1: %\documentstyle[aaspp4,epsf]{article} % this is for small print
2: %\documentstyle[12pt, aaspp4]{article}
3: %%\documentstyle[aaspp4]{article}
4: %\documentstyle[aas2pp4,epsf]{article}
5: %\documentstyle[12pt,aaspp]{article} % this is for a preprint
6: %(single-spaced\)
7: %\documentstyle[11pt,aaspp4]{article} % this is for smallfont preprint
8:
9: %\documentstyle[12pt,aasms4]{article} % this is for submittal
10: %(double-spaced)
11: \documentclass[12pt,preprint]{aastex} % These are for v5.0
12: %\documentclass{aastex} % These are for v5.0
13: %\documentclass[preprint]{aastex} % but some stuff doesn't work!
14: %\documentclass[preprint2]{aastex}
15:
16: \newcommand{\ergcm}{ergs~cm$^{-2}$~s$^{-1}~$}
17: \newcommand{\ergs}{ergs~s$^{-1}~$}
18: \newcommand{\mdot}{$\dot{{m}}$ }
19: \def\ni{\noindent}
20: \def\about{$\sim$}
21: \def\Msun{$M_\odot$}
22: \def\erg/cm2sec{ergs~cm$^{-2}$~s$^{-1}$}
23: \def\ergcm2{ergs~cm$^{-2}$}
24: \def\cm2{cm$^2$}
25: \def\la{\hbox{\rlap{\raise.3ex\hbox{$<$}}\lower.8ex\hbox{$\sim$}\ }}
26: \def\ga{\hbox{\rlap{\raise.3ex\hbox{$>$}}\lower.8ex\hbox{$\sim$}\ }}
27: \def\Lx{L$_x$}
28: \def\deg{$^{\circ}$~}
29:
30: %\def\/cm2sec{~cm$^{-2}$~s$^{-1}$}
31:
32: \received{}
33: \accepted{}
34: %\journalid{}{}
35: %\articleid{}{}
36: %\pagestyle{empty}
37: \begin{document}
38:
39:
40: \title{EXITE2 Observation of the SIGMA Source GRS 1227+025}
41:
42: \author{J.E. Grindlay}
43: \affil{Harvard-Smithsonian Center for Astrophysics, Cambridge, MA 02138}
44:
45: \author{Y. Chou}
46: %\affil{Graduate Institute of Astronomy, National Central University,\\
47: %Chung-li, Taiwan}
48: \affil{National Tsing Hua University, Hsinchu, Taiwan, Republic of China }
49:
50: \author{P.F. Bloser}
51: \affil{NASA/Goddard Space Flight Center, Greenbelt, MD 20771 }
52:
53: %\author{J.A. Jenkins}
54: %\affil{Harvard-Smithsonian Center for Astrophysics, Cambridge, MA 02138}
55:
56: \and
57: \author{T. Narita}
58: \affil{College of the Holy Cross, Worcester, MA 01610}
59:
60: \email{jgrindlay@cfa.harvard.edu}
61:
62: \begin{abstract}
63:
64: We report the EXITE2 hard X-ray imaging of the sky around
65: 3C273. A 2h observation on May 8, 1997, shows a $\sim$260 mCrab source
66: detected at $\sim$4$\sigma$ in each of two bands (50-70 and 70-93 keV)
67: and located $\sim$30' from 3C273 and consistent in
68: position with the SIGMA source GRS1227+025. The EXITE2 spectrum
69: is consistent with a power law with photon index 3 and large low
70: energy absorption, as indicated by the GRANAT/SIGMA results. No source was
71: detected in more sensitive followup EXITE2 observations in 2000 and
72: 2001 with 3$\sigma$ upper limits of 190 and 65 mCrab, respectively.
73: Comparison with the flux detected by SIGMA shows the source to be
74: highly variable, suggesting it may be non-thermal and beamed and thus
75: the first example of a ``type 2'' (absorbed) Blazar. Alternatively it
76: might be (an unprecedented) very highly absorbed binary system
77: undergoing accretion disk instability outbursts, possibly either
78: a magnetic CV, or a black hole X-ray nova.
79:
80:
81: \end{abstract}
82:
83: \section{Introduction}
84:
85:
86: GRS1227+025 was discovered serendipitously with SIGMA
87: \citep{Bassani91, Jourdain92} in the 40-80 keV energy band during an
88: observation of the QSO 3C273. Separated from 3C273 by only $\sim$15',
89: GRS1227+025 (hereafter GRS1227) was detected at 5.5$\sigma$ whereas
90: 3C273 was not detected. The source has a significantly steeper
91: spectrum (photon index 3$^{+1.3}_{-0.9}$, \cite{Bassani91})
92: compared to 3C273, which was detected in subsequent observations, with
93: its usual flatter photon index of $\sim$1.5. Upper limits (15-30 keV)
94: with ART-P (simultaneous with SIGMA) showed that the spectrum must be
95: heavily absorbed: N$_H \sim$1.5$\times$10$^{25}$ cm$^{-2}$ for a power
96: law spectrum with photon index 3 \citep{Bassani93}. Followup observations
97: done with SIGMA and OSSE (which could only measure the combined flux
98: of 3C273 and GRS1227) showed that GRS1227 is probably variable by a
99: factor 2-5 on $\sim$1 day time scales \citep{Bassani93, Johnson95}.
100: No obvious soft X-ray counterpart was found in the SIGMA
101: error circle (5 arcmin, 1$\sigma$) with ROSAT (e.g. \cite{Leach96}),
102: although an Einstein source (1E 1227+0224) in the SIGMA error
103: circle (1$\sigma$) was identified with a QSO at z=0.57 that would
104: have to be of ``type 2'' (heavily self-absorbed) if it were indeed
105: the counterpart \citep{Grindlay93}.
106:
107:
108: \section{EXITE2 Observation and Data Analysis}
109:
110:
111: The second-generation Energetic X-ray Imaging Telescope Experiment
112: (EXITE2) was a phoswich (NaI(Tl)/CsI(Na)) imaging telescope
113: designed to detect and image cosmic X-ray sources in the broad hard
114: X-ray energy band (20 - 600 keV) from a high-altitude balloon. The
115: details of the payload are described in \cite{Lum94, Manandhar95,
116: Chou98, Chou01, Chou03}. EXITE2 flew three
117: times (from Ft. Sumner, New Mexico) in 1997, 2000 and 2001 for
118: scientific observations, which included the Crab Nebula, the X-ray
119: binaries Cyg X-1 and 4U 0614+09, the microquasar
120: GRS 1915+105, the Seyfert galaxy NGC 4151 and
121: QSO 3C273/GRS1227. The data analysis methods and
122: the observation results of the Crab and Cyg X-1 are reported in
123: separate papers \citep{Bloser02a, Chou03}. Other results
124: from the 2000 and 2001 flights, including separate experiments
125: for the development of imaging Cd-Zn-Te detectors (Bloser et
126: al 2002b, Jenkins et al 2003) are reported elsewhere.
127:
128: For the 1997 flight, during UT03:20 to UT05:32 on May 8 (local time 21:20
129: to 23:32 on May 7), the telescope was pointing to within 5 arcmin of
130: the QSO 3C273 at (night) flight altitude $\sim$115,000 feet. The data were
131: analyzed by the EXITE2 standard analysis methods described in
132: Chou (2001), Bloser et al. (2002a) and Chou et al. (2003).
133: The sky images showed that no significant
134: ($>$3$\sigma$) detection of 3C273 was obtained during the $\sim$2 hour
135: observation. However, a source $\sim$30' away from 3C273 was detected
136: in the EXITE2 energy band 5 (50-70 keV) and band 6
137: (70-93 keV) at 3.9$\sigma$ and
138: 4.1$\sigma$, respectively. We estimated aspect errors as \la 5' from
139: agreement of the Crab and Cyg X-1 images with their source positions.
140: The image for band 6 is shown as Figure~\ref{3c273img}.
141:
142: The source detected by EXITE2 in 1997 near 3C273 was located
143: at $\alpha$(2000)= 12$^h$31$^m$13$^s$, $\delta$(2000)= +2$^\circ$18',
144: with positional uncertainty $\pm \, \sim$7 arcmin
145: (1$\sigma$). Although
146: $\sim$18' away from the GRS1227 location reported by Jourdain et
147: al (1992) and plotted in
148: Figure~\ref{grspos}, the 2$\sigma$ error circles for SIGMA vs. EXITE2
149: overlap. The detected energy bands and the combined 5$\sigma$
150: significance of the EXITE2 detection are both comparable to the
151: original SIGMA detection, further supporting the reality of a
152: highly variable source at this location and distinct from 3C273.
153: The fluxes measured during the 2 hours observation were
154: (6.6$\pm$1.7)$\times$10$^{-4}$ photons cm$^{-2}$ s$^{-1}$ keV$^{-1}$ and
155: (3.5$\pm$0.85)$\times$10$^{-4}$ photons cm$^{-2}$ s$^{-1}$
156: keV$^{-1}$, for band 5 and 6 respectively, a factor
157: of $\sim$10 brighter than SIGMA 1990 observations.
158: The combined band 5-6 flux was (4.94$\pm$0.91)$\times$10$^{-4}$ photons
159: cm$^{-2}$ s$^{-1}$ keV$^{-1}$ ($\sim$260 mCrab). No source was
160: significantly detected in band 4 (37-50 keV) or band 7 (93-127 keV)
161: with 3$\sigma$ upper limits 1.8$\times$10$^{-3}$ photons cm$^{-2}$
162: s$^{-1}$ keV$^{-1}$ and 1.2$\times$10$^{-4}$ respectively.
163: We found that a spectrum of power law index $\sim$3 as reported by
164: Bassani et al. 1991 for GRS1227 is consistent with
165: the EXITE2 1997 observation results (see Figure~\ref{grsspec}).
166: Alternatively, the source could still
167: have a flatter spectrum (e.g. photon index $\sim$1.7 as for typical
168: AGN) with high energy cutoff at $\sim$100 keV.
169:
170: Additional observations of the field containing 3C273/GRS1227
171: were made in the EXITE2 flights
172: on September 19 2000 (UT20:45 to UT21:30) and May 23 2001 (UT03:09 to
173: UT06:17 on May 24). No source was detected during the $\sim$45 minute
174: observation in the 2000
175: flight. The 3$\sigma$ upper limit for the combined bands 5-6 (50 to
176: 93 keV) was 3.6$\times$10$^{-4}$ photons cm$^{-2}$ s$^{-1}$
177: keV$^{-1}$ ($\sim$190 mCrab), which is a factor 1.5 below the 1997
178: value. For the 2001 observation, the telescope was pointed to the
179: combined position of SIGMA/EXITE2, the intersection of
180: the 2$\sigma$ error circles
181: for EXITE2 and SIGMA
182: at RA=12$^h$30$^m$14$^s$, DEC= +2$^\circ$14'53'' (J2000),
183: as shown in Figure~\ref{grspos}.
184: No source was detected during a $\sim$3
185: hour observation. The 3$\sigma$ upper limit
186: for the combined bands 5-6 (50 to 93 keV) was 1.25$\times$10$^{-4}$
187: photons cm$^{-2}$ s$^{-1}$ keV$^{-1}$, or $\sim$65 mCrab, which
188: is comparable to the original SIGMA detection. The upper
189: limits for both EXITE2 observations are also plotted in Figure~\ref{grsspec}.
190: The combined EXITE2-SIGMA observations indicate that the source is
191: highly variable (by a factor of $\sim$10) at hard X-ray energies.
192:
193: \section{Discussion}
194:
195: The GRS1227 source presents a major puzzle. Both EXITE2 and
196: SIGMA detect a source at approximately the same location
197: (within 2$\sigma$) and with similar spectral properties which
198: are very distinct from 3C273. Although the EXITE2 detection is
199: marginal, as was the original SIGMA detection, the combined results
200: are more likely to be real. GRS1227 must be highly
201: self-absorbed and yet highly
202: variable. The SIGMA variations are a factor of $\sim$2 and
203: SIGMA-EXITE2 variation is nearly a factor of $\sim$20. This is a
204: combination not seen before, and suggests a new class of object: a
205: type 2 (i.e. heavily absorbed) Blazar, since Blazars (BL Lacs) are
206: usually the only AGN which show such
207: extreme variability. However, it might also be
208: a Narrow Line Seyfert 1 (NLSy) galaxy (e.g. RX J2217.9--5941, see
209: \cite{Grupe01}) but must still be highly self-absorbed. It may
210: be similar to the NLSys with significant internal absorption
211: and significant variability described by \cite{Fabian99}.
212: Alternatively it might be (an unprecedented) very highly absorbed
213: binary system, possibly a low luminosity magnetic CV undergoing
214: outburst, since intermediate polars at high inclination can also show
215: significant internal absorption.
216:
217: From the EXITE2 vs. SIGMA source error circles (2$\sigma$) shown in
218: Figure~\ref{grspos}, the combined source error box is $\sim$15'$\times$6',
219: centered on the EXITE2-SIGMA combined position RA=12$^h$30$^m$14$^s
220: (\pm$4'), DEC= +2$^\circ$14'53''($\pm$4') (J2000). At this
221: relatively uncrowded high latitude position, even a $\sim$2' position
222: could enable identification with an optical or radio variable.
223: However the need for a \la1'
224: position is obvious: if this is, as suspected, a highly absorbed
225: non-thermal source (BL Lac), it may not be at all conspicuous in the
226: optical and appear as a simple elliptical galaxy. The radio galaxies
227: numbered (2) and (4) in Figure~\ref{grspos}, or [SRK80]122803+023451
228: and 87GB[BWE91]1227+0230, are thus possible candidates. However, in
229: the event that the radio emission is confined to the core, GRS1227
230: may not even be bright at radio wavelengths since it is likely
231: Compton-thick. Given the space density of galaxies cataloged (SIMBAD
232: and NED) within the combined EXITE2/SIGMA error circles
233: (cf. Figure~\ref{grspos}), a \la1' source position is obviously needed
234: for a unique galaxy identification.
235:
236:
237: Source variability measurements can partly constrain the
238: underlying source nature. The generalized condition for
239: maximum variability of an isotropically emitting source powered
240: by accretion at rate \mdot and radiating luminosity L$_x$ is \\
241:
242: $\Delta$L$_x$ \la 2 $\times$ 10$^{41}$ $\eta_{0.1}$ $\Delta$t ergs/s, \\
243:
244: \ni
245: where $\eta_{0.1}$ = L$_x$/\mdot c$^2$ \about 0.1 is the
246: likely maximum efficiency factor
247: (cf. \cite{Guilbert83}). This can impose direct constraints
248: on the source luminosity (and thus distance and nature) for an
249: observed luminosity variation $\Delta$L$_x$ over time interval
250: $\Delta$t. Therefore, a factor of 2
251: increase on \about1d timescales (already suspected from SIGMA)
252: would imply the source is beamed and thus a Type 2 Blazar
253: if it were identified with
254: an AGN at z \ga 0.1 (like 3C273) with L$_x$ \ga10$^{45}$ erg/s.
255:
256: GRS1227 exhibits an unusual spectral shape. The simultaneous
257: detection by SIGMA (\ga40 keV) and upper limit at \la15 keV by
258: the ART-P X-ray imager on GRANAT (Bassani 1991)
259: indicated that the source is
260: highly self-absorbed (N${_H} \sim$10$^{24-25}$ cm$^{-2}$) and thus a
261: Compton-thick absorber (see \cite{Matt99} and
262: references therein). If the spectrum above $\sim$40-50 keV is indeed
263: a power law with photon index $\sim$3, it is consistent with a
264: Blazar or NLSy. Although NLSys are also highly variable,
265: these have not been reported
266: as Type 2 objects. As mentioned above, the EXITE2 spectral
267: shape (Figure 3) is equally consistent with a flatter spectrum with photon
268: index 1.7 but high energy cutoff at $\sim$100 keV, as expected for
269: a highly absorbed Seyfert 2 or Type 2 QSO. However the remarkable
270: variability required argues against the Seyfert or Type 2 QSO
271: interpretation and in favor of a Blazar, with highly variable
272: non-thermal emission. In this case, large amplitude optical
273: variability might also be expected, since \cite{Liller75} found a $\sim$5
274: magnitude brightening for the BL Lac PKS 1510-089 on the Harvard
275: plates. Given the offset from 3C273, it is possible that significant
276: optical variables could have been missed, and a search using
277: the Harvard plates is in progress.
278:
279: A magnetic CV interpretation is also possible even if the spectrum is
280: indeed a (moderately steep) power law: the magnetic CV AE Aqr can be
281: fit with a similar power law spectrum \citep{Eracleous96}, and
282: has been reported as a high energy gamma-ray source
283: (cf. Meintjes and de Jager 2000 and references therein). A magnetic CV
284: origin for GRS1227 might also explain the extreme flare variations:
285: the object could be a dwarf nova (like GK Per) undergoing outbursts.
286: Although the X-ray emission in classical dwarf novae outbursts is
287: usually soft, an outburst in a magnetic CV may contain significant
288: non-thermal emission. Historical optical measurements of the
289: could again provide an important test, since
290: the optical flux (from the disk) would be expected to
291: increase by $\sim$5 mag. If the EXITE2 1997 detection
292: was the peak of an outburst like that of GK Per, the expected X-ray
293: optical flux ratio F$_x$/F$_v$$\sim$10 would imply an outburst
294: magnitude V$\sim$10, and thus quiescence magnitude V$\sim$15-16, which
295: is probably too bright to have been missed as a blue object in the Palomar
296: Green survey (Green, Schmidt and Liebert 1986), which revealed
297: no such objects near 3C273.
298:
299: A high latitude black hole X-ray binary, an X-ray
300: nova (XN) like XTE J1118+480. which showed evidence for
301: magnetic flares and non-thermal optical emission (Merloni,
302: DiMatteo and Fabian 2000) is
303: conceivable. Here the power law hard X-ray
304: spectrum is natural; this is seen in
305: the very high state of BH novae (e.g. Nova Muscae, \cite{Esin97})
306: and is expected for the jet emission probable in a system like
307: XTE J1118+480. It
308: too would have to be highly self-absorbed, but -
309: as with the CV hypothesis - is plausible if the system were at
310: inclination $\sim$90$^\circ$, and self-absorbed at the required column
311: density of N$_H\sim$10$^{25}$ cm$^{-2}$. Such a highly absorbed XN
312: would be missed by the ASM on RXTE, and indeed no evidence for
313: an ASM source is found at the time of the 1997 EXITE2
314: observation (or the 2000 and 2001 EXITE2
315: observations; cf. Figure 4). However, an optical
316: outburst is again expected and should again be searched on historical
317: monitoring (e.g. plate) data. Thus the galactic
318: binary/transient scenarios might be distinguished from the
319: Type 2 BLazar or NLSeyfert interpretation by historical optical or
320: radio variations, which are probably larger in the binary
321: case given the relatively low luminosity secondary star expected.
322: The lack of any (large amplitude) optical variable
323: would support the Type 2 Blazar or NLSeyfert interpretation.
324:
325: Definitive confirmation of the source existence,
326: and of course a refined hard X-ray position and
327: spectrum, are obviously needed. Long after this paper
328: was initially submitted, we have carried out an
329: observation with INTEGRAL (IBIS telescope) to attempt to
330: further measure the source position, spectrum and variability.
331: Data have just been received and will be reported separately.
332: An initial INTEGRAL/IBIS observation of 3C273 has been
333: reported recently by Courvoisier et al (2003) which also
334: provided a sensitive new upper limit ($\sim$10 mCrab) for
335: emission from the position of GRS1227 -- a factor of $\sim$6
336: below the SIGMA detection and 2001 EXITE2 upper limit
337: and $\sim$25 below the 1997 EXITE2 detection. Thus the source
338: variability at \ga40 keV must be at least this factor.
339: Wide-field XMM
340: observations of the 3C273 field
341: should also be examined for evidence for a highly variable, highly
342: absorbed (\la 10 keV) X-ray source which would enable identification.
343:
344: \acknowledgments The authors thank T Gauron, J. Gomes, V. Kuosmanen,
345: F. Licata, G. Nystrom, A. Roy, R. Scovel (SAO) and J. Apple (MSFC) for
346: for the EXITE2 detector and gondola development, and NSBF personnel
347: for their excellent support of the balloon launches and flight
348: operations. This work was partially supported by NASA grants
349: NAGW-5103 and NAG5-5279. PFB is a National Research Council Research
350: Associate at NASA/Goddard Space Flight Center. TN acknowledges support
351: from the College of the Holy Cross.
352:
353:
354:
355:
356: \begin{thebibliography}{}
357:
358: \bibitem[Bassani et al.(1991)]{Bassani91} Bassani L. et al. 1999,
359: Proc. of 22nd Int. Cosmic Ray Conf., 1, 173
360: \bibitem[Bassani et al.(1993)]{Bassani93} Bassani L. et al. 1993,
361: \aap. 97, 89
362: \bibitem[Bloser et al.(2002a)]{Bloser02a} Bloser, P. F., Chou,
363: Y., Grindlay, J. E., Narita, T., \& Monnelly, G., 2002a, Astroparticle
364: Physics, 17, 393
365: \bibitem[Bloser et al.(2002b)]{Bloser02b} Bloser, P., Narita, T.,
366: Jenkins, J., Perrin, M., Murray, R. and Grindlay, J. 2002b,
367: Proc. SPIE, 4497, 88
368: \bibitem[Chou et al.(1998)]{Chou98} Chou, Y., et al. 1998, in
369: Conf. Record of the 1998 IEEE Nucl. Sci. Symposium (Piscataway, New
370: Jersey: IEEE), 210
371: \bibitem[Chou(2001)]{Chou01} Chou, Y., 2001, Ph.D thesis, Harvard
372: University
373: \bibitem[Chou et al.(2003)]{Chou03} Chou, Y., Bloser, P. F.,
374: Grindlay, J. E., Jenkins, J. A., \& Narita, T., 2003, submitted to ApJ
375: \bibitem[Courvoisier(2003)]{Courvoisier03} Courvoisier, T., Beckmann, V.,
376: Bourban, G., Chenevez, J., Chernyakova, M., Deluit, S., Favre, P.,
377: Grindlay, J. et al. 2003, \aap, in press (astro-ph/0308212)
378: \bibitem[Eracleous \& Horne(1996)]{Eracleous96} Eracleous,
379: M. and Horne, K. 1996, \apj, 471,427
380: \bibitem[Esin et al.(1999)]{Esin97} Esin, A. A., McClintock,
381: J. E., \& Narayan, R., 1997, \apj, 486, 865
382: \bibitem[Fabian \& Iwasawa(1999)]{Fabian99} Fabian,
383: A. C. and Iwasawa, K., 1999, \mnras, 303, L34
384: \bibitem[Green(1986)]{Green86} Green, R. F., Schmidt, M., and
385: Liebert, J., 1986, Ap.J. Suppl., 61, 305
386: \bibitem[Grindlay(1993)]{Grindlay93} Grindlay, J. E. 1993,
387: A\&AS, 97, 113
388: \bibitem[Grupe et al.(2001)]{Grupe01} Grupe, D., Thomas,
389: H. -C. and Leighly, K. M., 2001, \aap, 369, 450
390: \bibitem[Guilbert et al.(1983)]{Guilbert83} Guilbert, P., Fabian,
391: A., \& Rees, M. 1983, \mnras, 205, 593
392: \bibitem[Johnson et al.(1995)]{Johnson95} Johnson, W. N. et
393: al. 1995, \apj, 445, 182
394: \bibitem[Jourdain et al.(1992)]{Jourdain92} Jourdain, E. et
395: al. 1992, \apj, 395, L69
396: \bibitem[Leach \& McHardy(1996)]{Leach96} Leach,
397: C. M. \& McHardy I. M., 1996, \mnras, 278, 465
398: \bibitem[Liller \& Liller(1975)]{Liller75} Liller ,
399: M. H. and Liller, W., 1975, \apj, 199, L133
400: \bibitem[Lum et al(1994)]{Lum94} Lum, K. S. K. et al. 1994,
401: IEEE Trans. on Nucl. Sci., NS-41, 1354
402: \bibitem[Manandhar(1995)]{Manandhar95} Manandhar, R. P. 1995,
403: Ph.D. Dissertation, Harvard University
404: \bibitem[Matt et al.(1999)]{Matt99} Matt, G., Pompilio, F. \& La
405: Franca, F., 1999 New Astro. 4, 91
406: \bibitem[Meintjes(2000)]{Meintjes00} Meintjes, P.J. and de Jager, O.C.
407: 2000, \mnras, 311, 611
408: \bibitem[Merloni(2000)]{Merloni00} Merloni, A., DiMatteo, T. and
409: Fabian, A. 2000, \mnras, 318, L15
410:
411: \end{thebibliography}
412:
413: \clearpage
414:
415:
416:
417:
418: \begin{figure}
419: \plotone{f1.eps}
420: \caption{Image of GRS1227+025 obtained with EXITE2 on UTMay 8.17, 1997,
421: in the 70 - 93 keV band. Contours begin at 2.5$\sigma$ and
422: increase in steps of 0.5$\sigma$ to $\sim4\sigma$.
423: The source is independently detected at $\sim4\sigma$
424: in the 50-73 keV band. The expected position of 3C273 is
425: shown by the cross; the source detected is at a position (J2000)
426: compatible with the SIGMA detection of GRS1227+025. \label{3c273img}}
427: \end{figure}
428:
429: \clearpage
430:
431:
432:
433: \begin{figure}
434: \plotone{f2.eps}
435: \caption{GRS1227+025 source positions (plotted on J2000 grid) measured
436: (2$\sigma$ radius error circles) by
437: SIGMA ($\sim$10') and by EXITE2 ($\sim$14').
438: The positions of
439: 3C273 and 1E 1227+0224 are also shown. The cross marked inside the
440: intersection of the 2$\sigma$ error circles is combined position of
441: EXITE2 and SIGMA. The numbered source positions
442: (X) are galaxies (1,3,5,6) and radio galaxies (2,4) found in a
443: SIMBAD and NED search of the intersecting error circle for
444: EXITE2/SIGMA.
445: \label{grspos}}
446: \end{figure}
447:
448: \clearpage
449:
450: \begin{figure}
451: \plotone{f3.eps}
452: \caption{Spectrum measured by the 1997 EXITE2
453: observation (triangle data points and dotted curve), and the upper
454: limit from the 2000
455: observation (asterisk data point) and 2001 observation (``X'' data
456: point) vs. that measured in the SIGMA/ART-P observations
457: (square and diamond points; from Jourdain et al. (1992) and dashed
458: curve, from Bassani et al. (1991)). Each of the three curves is for a
459: power law with photon index 3 and hydrogen column density
460: N$_H$=1.5$\times$10$^{25}$cm$^{-2}$ to fit the SIGMA/ART-P observation
461: results. \label{grsspec}}
462: \end{figure}
463:
464: \clearpage
465: \begin{figure}
466: \plotone{f4.eps}
467: \caption{RXTE/ASM light curve for 3C273 showing times of EXITE2
468: observations of the field and 1997 detection vs. 2000 and 2001
469: upper limits of GRS1227+025. The increased ASM count rate due
470: to solar background is evident for the times (marked) when the
471: Sun is closest ($\sim$4\deg) to 3C273 or GRS1227+025.}
472: \end{figure}
473:
474: \end{document}
475:
476:
477:
478:
479:
480: