Astron. Astrophys. 319, 855-862 (1997)

3. The outburst ephemeris

Table 4 lists the known outbursts from 4U 1630-47 and is an updated version of that in P95. The derivation of the outburst times, except those for cycles 11, 12, 15, and 16 are reported in P95. Outburst 16 was observed by both the All Sky Monitor (ASM) and Proportional Counter Array instruments on the Rossi X-ray Timing Explorer (XTE). The ASM intensity of 4U 1630-47 increased from 20 mCrab to about 200 mCrab on around 1996 March 20 and remained steady at this level until at least 1996 May 1 (Levine et al. 1996; Marshall 1996). We therefore assume an outburst start time of 1996 March 20 with an uncertainty of days. Since the observations presented here only provide brief snapshots of the outbursts, it is necessary to estimate the outburst phases at which they occurred so that the times of maxima can be derived.

Table 4. 4U 1630-47 outburst observations

The TTM instrument on the Mir Space Station observed 4U 1630-47 in outburst on 1989 March 24-25 (in 't Zand 1992). Using the best-fitting thermal bremsstrahlung spectrum of in 't Zand (1992) of kT=  keV and absorption of  H atoms cm^-2 gives a 2-28 keV intensity of  erg cm^-2  s^-1. This is 30% higher than the intensity observed by Ginga 22 days earlier of  erg cm^-2  s^-1 and suggests that the Ginga observation may have occurred during the rising phase of the outburst. We therefore assume that outburst maximum occurred mid-way between the TTM and Ginga observations. Since the other outbursts have a typical e -folding time of 50 days, we have made a conservative estimate for the uncertainty in outburst maximum of 25 days (Table 4).

If the excess emission observed by Ginga during the 1987 October observation originates from 4U 1630-47, we can use the known outburst properties to estimate the time of outburst maximum. Since the luminosity and power-law spectral shape are similar to those during the final EXOSAT observation reported in P86 which occurred 120 days after the start of the 1984 outburst, we have made a conservative estimate for the 1987 outburst start time by subtracting 120 days from the time of observation. We assign an uncertainty of 50 days to the outburst start time derived in this way (Table 4).

The 1-50 keV luminosity observed by ASCA in 1994 September is a factor 4 less than observed during the second EXOSAT observation of 4U 1630-47 reported in P86. However, as discussed in Sect.  2.2, the spectral shapes are similar. This suggests that the ASCA observation took place at a similar, or later, outburst phase than the second EXOSAT observation which occurred 40 days after the start of the 1984 outburst (P95). The difference in luminosities and an e -folding time of 50 days suggests that the ASCA observation may have occurred up to 70 days later in the outburst than the second EXOSAT observation. We therefore assume that the ASCA observation took place midway between these two extremes and subtract 75 days from the observation time to derive the outburst start time (Table 4). We assign an uncertainty of  days to the outburst start time derived in this way.

Inspection of the outburst times derived above reveals that the Ginga outburst times are in agreement with the ephemeris given in P95, but that the ASCA and XTE outburst times are 100-150 days late. Similar behavior has been noted before. The 1977 outburst occurred 50-70 days late (Kaluzienski & Holt 1977), and may have lasted for up to six months (Kaluzienski et al. 1978; Sims & Watson 1978). The subsequent outburst was observed by Einstein and appeared to start 120 days early (Table 4). This implies, that at times, 4U 1630-47 can undergo more complex outburst behavior. A linear fit to all the outburst times listed in Table 4 gives an unacceptable fit with a of 85 for 11 dof. The mean outburst recurrence interval is 610.0 day, and the outburst epoch of JD 2,440,368.5. These values are inconsistent with the ephemeris reported in P95. The average deviation of the outburst times from this linear relation is  day (Fig. 3). If the last two outbursts are ignored then the ephemeris is consistent with that of P95 which is shown as a dotted line in Fig. 3. This sort of behavior is reminiscent of the superoutbursts observed from SU UMa stars. In these systems, the superoutbursts follow a linear ephemeris for 10-20 cycles, with a standard deviation of only 5-10% of the corresponding period. Occasionally, the mean cycle length switches to one of another 2-3 period values which are characteristic of each star (e.g. Vogt 1980). Observations of subsequent outbursts from 4U 1630-47 will reveal whether the recent anomalous behavior continues, or whether the timing reverts to that predicted by the ephemeris in P95.

Fig. 3. The time residual in days after a linear fit of the outburst times to cycle number. The error bars reflect the uncertainties in deriving the outburst start times (see text). The dotted line shows the best-fit ephemeris of P95 which excludes the cycle 11, 12, 15, and 16 measurements

© European Southern Observatory (ESO) 1997

Online publication: July 3, 1998
helpdesk.link@springer.de