:
981: \subsection{Deep imaging}
982:
983: No other candidates have been identified in our deep IR imaging within
984: or close to the X-ray position apart from those indicated in Fig.
985: ~\ref{charts}. We have used our deepest images of the field, i.e.
986: those obtained from the co-addition of the repeated \vlt/\isaac\
987: $H$-band, the \hst/\nicmos\ $160W$, and the \vlt/\naco\ $K_s$-band
988: observations, to set constraining upper limits on the flux of an
989: hypotethically undetected CCO counterpart. We obtained $H\sim 23$ and
990: $K \sim 22.1$ (both at $3 \sigma$) from the \hst/\nicmos\ and
991: \vlt/\naco\ observations, respectively. These values represent the
992: deepest upper limits obtained so far for this source.
993: %We note that
994: %the \naco\ $K_s$-band upper limit is not much deeper than the
995: %corresponding \hst\ one, despite of the large amount of observing time
996: %invested on an 8m telescope. This is not surprising and it is
997: %probably due to the fact that the limiting magnitudes reachable with
998: %\naco\ critically depend on the quality of the AO correction to the
999: %image PSF. In our case, this was affected by the relative distance
1000: %between the target and the reference guide star which falls outside
1001: %the \naco\ field of view, and by the very much varying seeing
1002: %conditions.
1003:
1004:
1005:
1006: \section{Discussion}
1007:
1008: Although our comprehensive study (astrometry, variability, multi-band
1009: photometry) of the potential CCO counterparts did not single out an
1010: high confidence candidate, position-wise, object \#5 (inside the 68\% c.l. region),
1011: objects \#3 and \#6 (inside the 99\% c.l. region), as well as object \#1
1012: (marginally consistent, at the $\sim3\sigma$ level) cannot be ruled out.
1013: On the other hand, objects
1014: %\#1 (the originally suggested counterpart),
1015: \#2, \#4 and \#7 may be disregarded.
1016: %No evidence for IR variability at the 6
1017: %hours X-ray period has been found for any of them, nor for variability
1018: %on a week or months time scales, with the only possible exception of
1019: %object \#6. Furthermore, none of the objects stands out for peculiar
1020: %colors with respect to the field stellar population.
1021: The field is very crowded, with a source density in the combined \vlt/\naco\
1022: $K_s$ image of $>1.1$ objects per square arcsec at the sensitivity
1023: limit of $K_s \sim 22.1$.
1024: %position-wise, object \#5 (inside the 68\% c.l. region),
1025: %objects \#3 and \#6 (inside the 99\% c.l. region), as well as object \#1
1026: %(marginally consistent, at the $\sim3\sigma$ level)
1027: %have to be considered
1028: %as possible candidates, while objects
1029: %%\#1 (the originally suggested counterpart),
1030: %\#2, \#4 and \#7 may be disregarded as less likely. We will discuss
1031: %our results and their implications against possible scenarii for
1032: %the nature of 1E.
1033: Before discussing the implications of our results on possible pictures
1034: for 1E 1613, we note that
1035: %The bulk of the sources detected in the field show very red colors.
1036: none of the possible candidates stands out for peculiar
1037: colors with respect to the field very red stellar population. The
1038: average H-K$\sim2$ requires a large interstellar reddening,
1039: consistent with A$_V \sim 20-25$.
1040: Such a reddening is much larger than the value of A$_V \sim 3.3-6.6$
1041: expected at the distance of the X-ray source,
1042: %one expected at the
1043: %distance of the X-ray source, which is A$_V \sim 3.3-6.6$,
1044: according to
1045: the measured N$_H$ of 1E 1613 \citep{deluca06}, to the results of neutral H study towards
1046: RCW103 \citep{reynoso04} as well as to spectrophotometry of the SNR
1047: \citep[e.g. ][]{leibowitz83}.
1048:
1049: Thus, if one of the plausible candidates is indeed
1050: physically associated with 1E 1613, it must
1051: have very peculiar, red intrinsic colors.
1052:
1053: In the frame of the binary system scenario
1054: for 1E 1613 \citep{deluca06}, in principle, one could expect the companion
1055: star to be significantly different from a main sequence
1056: star of comparable mass. This could be the result of an early phase
1057: of irradiation by photons and charged particles from the newborn neutron
1058: star, which could have left the companion away from thermal equilibrium
1059: (Kelvin-Helmoltz time scale to thermal relaxation
1060: %would be of order 10$^4$ y, thus
1061: would be much larger than the age of the system).
1062: Tidal interaction along a very eccentric orbit could also play some role.
1063: %Such a picture could apply to the binary scenario discussed by
1064: %\citet{deluca06}, and .
1065: However, the hypothesis that any of the possible candidates be the companion of 1E 1613
1066: is very unlikely. The observed colors and magnitudes
1067: would require an unrealistically low temperature for the star
1068: ($\sim1000 - 1500$ K), implying (at a distance of 3.3 kpc and for A$_V=6.6$),
1069: a photospheric radius of (1-2)$\times10^{11}$ cm, exceeding the
1070: Roche lobe dimension ($\sim4\times10^{10}$ cm, assuming a 1.4 M$_\odot$ neutron star,
1071: a 0.5 M$_\odot$ companion
1072: and an orbital period of 6.67 hours) and comparable to the system orbital
1073: separation ($\sim1.5\times10^{11}$ cm under the same assumptions).
1074:
1075: Within the isolated magnetar scenario
1076: \citep{deluca06}, a fallback disc is required
1077: in order to quench the neutron star rotation to a period of 6.67 hours in
1078: $\sim2000$ yr. Could one of the possible candidates
1079: be the fallback disc itself? Evidence for a debris disc surrounding
1080: the anomalous X-ray pulsar 4U 0142+61, recently obtained \citep{wang06},
1081: make such an hypothesis not unrealistic \citep[a faint IR source
1082: at the position of the CCO in Vela Jr. could also be related to a debris disc
1083: surrounding the compact object, see][]{mignani07b}. While the physics
1084: of the possible disc surrounding AXP 4U 0142+61 is not understood
1085: \citep[passive or viscous? ][]{wang06,ertan07}, we note that
1086: the colors of our candidates are similar to the case of 4U 0142+61,
1087: but their F$_{Ks}$/F$_X$
1088: ratio (in the range 0.7-2$\times10^{-3}$
1089: at the epoch of the NACO observations) is about one order
1090: of magnitude larger. The upper limits to the far IR emission set by
1091: \citet{wang07} are not constraining.
1092:
1093: Unfortunately, mostly because of the faintness of the IR sources, results of the
1094: temporal analysis could not offer conclusive clues.
1095: %While the detection of 6.67 hr periodicity would have unambiguously nailed down
1096: %the source counterpart,
1097: The upper limits we could set
1098: do not put stringent constraints. They could be consistent
1099: with a binary scenario, since any orbital modulation would obviously
1100: depend on the inclination of the system with respect to the plane of the sky.
1101: On the other side, the isolated neutron star scenario
1102: does not allow for firm predictions about a possible IR periodicity.
1103:
1104: We also tried to exploit the source long-term fading seen in X-rays
1105: between the NICMOS and the NACO observations looking for long-term
1106: variability to pinpoint the IR counterpart of 1E 1613.
1107: %The factor $\sim3.5$ decrease in the X-ray flux between the
1108: %epochs of the NICMOS and NACO
1109: %observations lead to expect some correlated long-term variability
1110: %for the IR counterpart of 1E 1613.
1111: In the binary scenario, IR variability could be due to the reprocessing of
1112: the compact source X-ray radiation by the companion star (and possibly by an
1113: accretion disc, if any). The same could be true for the isolated magnetar
1114: scenario, X-ray reprocessing taking place in a fallback disc.
1115: We stress, yet again, that in both pictures the effect would be strongly
1116: dependent on the geometry of the system with respect to the line of sight.
1117: Indeed, we obtained evidence for a possible $\sim1$ mag fading of one of the candidates,
1118: which
1119: yields a constant F$_{Ks}$/F$_X$ ratio in case of association to 1E 1613.
1120: %would be consistent (in case of association to 1E 1613) with a constant F$_{Ks}$/F$_X$ ratio.
1121: %which could be consistent with the above expectations.
1122: However, in view of the scatter seen in the correlation of
1123: NICMOS and NACO photometric measurements for sources of comparable magnitude, such a possible
1124: evidence for IR variability is not strong enough to claim an identification.
1125:
1126: %The observations we analysed were collected along a $\sim6$ years time span.
1127: %The X-ray emission of 1E dramatically faded in the same time span, its flux
1128: %at the epoch of the NACO observation being a factor $\sim3.5$ and a factor
1129: %$\sim7$ lower than at the epoch of the NICMOS and ISAAC observations,
1130: %respectively.
1131: %... in binary scenario variation
1132: %... in isolated scenario disc IR should vary
1133: %... src 6 has a possible variation of the ``correct'' factor, but difficult
1134: % to assess wether it is significant
1135:
1136: %The most likely conclusion of our investigation is that none of the possible
1137: %candidates is physically associated to the X-ray source and that 1E 1613 remains undetected
1138: %in the IR down to Ks$>$22.1.
1139:
1140: With no compelling reasons to associate any of the possible
1141: candidates to 1E 1613, we explore the implications of a source non-detection down
1142: to Ks$>$22.1.
1143:
1144: %Such a result has important implications for the binary
1145: %scenario.
1146: Considering the binary scenario, accounting for
1147: uncertainties on the distance and reddening,
1148: such upper limit is only consistent with an M6-M8 dwarf,
1149: i.e. a very underluminous companion.
1150: %The upper limit points to a very underluminous
1151: %companion, being only consistent with an M6-M8 dwarf (accounting for
1152: %uncertainties on the distance and reddening).
1153: It is rather unlikely that such a small star may power the mechanism
1154: proposed by \citet{deluca06},
1155: %for the binary scenario,
1156: where accretion
1157: of wind from the companion --
1158: modulated along an eccentric orbit --
1159: %is proposed to
1160: explains the
1161: phenomenology of the low state. Such a picture would require
1162: an accretion rate of $\sim10^{-13}\,M_{\odot}\,y^{-1}$,
1163: implying a red dwarf wind mass loss
1164: larger by at least a factor of a few, which seems somewhat too
1165: high for such a small star.
1166: Thus, an alternative process is required to
1167: explain the X-ray phenomenology.
1168: Even the survival to the supernova explosion of a
1169: binary system with such an extreme mass ratio
1170: seems problematic.
1171: It would require an ad-hoc kick for the neutron star
1172: to avoid disruption of the system.
1173: %We note, however, that
1174: %the position of 1E 1613, within $\sim$15$''$ of the geometrical center of RCW103,
1175: %sets a rather tight upper limit to the (projected) velocity
1176: %of $120d_{3.3} \tau_{2000} ^{-1}$ km s$^{-1}$, where $d_{3.3}$ is
1177: %the distance in units of 3.3 kpc and $\tau_{2000}$ is the
1178: %age in units of 2000 y.
1179:
1180: On the other side, the lack of an IR counterpart fits well within the
1181: isolated magnetar scenario. The upper limit to the ratio
1182: F$_{Ks}$/F$_X<1.5\times10^{-4}$
1183: (at the epoch of the NACO observation) is fully consistent with the values
1184: observed for all magnetars identified in the IR
1185: \citep[see e.g.][]{fesen06}, including 4U 0142+61
1186: \citep{wang06}.
1187: In the isolated magnetar scenario,
1188: we cannot rule out the possibility that one of the candidate sources be
1189: the residual disc surrounding 1E 1613.
1190: However, Occam's razor argues against such a conclusion, since
1191: all possible candidates are undistinguishable from normal, background stars.
1192: %The large chance alignement probability, coupled to the lack
1193: %of peculiar colors and of unambiguous variability for the possible
1194: %counterparts favor their interpretation as unrelated, background stars.
1195:
1196: Recent SWIFT monitoring of 1E 1613 shows that the X-ray source continued to fade in
1197: the 2006-2007 time interval
1198: and that its flux is approaching the 1999, pre-outburst level (see Fig.~\ref{multi_lc}).
1199: Historical observations point to a large flux variability
1200: over a $\sim10$ yr time scale \citep{gotthelf99,deluca06}, so that a rebrightening
1201: is likely to occur. A factor $\sim100$ increase in the X-ray luminosity
1202: (as in the 1999-2000 outburst) would yield a dramatic change in the irradiation
1203: %conditions
1204: of any possible object/structure (companion star/fallback disc)
1205: linked to 1E 1613.
1206: Thus, if a new outburst from 1E 1613 will occur,
1207: a fast follow-up in the IR will be crucial in order to
1208: conclusively address the issue of the IR counterpart for 1E 1613
1209: and to shed light on its nature. The images described in this work
1210: will be a reference to search for variability
1211: of the counterpart.
1212: %Even a negative result, on the other side, will be
1213: %useful in constraining the properties of a disc and/or of a companion star.
1214:
1215:
1216: %\pagebreak
1217:
1218: \acknowledgments
1219: This research is partially supported by the Italian Space Agency (contract
1220: ASI I/011/07/0 in support to the Swift mission).
1221:
1222:
1223:
1224: \begin{thebibliography}{}
1225: \bibitem[Becker \& Aschenbach(2002)]{becker02} Becker, W., Aschenbach, B.,
1226: 2002, MPE report 278, 64
1227: \bibitem[Cram\'er(1945)]{stat} Cram\'er, H., 1945, Mathematical methods of
1228: statistics, Princeton University Press, Princeton (USA)
1229: \bibitem[De Luca(2007)]{deluca07} De Luca, A.,
1230: 2007, in Proc. of ``40 years of Pulsars: Millisecond pulsars,
1231: Magnetars, and more'', AIP Conf.Ser., ed. by Z. Wang,
1232: C. Bassa, A. Cummings, A, Kaspi, V., in press, arXiv:0712.2209
1233: \bibitem[De Luca et al.(2006)]{deluca06} De Luca, A., Caraveo, P.A.,
1234: Mereghetti, S., Tiengo, A., Bignami, G.F., 2006, Science 313, 814
1235: \bibitem[Ertan et al.(2007)]{ertan07} Ertan, \"U., Erkut, M.H., Eksi, K.Y., Alpar, M.A., 2007, ApJ657, 441
1236: \bibitem[Fesen et al.(2006)]{fesen06} Fesen, R.A., Pavlov, G.G., Sanwal, D., 2006, ApJ 636, 848
1237: \bibitem[Garmire et al.(2000)]{garmire00} Garmire, G.P., Pavlov, G.G., Garmire, A.B., Zavlin, V.E., 2000,
1238: IAU Circ. 7350
1239: \bibitem[Gotthelf et al.(1999)]{gotthelf99} Gotthelf, E.V., Petre, R.,
1240: Vasisht, G., 1999, ApJ 514, L107
1241: \bibitem[Leibowitz \& Danziger(1983)]{leibowitz83} Leibowitz, E.M. \&
1242: Danziger, I.J., 1983, MNRAS 204, 273
1243: \bibitem[Li(2007)]{li07} Li, X.-D., 2007, ApJ 666, L81
1244: \bibitem[Mignani et al.(2004)]{mignani04} Mignani, R.P., De Luca, A.,
1245: Caraveo, P.A., 2004, in ``Young Neutron Stars and their environments'', ASP
1246: Conference Series, ed. by F. Camilo, B.M. Gaensler, San Francisco:
1247: ASP, p.391
1248: \bibitem[Mignani et al.(2007a)]{mignani07} Mignani, R.P., Zaggia, S.,
1249: Dobryzca, D., et al., 2007a, in Proc. of ``40 years of Pulsars: Millisecond pulsars,
1250: Magnetars, and more'', AIP Conf.Ser.,
1251: ed. by Z. Wang, C. Bassa, A. Cummings, V. Kaspi, in press
1252: \bibitem[Mignani et al.(2007b)]{mignani07b} Mignani, R.P., De Luca, A., Zaggia,
1253: S., et al., 2007b, A\&A 473, 883
1254: \bibitem[Patat(2004)]{patat04} Patat, F., 2004, The Messenger 118, 11
1255: \bibitem[Pavlov et al.(2002)]{pavlov02} Pavlov, G.G., Sanwal, D., Garmire,
1256: G.P., Zavlin, V.E., 2002, in ``Neutron Stars in Supernova Remnants'', ASP
1257: Conference Series, ed. by P.O. Slane and B.M. Gaensler, San Francisco:
1258: ASP, p.247
1259: \bibitem[Pavlov et al.(2004)]{pavlov04} Pavlov, G.G., Sanwal, D., Teter,
1260: M.A., 2004, in ``Young Neutron Stars and their environments'', ASP
1261: Conference Series, ed. by F. Camilo, B.M. Gaensler, San Francisco:
1262: ASP, p.236
1263: \bibitem[Persson et al.(1998)]{persson98} Persson, S.E., Murphy, D.C.,
1264: Krzeminski, W., Roth, M., Rieke, M.J., 1998, AJ 116, 2475
1265: \bibitem[Pizzolato et al.(2008)]{pizzolato08} Pizzolato, F., Colpi, M., De Luca, A.,
1266: Mereghetti, S., Tiengo, A., 2008, ApJ in press, arXiv:0803.1373
1267: \bibitem[Reynoso et al.(2004)]{reynoso04} Reynoso, E.M., Green, A.J.,
1268: Johnston, S., et al., 2004, PASA 21, 82
1269: \bibitem[Sanwal et al.(2002)]{sanwal02} Sanwal, D., Garmire, G.P., Garmire,
1270: A., Pavlov, G.G., Mignani, R., 2002, BAAS 34, 764
1271: \bibitem[Tuohy \& Garmire(1980)]{tuohy80} Tuohy, I., Garmire, G.P., 1980,
1272: ApJ 239, L107
1273: \bibitem[Wang et al.(2006)]{wang06} Wang, Z., Chakrabarty, D., Kaplan, D.L., 2006, Nature 440, 772
1274: \bibitem[Wang et al.(2007)]{wang07} Wang, Z., Kaplan, D.L., Chakrabarty, D.,
1275: 2007, ApJ 655, 261
1276: \bibitem[Zaggia et al.(1997)]{zaggia97} Zaggia, S.R., Piotto, G., Capaccioli, M., 1997, A\&A 327, 1004
1277: \end{thebibliography}
1278:
1279: \clearpage
1280:
1281: \begin{table}
1282: \begin{center}
1283: \caption{Summary of the \vlt/\naco\ observations of the field of 1E 1613, with the observing epochs, the observations start time (UT), the filter, the exposure times, the average seeing and airmass values.eld. }
1284: \begin{tabular}{cccccc} \\ \hline
1285: yyyy-mm-dd & Time (UT) & Filter & T (s) & Seeing (``) & Airmass \\ \hline
1286: 2006-05-24 & 00:49:53 & Ks & 2280 & 0.84 & 1.57\\
1287: & 01:57:36 & Ks & 2040 & 0.64 & 1.36\\
1288: & 02:45:14 & Ks & 1200 & 1.10 & 1.23\\
1289: & 03:14:21 & Ks & 2280 & 1.02 & 1.18\\
1290: & 04:04:24 & Ks & 2280 & 0.91 & 1.13\\
1291: & 04:54:17 & Ks & 2280 & 0.81 & 1.12\\
1292: & 05:44:25 & Ks & 1200 & 0.80 & 1.14\\
1293: & 06:11:50 & Ks & 2280 & 0.90 & 1.16\\
1294: & 07:02:02 & Ks & 720 & 1.06 & 1.24\\
1295: & 07:19:58 & Ks & 2280 & 0.95 & 1.28\\
1296: & 08:09:50 & Ks & 1920 & 1.00 & 1.44\\
1297: & 08:56:06 & Ks & 360 & 0.89 & 1.67\\
1298: 2006-05-25 & 00:55:12 & Ks & 2280 & 0.61 & 1.54\\
1299: & 01:57:43 & Ks & 2280 & 0.66 & 1.34\\
1300: & 02:47:40 & Ks & 2280 & 0.60 & 1.22\\
1301: & 03:37:25 & Ks & 2280 & 0.65 & 1.15\\
1302: & 04:27:26 & Ks & 2280 & 0.64 & 1.12\\
1303: & 05:17:22 & Ks & 2280 & 0.62 & 1.12\\
1304: & 06:08:53 & Ks & 2280 & 1.00 & 1.16\\
1305: & 06:59:57 & Ks & 2280 & 0.96 & 1.24\\
1306: & 07:50:24 & Ks & 600 & 0.72 & 1.38\\
1307: & 08:06:16 & Ks & 2280 & 1.00 & 1.44\\ \hline
1308: \end{tabular}
1309: \label{NACOdata}
1310: \end{center}
1311: \end{table}
1312:
1313: \begin{table}
1314: \begin{center}
1315: \caption{Summary of the \vlt/\isaac\ observations of the field of 1E 1613, with the observing epochs, the observations start time (UT), the filter, the exposure times, the average seeing and airmass values. }
1316: \begin{tabular}{cccrcc} \\ \hline
1317: yyyy-mm-dd & Time (UT) & Filter & T (s) & Seeing (``) & Airmass \\ \hline
1318: 2001-04-10 & 07:33:07 & J & 2000 & 0.68 & 1.12 \\
1319: & 07:33:07 & J & 2000 & 0.68 & 1.12 \\
1320: 2001-05-12 & 06:04:33 & Ks & 2000 & 0.95 & 1.13 \\
1321: 2001-05-13 & 04:42:57 & Ks & 2000 & 1.00 & 1.13 \\
1322: & 05:39:35 & Ks & 2000 & 1.41 & 1.12 \\
1323: 2001-06-06 & 05:55:32 & H & 1000 & 0.64 & 1.28 \\
1324: & 06:46:57 & H & 200 & 1.04 & 1.34 \\
1325: 2001-07-23 & 01:48:12 & H & 1000 & 1.03 & 1.14 \\
1326: & 02:20:50 & H & 1000 & 0.98 & 1.18 \\
1327: & 02:50:03 & H & 1000 & 1.02 & 1.22 \\
1328: 2001-07-29 & 00:17:16 & H & 1000 & 0.73 & 1.12 \\
1329: & 02:45:57 & H & 950 & 0.70 & 1.26 \\
1330: & 03:14:54 & H & 1000 & 0.84 & 1.33 \\
1331: & 03:43:08 & H & 1000 & 0.70 & 1.43 \\
1332: 2001-07-30 & 00:27:13 & H & 1000 & 1.00 & 1.12 \\
1333: & 01:04:19 & H & 1000 & 1.00 & 1.13 \\
1334: & 01:32:56 & H & 1000 & 1.00 & 1.15 \\
1335: & 02:01:25 & H & 1000 & 1.00 & 1.18 \\ \hline
1336: \end{tabular}
1337: \label{ISAACdata}
1338: \end{center}
1339: \end{table}
1340:
1341:
1342: \begin{table}
1343: \begin{center}
1344: \caption{Summary of the \hst/\nicmos\ observations of the field of 1E 1613, with the observing epochs, the observations start time (UT), the filter and the exposure times. }
1345: \begin{tabular}{cccr} \\ \hline
1346: yyyy-mm-dd & Time (UT) & Filter & T (s) \\ \hline
1347: 2002-08-15 & 02:25:38 & $160W$ & 2590 \\
1348: & 04:10:28 & $160W$ & 2590 \\
1349: & 06:05:15 & $160W$ & 2590 \\
1350: & 09:09:37 & $160W$ & 2590 \\
1351: & 11:12.32 & $110W$ & 935 \\
1352: & 12:40:18 & $205W$ & 1007 \\
1353: 2002-10-08 & 08:50:00 & $160W$ & 2590 \\
1354: & 10:27:32 & $160W$ & 2590 \\
1355: & 12:04:01 & $160W$ & 2590 \\
1356: & 13:40:14 & $160W$ & 2590 \\
1357: & 15:19:28 & $110W$ & 935 \\
1358: & 15:42:17 & $205W$ & 1007 \\ \hline
1359: \end{tabular}
1360: \label{NICMOSdata}
1361: \end{center}
1362: \end{table}
1363:
1364:
1365:
1366:
1367: \begin{table}
1368: \begin{center}
1369: \caption{Time-averaged magnitudes and associated r.m.s. (in parenthesis) for all candidate counterparts. Different columns refer to different telescope/instrument/filter combinations. }
1370: \begin{tabular}{ccccc} \\ \hline
1371: Id. & \vlt/\naco\ & \vlt/\isaac\ & \hst/\nicmos\ \\
1372: & ($K_s$) & ($H$) & ($160W$) \\ \hline
1373: 1 & 18.01 (0.09)& 19.082 (0.29) & 19.51 (0.09)\\
1374: 2 & 15.50 (0.03)& 16.411 (0.03) & 16.38 (0.02)\\
1375: 3 & 19.71 (0.20)& - & 21.25 (0.08)\\
1376: 4 & 19.71 (0.21)& - & 21.07 (0.04)\\
1377: 5 & 19.66 (0.37)& - & 21.44 (0.16)\\
1378: 6 & 18.56 (0.27)& - & 20.49 (0.16)\\
1379: 7 & - & - & 21.43 (0.07) \\\hline
1380: \end{tabular}
1381: \label{tab:rms}
1382: \end{center}
1383: \end{table}
1384:
1385:
1386: \begin{table}
1387: \begin{center}
1388: \caption{Multi-band magnitudes for all the candidate counterparts.
1389: Values are computed on the average images. To allow for an easier comparison
1390: between different measurements, \hst\ magnitudes have been renormalized to the
1391: Johnson's system (see text). Quoted uncertainties include statistical errors
1392: only.}
1393: \begin{tabular}{c|ccc|ccc|c} \\ \hline
1394: ID &\multicolumn{3}{c}{\isaac} & \multicolumn{3}{c}{\nicmos} & \multicolumn{1}{c}{\naco} \\
1395: & J & H & K & J& H & K & K \\ \hline
1396: \hline
1397: 1 & 22.10$\pm$0.10 & 19.36$\pm$0.03 & 17.93$\pm$0.02 & 22.10$\pm$0.30 & 19.50$\pm$0.02 & 17.98$\pm$0.02 & 17.94$\pm$0.03\\
1398: 2 & 17.94$\pm$0.01 & 16.38$\pm$0.01 & 15.49$\pm$0.01 & 18.05$\pm$0.01 & 16.38$\pm$0.03 & 15.52$\pm$0.02 & 15.42$\pm$0.02\\
1399: 3 & - & - & - & - & 21.25$\pm$0.20 & 19.72$\pm$0.04 & 19.69$\pm$0.07\\
1400: 4 & - & - & - & - & 21.07$\pm$0.01 & 19.17$\pm$0.21 & 19.92$\pm$0.09\\
1401: 5 & - & - & - & - & 21.43$\pm$0.01 & 19.22$\pm$0.21 & 19.89$\pm$0.08\\
1402: 6 & - & - & - & - & 20.49$\pm$0.17 & 17.70$\pm$0.12 & 18.74$\pm$0.05\\
1403: 7 & - & - & - & - & 21.43$\pm$0.15 & 20.57$\pm$1.27 & 20.23$\pm$0.12\\ \hline
1404: \end{tabular}
1405: \label{multiphot}
1406: \end{center}
1407: \end{table}
1408:
1409: %\pagebreak
1410:
1411: \clearpage
1412:
1413:
1414: \begin{figure*}
1415: \includegraphics[angle=0,width=7cm]{f1a.eps}
1416: \includegraphics[angle=0,width=7cm]{f1b.eps}
1417: \includegraphics[angle=0,width=7cm]{f1c.eps}
1418: \caption{Inner part of the field of 1E 1613 as
1419: observed by the \vlt$/$\isaac\ ($H$-band), the \hst$/$\nicmos\
1420: ($160W$) and \vlt$/$\naco\ ($K_s$).
1421: The black box on the \vlt$/$\isaac\ image marks the portion of
1422: the field shown in the \hst$/$\nicmos\ and \vlt$/$\naco\ images.
1423: North to the top, East to the left. In each case, the
1424: images are the result of the co-addition of repeated integrations (see
1425: Tab. 1--3), with corresponding total integration times of 14000s, 10360
1426: s and 40000 s, respectively. The inner ellipse corresponds to the 68\% c.l.
1427: error region, while
1428: %and accounts for the overall uncertainty
1429: %of the IR image astrometry calibration (\S 3.4).
1430: the outer ellipse corresponds to the 99\% c.l. region.
1431: The semiaxes of the 68\% and 99\% ellipses were computed by summing
1432: (in quadrature) the uncertainty on the IR image astrometric
1433: calibration (see \S3.4) to the uncertainty on the Chandra coordinates
1434: (\S2), then multiplying the resulting values by $(-2log(1-0.68))^{(1/2)}$
1435: and by $(-2log(1-0.99))^{(1/2)}$, respectively \cite[see e.g.][]{stat}.
1436: %The circles (0\farcs6 radius) correspond
1437: %to the source X-ray position as determined from the \chan\ astrometry
1438: %(see \S3.1) and accounts for the overall uncertainty of the IR image
1439: %astrometry calibration (\S 3.2).
1440: Candidate counterparts detected
1441: within or close to the X-ray error region are numbered. Object \#1 is
1442: the originally proposed counterpart of \cite{sanwal02}. Objects 3$\div$7
1443: are resolved only in the \hst$/$\nicmos\ and \vlt$/$\naco\ images.
1444: \label{charts}}
1445: \end{figure*}
1446:
1447:
1448: %\begin{figure}
1449: %\includegraphics[angle=-90,width=8cm]{rcw_naco_lc.ps}
1450: %\caption{Top to bottom: \vlt$/$\naco\ $K_s$-band lightcurves for the candidate
1451: % counterparts \#1-6. Difference with respect to the average
1452: %magnitude is plotted as a function of the time. Horizontal dotted lines mark
1453: %the r.m.s. variability for each source.
1454: %\label{naco_lc}}
1455: %\end{figure}
1456:
1457: %\begin{figure}
1458: %\includegraphics[angle=-90,width=8cm]{rcw_naco_folding.ps}
1459: %\caption{Top to bottom: folded \vlt$/$\naco\ $K_s$-band lightcurves for the
1460: % candidate counterparts \#1-6. Black and red points represent
1461: % flux measurements performed in the first and in the second night,
1462: %respectively. Two phase intervals are plotted for clarity.
1463: %\label{folding}}
1464: %\end{figure}
1465:
1466: \begin{figure}
1467: \includegraphics[angle=-90,width=9cm]{f2a.eps}
1468: \includegraphics[angle=-90,width=8cm]{f2b.eps}
1469: \caption{(left)
1470: Top to bottom: \vlt$/$\naco\ $K_s$-band lightcurves for the candidate
1471: counterparts \#1-6. Difference with respect to the average
1472: magnitude is plotted as a function of the time. Horizontal dotted lines mark
1473: the r.m.s. variability for each source. (right) Top to bottom: folded \vlt$/$\naco\ $K_s$-band lightcurves for the
1474: candidate counterparts \#1-6. Black and red points represent
1475: flux measurements performed in the first and in the second night,
1476: respectively. Two phase intervals are plotted for clarity.
1477: Error bars
1478: in both panels account for statistical uncertainties only.
1479: The r.m.s. variability for each source - plotted in the left panel
1480: only - is representative of the random errors affecting our
1481: measurements (see text).
1482: \label{naco_lc}}
1483: \end{figure}
1484:
1485:
1486: %\begin{figure}
1487: %\includegraphics[angle=-90,width=8cm]{totfolded.eps}
1488: %\caption{({\em alternativa a fig.3}) Top to bottom: folded \vlt$/$\naco\ $K_s$-band lightcurves for the
1489: % candidate counterparts \#1-6. The specific flux $F_{\lambda}$ (@2.16$\mu$), renormalized to
1490: %its average value, is plotted as a function of the
1491: %phase along the 6.67 hour cycle. Two phase intervals are plotted for clarity.
1492: %\label{folding2}}
1493: %\end{figure}
1494:
1495: \begin{figure}
1496: \includegraphics[angle=0,width=8cm]{f3.eps}
1497: \caption{The r.m.s. variability for all sources detected in the \vlt$/$\naco\
1498: field is plotted as a function of the source's average magnitude.
1499: A larger variability for fainter sources is apparent.
1500: Black
1501: squares represent the possible candidate counterparts (sources 1-6, see text).
1502: \label{fig:rms}}
1503: \end{figure}
1504:
1505:
1506:
1507: \begin{figure}
1508: \includegraphics[angle=0,width=7cm]{f4.eps}
1509: \caption{Upper panel: X-ray lightcurve of 1E 1613 in the time range
1510: 1999, September to 2007, July,
1511: updated from
1512: \citet{deluca06}. Data have been
1513: collected with the Chandra/ACIS, Chandra/HRC, XMM-Newton/EPIC and Swift/XRT
1514: instruments.
1515: Swift/XRT as well as Chandra/HRC fluxes have been computed
1516: assuming the source spectral shape to be the same as observed by XMM-Newton
1517: in 2005 \citep{deluca06}. Error bars are at $1\sigma$ confidence level.
1518: Details on the analysis of the X-ray dataset will be
1519: reported elsewhere. Vertical dotted lines mark the epochs of the IR $K_s$ observations.
1520: Middle panel: $K_s$-band lightcurves for candidate counterparts
1521: \#1,2,3,7. Error bars
1522: account for $1\sigma$ statistical uncertainty. Horizontal dotted lines
1523: mark the r.m.s. observed in the correlation of NICMOS and NACO
1524: measurements for sources in the K$\sim$18.5-20 magnitude range,
1525: which could be used as an estimate of the overall uncertainty.
1526: Lower panel: same as middle panel for candidate counterparts \#4-6,
1527: showing a possible flux decrease.
1528: The possible variation for source \#6 is about twice the
1529: uncertainty in the NICMOS-to-NACO comparison.
1530: \label{multi_lc}}
1531: \end{figure}
1532:
1533: \begin{figure*}
1534: \includegraphics[angle=0,width=10cm]{f5a.eps} \\
1535: \includegraphics[angle=0,width=10cm]{f5b.eps}
1536: \caption{($J,J-K_s$) and ($H,H-K_s$) CMDs of the field of 1E 1613 obtained from \vlt/\isaac\ (upper pair) and the \hst/\nicmos\ (lower pair) observations. The locations of the candidate counterparts (see Fig.~\ref{charts}) are plotted in red.
1537: \label{cmd}}
1538: \end{figure*}
1539:
1540: \end{document}
1541:
1542:
1543: