1: %\documentstyle[manuscript]{aastex}
2: \documentclass[12pt,preprint]{aastex}
3: \def\lsim{\lower.5ex\hbox{$\; \buildrel < \over \sim \;$}}
4: \def\gsim{\lower.5ex\hbox{$\; \buildrel > \over \sim \;$}}
5: %\received{2003 February 6}
6:
7: \begin{document}
8:
9: \title{Possible Photometric Evidence of Ejection of Bullet Like Features in the Relativistic Jet source SS433}
10:
11: \author{Sandip K.\ Chakrabarti$^{1,2}$, S. Pal$^{2}$, A. Nandi$^1$, B.G. Anandarao$^3$, Soumen Mondal$^3$}
12:
13: \affil{$^1$S.N. Bose National Center for Basic Sciences, JD-Block, Salt Lake, Kolkata, 700098\\ India\\
14: $^2$ Centre for Space Physics, P-61 Southend Gardens, Kolkata, 700084, India\\
15: $^3$ Physical Research Laboratory, Navarangapura, Ahmedabad, 380009, India\\
16: e-mail: chakraba@bose.res.in, space\_phys@vsnl.com, anuj@bose.res.in, anand@prl.ernet.in, soumen@prl.ernet.in}
17:
18: \begin{abstract}
19: SS433 is well-known for its precessing twin jets having
20: optical bullets inferred through {\it spectroscopic} observation of $H_\alpha$
21: lines. Recently, Chakrabarti et al. (2002) described processes which may be
22: operating in accretion disk of SS433 to produce these bullets. In a recent multi-wavelength campaign,
23: we find sharp rise in intensity in time-scales of few minutes in X-rays, IR and radio waves
24: through {\it photometric } studies. We interpret them to be
25: possible evidence of ejection of bullet-like features from accretion disks.
26: \end{abstract}
27:
28: \keywords {SS433 --- X-ray, Infra-red and radio sources --- stars: individual (SS 433) ---
29: stars: winds, outflows, mass loss }
30:
31: \noindent ASTROPHYS. J. LETTERS, 595, L45, 2003
32:
33: \section{Introduction}
34:
35: SS433 is a well studied bright emission line compact system which is ejecting matter
36: in symmetrically opposite directions at a speed of $v_{jet} \sim 0.26$c.
37: It has a mass-losing (${\dot M} \sim 10^{-4} M_\odot/$yr) companion
38: orbiting in $13.1$d. The jets are precessing in $162.15$d around the symmetry axis. The velocity is
39: remarkably constant to within a percent or so (Margon, 1984; Gies et al. 2002). As a result of the remarkable
40: constancy in the precession time scales and the velocity, the instantaneous locations of the red and blue-shifted $H_\alpha$
41: lines are well predicted by the so-called `kinematic-model' (Abell \& Margon, 1979) and the compilation of
42: twenty years of timing properties (Eikenberry et al. 2001) suggest that kinematic model can explain the general
43: variation of the red and blue-shifts very well.
44:
45: A very exciting observation, made within years of the discovery of SS433, suggests that the
46: passage of the jets is not continuous, but as if through successive and discrete bullet-like entities,
47: at least in the optical region (Grandi, 1981; Brown et al. 1991). The $H_\alpha$ lines were seen to brighten up
48: and fade away without changing their red/blue-shifts, indicating the brightened bullets
49: are radially ejected and do not have any rotational velocity component. Since the bullets
50: of energy $\sim 10^{35}$ ergs do not change their speed for a considerable time
51: ($\sim 1-2 $ days), Chakrabarti et al (2002) postulated that they must be ejected from accretion
52: disk itself. They presented a mechanism to produce quasi-regular bullets.
53: Using results of numerical simulations involving oscillation of shocks in accretion disks,
54: they concluded that in the normal circumstances, $50-1000$s interval is
55: expected in between the bullet ejection. These bullets would be ejected from X-ray emitting
56: region and propagate through optical, infra-red ($\sim 10^{13-14}$cm) and finally
57: to radio emitting region at $\gsim 10^{15}$cm (roughly the distance
58: covered in a day with $v\sim v_{jet}$) or so. Thus
59: if the object is in a low or quiescence state, each individual bullet flaring and dying away in
60: a few minutes time scale, should be observable not only in optical wavelengths
61: (Grandi, 1981; Margon, 1984; Brown et al. 1991; Gies et al. 2002) but also in all the
62: wavelengths, including X-ray, IR and radio emitting regions.
63: So far, no such observations of individual bullets has been reported in the literature
64: and it was necessary to make a multi-wavelength observation at relatively quieter states.
65:
66: In this {\it Letter}, we present some results of our multi-wavelength studies.
67: From the long term analysis of radio flares (Bonsignori-Facondi, Padrielli, Montebugnoli
68: and Barbieri, 1986; Vermeulen et al. 1993) it is known that in between
69: big flares which occupy $\sim 20\%$ of the time the object may go to
70: very quiescence state. So, it is likely that one could `catch' these bullets
71: in action provided observations are carried with very short time resolution. Our multi-wavelength
72: observations lasted during 25th of Sept., 2002 to 6th of October, 2002 with
73: X-ray, infra-red, optical, radio observations made simultaneously on the 27th of September, 2002.
74: Here, we report only X-ray, infra-red and radio observations of 27th and 29th of September, 2002.
75: Optical studies required longer integration times and these results
76: along with other days of observations would be reported elsewhere (Chakrabarti et al. 2003).
77:
78: Our main results indicate that there are considerable variations in the timescale of minutes
79: in all the wavelengths. These may be called micro-flares. When Fourier transform is made,
80: some excess power is observed in $2-8$ minutes time scale (often beyond 3$\sigma$ level).
81: The X-ray count rate was found to increase by $15-20$\% within a minute. Since the emitting
82: regions of X-ray, IR and radio are not well known with absolute certainly, while duration of the
83: flares last less than a few minutes, we could not prove beyond doubt that
84: there are indeed correlations among the micro-flares observed
85: in these wavelengths. However variabilities we find are not flicker type or shot noise type in the sense
86: that the power density spectrum (PDS) is not of $1/f$ type and the duration is not very short (i.e., $<1$s).
87: We therefore believe that we may have found evidences of bullet ejection through these observations
88: in other wavelengths.
89:
90: \section{Observations and Data Reduction}
91:
92: Radio observation was carried out with Giant Meter Radio Telescope (GMRT) at 1280MHz (bandwidth $16$ MHz)
93: which has $30$ antennas each of $45$m in diameter spreaded over $25$km
94: region (Swarup et al. 1991) near Pune, India along roughly $Y$
95: shaped array. The data is binned at every $16$ seconds. On 27th and 29th of
96: September, 2002, no. of antennas working were 28 and 13 respectively.
97: AIPS package was used to reduce the data.
98: Bad data was flagged using tasks UVFLG and TVFLG and the standard deviation
99: at each time bin using UVPLT package was computed. On 27th, 3C48 and
100: 3C286 were used as the flux calibrator while on the 29th only 3C48 was used.
101: Generally, the observation condition was very stable.
102:
103: Infrared observation was made using Physical Research Laboratory (PRL) 1.2m Mt.
104: Abu infrared telescope equipped with Near-Infrared Camera and Spectrograph
105: (NICMOS) having 256 $\times$ 256 HgCdTe detector array cooled to $77$K. The
106: filters used were standard J ($\lambda$=1.25 $\mu$m, $\Delta\lambda$= 0.30
107: $\mu$m), H ($\lambda$=1.65 $\mu$m, $\Delta\lambda$= 0.29 $\mu$m) and K$^\prime$
108: ($\lambda$=2.12 $\mu$m, $\Delta\lambda$= 0.36 $\mu$m) bands. The observational
109: data on 27th of September, 2002 for J and H bands are binned at every 10
110: seconds while that of K$^\prime$ band is binned at every 20 seconds. The
111: data reduction was performed using the IRAF software package.
112: All the object frames were de-biased, sky-subtracted and
113: flat-fielded using normalized dome flats. The sky frame was created by usual
114: practice of median combining of five position dithered images in which the
115: source was within the NICMOS field of 2$^\prime$ $\times$ 2$^\prime$. At each
116: dithered position ten frames were taken with each integration time of $10$
117: seconds. The nearby infrared bright standard star GL748 (Elias et al. 1982)
118: was used as the calibrator and it was observed for at least 15 minutes at each
119: filter band. We measured the stellar magnitudes using the aperture photometry
120: task (APPHOT) in IRAF. Our derived mean JHK$^\prime$ magnitudes on Sept. 27th
121: are 9.47$\pm$0.02, 8.48$\pm$0.02 and 8.32$\pm$0.02
122: respectively and the corresponding mean flux densities are 0.261$\pm$0.002,
123: 0.413$\pm$0.003 and 0.305$\pm$0.003 respectively. The magnitudes are converted
124: to flux density (Jansky) using the zero-magnitude flux scale of Bessell,
125: Castelli \& Plez (1998). To estimate reddening, we assumed visual extinction
126: A$_v$ =8.0 (Gies et al. 2002). Using the relation given in Bessell, Castelli
127: \& Plez (1998) the JHK$^\prime$ extinctions were found to be A$_J$= 2.32, A$_H$
128: = 1.84 , A$_K$ = 0.88. The de-reddened flux in JHK$^\prime$ are 2.24 Jy, 2.21
129: Jy and 1.67 Jy respectively. The differential magnitudes are determined using
130: two brightest stars (std1: J= 12.1, H=10.6, std2: J=12.5, H= 11.1 mag) in
131: the same frame of the object. The error in individual flux density measurement is usual propagation
132: error of the observed photometric magnitude. Photometric errors $\epsilon$
133: are calculated for individual frame of every star and for the subtracted
134: differential magnitude the final error was calculated as
135: $\sqrt{\epsilon_1^2 + \epsilon_2^2}$, where $\epsilon_1$ and $\epsilon_2$ are the
136: error-bars of the individual stars.
137:
138: X-ray observation was carried out using Proportional Counter Array (PCA)
139: aboard RXTE satellite. The data reduction and analysis was performed
140: using software (LHEASOFT) FTOOLS 5.1 and XSPEC 11.1. We extract
141: light curves from the XTE/PCA Science Data of GoodXenon mode. We combine
142: the two event analyzers (EAs) of 2s readout time to reduce the Good Xenon data
143: using the perl script {\bf make\_se}. Once {\bf make\_se} script was run on the
144: Good\_Xenon\_1 and Good\_Xenon\_2 pairs, the resulting file was reduced
145: as Event files using {\bf seextrct} script to extract light curves. Good time
146: intervals were selected to exclude the occultations by the earth and South
147: Atlantic Anomaly (SAA) passage and also to ensure the stable pointing.
148: We also extracted energy spectra from PCA {\bf Standard2} data in the energy range
149: $2.5$ - $20.0$ keV (out of five PCUs only data from 0, 2, 3 PCUs are
150: added together). For each spectrum, we subtracted the background data that
151: are generated using PCABACKEST v4.0. PCA detector response matrices are
152: created using PCARSP v7.10.
153:
154: \section{Results on short time-scale variabilities}
155:
156: The observational result of September 27th, 2002 is shown in Fig. 1
157: with UT (Day) along the X-axis.
158: The upper and middle panels show the radio and IR fluxes (uncorrected for reddening) in Jansky
159: and the lower panel shows X-ray counts per second in $2-20$keV.
160: Typical error-bars of the mean-flux measurements
161: (standard deviation in each time bin for radio and IR, and squared-root of counts
162: per binsize for X-rays) which are included in the Figure are: in radio
163: $\sim 1$mJy, in IR $0.5$mJy and in X-ray $\sim 3$ counts/s.
164: These observations correspond to an average flux of $10^{-14}$ergs/cm$^2$/s,
165: $5\times 10^{-10}$ergs/cm$^2$/s and $ 10^{-10}$ergs/cm$^2$/s respectively. In other words,
166: assuming isotropic emission, at a distance of $3$kpc for the source, the average radio, IR, and X-ray
167: luminosities are $1.1\times 10^{30}$ ergs/s, $5.5 \times 10^{34}$ergs/s and $10^{35}$ergs/s
168: respectively. Observations in radio and IR were carried out during
169: 25-30th September, 2002 and no signature of any persistent `flare' was observed.
170: The radio data clearly showed a tendency to go down from $1.0$Jy to $0.7$Jy reaching
171: at about $0.3$Jy on 28th/29th, while the X-ray data showed a
172: tendency to rise towards the end of the observation of the 27th.
173: The IR data in each band remained virtually constant. The H-band result was found to be higher
174: compared to the J and K$^\prime$ band results during 25th-29th September, 2002. A similar result
175: of turn around at about 4 micron was reported earlier by Fuchs (2003). This turn over could be
176: possibly due to free-free emission in optically thin limit.
177: Detailed discussion will be presented elsewhere (Chakrabarti et al. 2003).
178:
179: In Fig. 2, we present the same light curves as in Fig. 1 but plotted around the `local' mean,
180: i.e., mean values taken in each `spell' of observation.
181: We note that there are significant variations in a matter of minutes in observations
182: at all the wavelengths. From eye-estimate, we see variability time-scale to be
183: $T_{var} \sim 2-8$ minutes. The error-bars include errors in individual measurements plus the
184: standard deviation of the flux variation in the light curve.
185: To impress that the variability is real, we show in Fig. 3 the
186: differential flux density variation of IR observations in the J and H
187: bands during 27 September 2002 using differential photometry. The error-bars are also
188: shown. The 1 $\sigma$ error-bar (J=0.00035 Jy, H= 0.00085 Jy) of the differential flux
189: variation between SS433 and std1 for the whole light curve is a factor of
190: $3.5$ and $2.5$ in the J-band and H-band respectively in comparison to that
191: between two standards (J=0.0001 Jy, H=0.00035 Jy). The 1 $\sigma$ for the
192: light curve is more than a factor of 5$\sigma$ of single point measurement
193: error. Thus, the variation in the IR light curves of SS433 is likely to be intrinsic
194: and the analysis shows above 2$\sigma$ level variability in both bands.
195:
196: Could these variations be due to individual bullets? In order to be specific, we present in Fig. 4a,
197: one `micro-flare'-like event in radio from the data on 29th of Sept., 2002, when radio
198: intensity was further down $\sim 0.3$Jy so that the micro-flares could be prominently seen.
199: Here $0$s corresponds to $15$h$35$m UT. We observe brightening the source
200: from $0.35$Jy to $0.8$Jy in $\sim 75$s which faded away in another $\sim 75$s. That
201: is, the intensity became more than doubled in $\sim 1$ minute! Similarly in Fig. 4b,
202: where we presented a `micro-flare' from the 2nd (central) `spell' of X-ray data of 27th Sept. 2002
203: (Fig. 1-2), we also observe significant brightening and fading in $\sim 100$s.
204: Here $0$s corresponds to $16$h$5$m UT. The count rate went up more than $15\%$
205: or so in about a minute. The energy contained in the radio micro-flare, integrated
206: over their lifetime is about $I \nu \tau 4\pi D^2 10^{-23} = 1.1 \times 10^{33}$ergs
207: (Here, $I \sim 0.8$ is the intensity in Jansky, $\nu=1.28\times 10^9$Hz is the frequency
208: of observation, $\tau\sim 100s$ is the rise-time of the bullet, $D=9\times 10^{21}cm$
209: is the distance of SS433). Similarly, the energy contained in the X-ray micro-flare is
210: about $\frac{1}{2} \tau (N_{\gamma, max}-{\bar N_\gamma}) E_\gamma 4\pi D^2/A_{PCA}=
211: 2.7 \times 10^{35}$ergs (Here, $\tau \sim 100s$ is the rise-time of the flare,
212: $N_{\gamma, max}$ is the maximum photon count rate, ${\bar N}_\gamma$ is the
213: average photon number, $E_\gamma$ is the average photon energy, $A_{PCA}$ is the
214: area of the PCA detectors.). The spectroscopic study
215: yields an average flux of $2.41 \times 10^{-10}$ ergs/cm$^2$/s. With an estimated
216: duration of $100$s, and about $15\%$ energy going to the micro-flare (Fig. 4b), one obtains
217: the micro-flare energy to be $4.1 \times 10^{35}$ ergs in general agreement with
218: the result obtained from photometric study. Since the radio luminosity is very small,
219: even when integrated over $0.1$ to $10$GHz radio band (with a spectral index of
220: $\sim -0.5$) (Vermeulen et al, 1993) we find that almost all the injected energy
221: at X-ray band is lost on the way during its passage of $\sim 1-2$d.
222:
223: Though the variations we find are not periodic (strictly speaking they are
224: nor expected to be periodic, either), the power-density spectrum (PDS) does show
225: considerable power in frequencies $\sim 0.002-0.008$Hz.
226: Deviation of the PDS from a power-law background
227: $\propto \nu^{-\alpha}$ (e.g., Mineshige, Ouchi \& Nishimori, 1994) in all three
228: bands gives an estimate of excess power at low-frequencies.
229: We fit $\alpha=1.8$ for X-ray power, $1.9$ for IR power, and $1.6$
230: for radio power in PDS. X-ray power shows excess at $\sim 0.003$Hz
231: ($> 2.7 \sigma$), i.e., at $T_{var, x} \sim 5.5$ min. and
232: at $0.0077$Hz ($ >1.4\sigma$), i.e., at $T_{var, x} \sim 2.1$ min. IR power shows excess
233: at $0.0022$Hz ($>4\sigma$), i.e., $T_{var, ir} \sim 7.7$ min.
234: Radio PDS shows excess at $\sim 0.0023$Hz ($>3.2 \sigma$)
235: i.e., at $T_{var, r} \sim 7.2$ min. and at $\sim 0.003$Hz ($>1.6\sigma$),
236: i.e. at $T_{var,r}\sim 5.5$ min. respectively. Here 1$\sigma$ error in residual power
237: is the standard deviation computed separately for each PDS after
238: subtracting the power-law background $\nu^{-\alpha}$. Because the peaks in PDS are often marginally
239: significant we do not claim that we see quasi-periodic oscillations that are
240: observed in numerous black hole and neutron star candidates.
241:
242: In order to establish that the features we observe are really due to `bullets'
243: emitting at different wave bands, one should find correlations among them, or try to `follow' them
244: from one band to the other. Unfortunately, cross-correlation among our observations
245: did not yield sharp peaks, partly because the observations were of short duration.
246: Main problem is that the locations of the IR/radio emitting regions
247: themselves are very uncertain. Also, the average duration of an `event' ($\sim$ minute)
248: and average interval of the events ($2-8$ minutes) are very very short compared to
249: the travel time of the bullets to IR ($\sim 10^{4}$s) or radio ($\sim 10^5$s)
250: regions. However, we can exclude that the variabilities to be due to `fluctuations' at
251: the inner regions of the accretion disk -- the typical time-scale of such variabilities
252: (say, at $r\sim 3r_g$, where, $r_g=2GM/c^2$, $M$, $G$ and $c$ being the mass of the
253: black hole, $G$ being the gravitational constant and $c$ being the velocity of light)
254: of an $M=10M_\sun$ object would be $\sim 2\pi r/c \sim 20GM/c^3 \sim 10^{-3}$s, i.e., of
255: much shorter duration than what we see. Similarly, has it been due to
256: random or flicker noise, we should have seen $1/f^{\alpha}$ ($\alpha \sim 1$)
257: dependence of the PDS. However, the best fit of PDS has $\alpha \sim 1.6-1.8$ instead. Thus,
258: the origin of these features must be different and could be due to bullet-like ejections from the disk.
259:
260: In the spectrum, we find two strong Fe line features in all the three spells of X-ray observation.
261: The best fit was found to be the thermal bremsstrahlung model with two Fe lines
262: ($kT \sim 18$keV) having a reduced $\chi^2$ of around $1.2$ in each case.
263: We failed to fit with a model having a blackbody emission component. Thus, no evidence for
264: a Keplerian disk was found. The average flux was found to be $2.3 \times 10^{-10}$ ergs/cm$^2$/s.
265: This corresponds to a luminosity of $2.5 \times 10^{36}$ ergs/s. Since a bullet
266: has about $10-15\%$ of the total count (Fig. 4b), each bullet will have an energy
267: of around $2.5 \times 10^{35}$ ergs/s.
268:
269: \section{Concluding remarks}
270:
271: In this {\it Letter}, we presented results of our multi-wavelength observations which
272: were save at short time intervals. From the analysis of the observations
273: of radio, IR and X-ray in the quiescence state we conclude that we may be observing
274: ejection events of bullet-like features from the accretion disk in time scales of $\sim 2-8$ minutes.
275: Identification of small micro-flare events
276: with those those of bullet ejection is derived from the time scales of
277: variabilities, which are roughly the same in all these wavelengths.
278: We find their presence in X-ray ($\lsim 10^{11-12}$cm), IR ($\lsim 10^{13-14}$cm)
279: and radio ($\lsim 10^{15}$cm) emission regions. Vermeulen et al. (1993) found
280: evidence for optical bullets with life-time of $1-2$d. This is perhaps due to
281: the propagation of a burst of indistinguishable bullets and not due to
282: a single one. We identified micro-flare like features
283: in all these observations which may be signatures of the
284: bullets. Count rate of X-ray was seen to increase $15-20$\% in a matter of a minute.
285: One way to actually identify each bullet could have been to follow them from X-ray region
286: outwards. This will require very careful time delay measurements since the
287: distances of emission regions are not very accurately known to follow a
288: feature of duration of a minute. We exclude the
289: possibility that what we see were flicker noise since neither the duration
290: nor the PDS properties match with those of flicker noise. One could have
291: perhaps distinguished the energetic bullets by observing polarization
292: properties of the radio-emissions during the short-lived flares, but
293: unfortunately due to technical reason this observation could
294: not be carried out. We plan to do such an observation in near future.
295:
296: SKC thanks Dr. J. Swank of NASA/GSFC and Prof. A. P. Rao of GMRT/NCRA for giving
297: time in RXTE and GMRT for observations. He also thanks Mr. J. Kodilkar
298: of GMRT/NCRA for his assistance in analysing the radio data. This work is supported
299: in part by CSIR fellowship (SP) and a DST project (SKC and AN). Authors thank the
300: unknown referee for suggestions leading to considerable improvement of the observational aspects.
301:
302: \begin{thebibliography}{}
303:
304: \bibitem[]{201} Abell, G.\ O., \& Margon, B. 1979, Nature, 279, 701
305: \bibitem[]{202} Bessell, M.S., Castelli, F. \& Plez, B., 1998, A\&A, 333, 231
306: \bibitem[]{203} Bonsignori-Facondi, S.R., Padrielli, Montebugnoli, S. \& Barbieri, R., 1986, A\&A 166, 157
307: \bibitem[]{204} Brown, J.C., Cassinelli, J.P. \& Collins II, G.W 1991, ApJ, 378, 307
308: \bibitem[]{205} Chakrabarti, S.K., Goldoni, P., Wiita, P.J., Nandi, A. \& S. Das, 2002, ApJ, 576, L45
309: \bibitem[]{206} Chakrabarti, S.K. et al. 2003 (in preparation)
310: \bibitem[]{207} Eikenberry, S.S. et al. 2001, ApJ. 561, 1027
311: \bibitem[]{208} Elias, J.H., Frogel, J.A., Matthews, K. \& Neugebauer, G. 1982, AJ, 87, 1029
312: \bibitem[]{209} Fuchs, Y. in New Views on Microquasars, P. Durouchaux, Y. Fuchs \& J. Rodriguez (Eds.), 269 (CSP:Kolkata) astro-ph/0208432
313: \bibitem[]{210} Gies D.\ R., McSwain M.\ V., Riddle, R.\ L., Wang, Z., Wiita, P.\ J., \& Wingert,
314: D.\ W. 2002, ApJ, 566, 1069
315: \bibitem[]{211} Grandi, S. 1981, Vistas Astr. 25, 7
316: \bibitem[]{212} Margon, B. 1984, ARA\&A, 22, 507
317: \bibitem[]{213} Mineshige, S., Ouchi, N. B. \& Nishimori, H., 1994, PASJ, 46, 97
318: \bibitem[]{214} Vermeulen, R.\ C., et al. 1993, A\&A, 270, 189
319: \bibitem[]{215} Swarup, G. et al. 1991, Curr. Sci. 60, 95
320:
321: \end{thebibliography}{}
322:
323:
324: \vfil\eject
325: \begin{figure}
326: \plotone{f1.eps}
327: \end{figure}
328:
329: \vfil\eject
330: \begin{figure}
331: \plotone{f2.eps}
332: \end{figure}
333:
334: \vfil\eject
335: \begin{figure}
336: \plotone{f3.eps}
337: \end{figure}
338:
339: \vfil\eject
340: \begin{figure}
341: \plotone{f4a.eps}
342: \end{figure}
343:
344: \vfil\eject
345: \begin{figure}
346: \plotone{f4b.eps}
347: \end{figure}
348:
349: \newpage
350:
351: \noindent Fig. 1: Multi wavelength observation of short-time variability in SS433 by Radio (upper panel),
352: Infra-red (middle panel) and X-ray (lower panel) on 27th of September, 2002. The observations
353: were made at Giant Meter Radio Telescope, Pune at 1.28GHz (Radio), 1.2m Mt. Abu Infra-red
354: Telescope at J, H and K$^\prime$ bands and RXTE satellite (2-20keV) respectively.
355:
356: \noindent Fig. 2: Observations at in Fig. 1 are plotted around mean taken in each spell
357: of observation. Considerable variations at time scale of a few minutes are observed.
358:
359: \noindent Fig. 3: Differential photometry of SS433 with respect to two brightest standard stars
360: (std1 and std2) in the same frame of the object are plotted. Different curves
361: are marked on upper-right corner in each panel. X-axis of the graph is the
362: relative time of measurements in seconds. The error bar for each individual
363: differential measurements are also shown. Differential flux variation of SS433
364: is above 2$\sigma$ level in comparison that of standards.
365:
366: \noindent Fig. 4: Individual flares in very short timescales are caught.
367: (a) A radio flare lasting $2.5$ minutes (observed on 29th Sept. 2002)
368: and (b) an X-ray flare (observed on 27th of Sept. 2002) lasting for about $3.5$ minutes. Each bin-size is $16$s.
369:
370: \end{document}
371:
372:
373:
374:
375: