Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 334, 873-894 (1998)

Previous Section Next Section Title Page Table of Contents

8. X-ray monitoring

Due to its position close to the Orion nebula, P1724 has been observed by ROSAT quite frequently: In 1990 it was observed with the ROSAT All-Sky Survey (RASS), and between 1991 and 1993 it was in the field of view of seven pointed ROSAT observations with both the Positional Sensitive Proportional Counter (PSPC) and the High Resolution Imager (HRI). For details on ROSAT and its instruments, we refer to Trümper (1983).

During the period 1 to 21 Sep 1995, J.-P. Caillault and co-workers performed a sequence of ten ROSAT HRI observations of the Trapezium region. The typical exposure time of each observation was 2 ksec, and the individual observations were separated by about two days. The X-ray sources in the Orion nebula cluster were monitored in this way for a period of 20 days (cf., Gagné et al. 1997 for the first results of this study). The details of all available ROSAT observations are given in Table 5.


Table 5. ROSAT pointed observations of P1724. ROR is the ROSAT observation number, the prefix `P' indicates observations with the PSPC detector, the prefix `H' those with the HRI detector. We list the dates of the start and end of the observation period, the exposure times [FORMULA], and the mean count rates [FORMULA] for each observation. For the sequence of monitoring observations we list the minimum and maximum count rates.

We obtained the data for all of these observations from the ROSAT data archive and analyzed them with the Extended Scientific Software Analysis System (EXSAS, Zimmermann et al. 1995). With one exception, P1724 was clearly detected as an X-ray source in all these observations. The exception is pointing 200700, which had a very short exposure time of only 700 sec, and in which P1724 was located at a large off-axis angle of [FORMULA], where the sensitivity of the detector is significantly reduced. From this observation we determined an upper limit to the count rate of P1724.

In 1993 a deep ASCA observation of the Orion region was performed (Yamauchi et al. 1996). Again, P1724 was clearly detected as one of the brightest X-ray sources within the region observed.

8.1. The X-ray spectrum

Since Preibisch et al. (1995) describe the spectral analysis in detail, we only summarize here the results of the spectral fits. The X-ray spectrum of P1724 was extracted from PSPC observation 200151. In this observation the count rate of P1724 showed no significant variations and thus we are confident that these data yield the quiescent X-ray spectrum. An isothermal plasma model clearly failed to reproduce the spectrum. A two-temperature model gave an acceptable fit with [FORMULA], [FORMULA], and [FORMULA].

The spectrum could be equally well fitted by a model with a continuous temperature distribution in the form of a power law up to a maximum temperature (cf. Schmitt et al. 1990 or Preibisch 1997 for a discussion of this model). The fit with this model gave [FORMULA], a maximum temperature of [FORMULA], and a power-law exponent of [FORMULA].

From their deep ASCA observation Yamauchi et al. (1996) also extracted spectra for P1724. A simultaneous two-temperature fit to the SIS and GIS spectra yielded [FORMULA], [FORMULA], [FORMULA], which agrees very well with the results of our fit to the ROSAT spectrum.

From the spectral parameters of the fits one may calculate the X-ray luminosity in the 0.1 to 2.4 keV ROSAT band. The X-ray luminosity computed for the two-temperature model agrees very well with that for the power law temperature distribution model. Assuming a distance of [FORMULA], we find [FORMULA] erg/sec. This is the quiescent X-ray luminosity during observation 200151 in March 1991. This observation was also used to determine the conversion factor between count rate and X-ray luminosity for other ROSAT observations.

8.2. X-ray variability

We extracted background-corrected lightcurves for P1724 from all ROSAT observations and analyzed them for signs of variability. While most of the lightcurves show only small-amplitude irregular variability, significant enhancements of the count rate can be found in the light curve extracted from the RASS data and from the HRI observation 200500.

The large flare discussed in detail by Preibisch et al. (1995) occurred during the third part of observation 200500 in Sep 1992. Since Preibisch et al. (1995) assumed a distance of 500 pc, the energetics of this flare have been slightly overestimated. However, even with an assumed distance of 460 pc the total flare energy radiated in the ROSAT X-ray band is [FORMULA] erg, and this is still one of the most energetic flares ever observed on a star. Only the IR protostar IRS 43 (=YLW 15) has been caught by ROSAT HRI at an even more energetic flare (Grosso et al. 1997).

During the RASS observations in 1990 (Fig. 14a) and also during the first part of observation 200500 in Oct 1991 (Fig. 14b), the X-ray lightcurve shows systematic enhancements of the count rate lasting for [FORMULA] hours. In both events the maxima are not narrow and rise and decay times are approximately equal, which is quite atypical for a flare. The variations might indicate rotational modulation (Neuhäuser et al. 1995b); however, this would imply a rotation period of [FORMULA] hours, clearly inconsistent with the observed photometric period of 5.7 days (cf. Sect. 2).

[FIGURE] Fig. 14a-c. X-ray lightcurves. We show ROSAT count rate versus time of observation. a ROSAT All Sky Survey (upper panel), b first part of HRI observation 200500 (middle panel), and c HRI observations in Sep 1995 (lower panel)

An alternative interpretation might be a flaring event which occurred on the back side of the star shortly (compared to the rotation period) before the flaring plasma became visible at the stellar limb. Two effects are relevant: (1) The flaring plasma cools, so that its X-ray emission decreases, but (2) more X-ray emitting plasma gradually becomes visible as it rotates into view. These two effects together might qualitatively account for the observed light curves (cf., Casanova 1994). During the X-ray flare of Sep 1990, the rotational phase at the beginning of the flare event was 0.013; and the beginning of the flare in Oct 1991 took place at phase 0.612. Hence, both flares took place roughly midway between minimum and maximum visible light, somewhat closer to minimum. Assuming that the X-ray flares originated at or near the dark spot area, this suggests that the flares may have occurred rather close to the limb of the star. We shall investigate this scenario quantitatively in more detail elsewhere (Neuhäuser et al., in preparation).

The lightcurve constructed from the HRI monitoring observations in 1995 is shown in Fig. 14c. While the variations appear quite smooth, the lightcurve cannot be described as a sinusoidal variation. Neither a period of 40 hours nor 5.7 days can reproduce the observed lightcurve. Hence, we do not see any evidence for rotational modulation in the X-ray data.

The data point for observation 201792 (JD 2449977) shows a significantly higher count rate than in the other observations of the series. A closer look at the lightcurve within observation 201792 indicates a decreasing count rate. However, due to the short exposure time of this observation (2.2 ksec) no meaningful analysis of this lightcurve is possible. Most probably P1724 showed another flare a few hours before the beginning of observation 201792.

The ASCA lightcurve presented in Yamauchi et al. (1996) shows variability of a factor of about 2. One data point in their binned lightcurve is roughly a factor of 3 above the mean, but no flare-like decay can be seen.

Finally, we used all available data to construct a long term X-ray lightcurve for P1724, covering approximately 5 years (Fig. 15). The variation in X-ray luminosity between 1990 and 1993 by about a factor of 2 is consistent with the variation found from the monitoring in Sep 1995 on a shorter time-scale.

[FIGURE] Fig. 15. Long term X-ray variation. We show X-ray luminosity in [FORMULA] versus observation date given in years. The first data point is from the ROSAT All-Sky Survey. The upper limit from observation 200700 is marked by an arrow. The X-ray luminosity reported by Yamauchi et al. (1996) for the (0.5 - 10 keV) ASCA band was transformed to the ROSAT band (0.1 - 2.4 keV) by using the spectral parameters from the 2T fit

8.3. Discussion of the X-ray properties

The mean quiescent X-ray luminosity of P1724 is [FORMULA]. This is quite high, even for an X-ray luminous TTS. By comparison with the X-ray luminosities derived by Gagné et al. (1995) for the TTS in the Orion nebula cluster, P1724 seems to be the most X-ray luminous TTS in Orion, surpassed only by the O and B stars. However, in the sample of Gagné et al. (1995) there are five stars of spectral type G or K with X-ray luminosities between 3 and [FORMULA]. Thus we conclude that the X-ray luminosity of P1724 is extreme but still consistent with that of the other X-ray luminous TTS in the Orion nebula cluster.

Using the bolometric luminosity of [FORMULA] from Sect. 5, we find a ratio [FORMULA]. This is in the typical range as reported by Gagné et al. (1995) for the most X-ray luminous G and K stars (between [FORMULA] and [FORMULA]). Similarly, the X-ray surface flux of [FORMULA] is in the typical range for X-ray active TTS, which is 0.03 to [FORMULA] (cf., Neuhäuser et al. 1995a, Preibisch 1997).

Gagné et al. (1995) thoroughly investigated the variability of the X-ray sources in the Orion nebula cluster. They found that at least 1/4 of the TTS show significant X-ray variability on time scales of about one year with typical amplitudes of factors of 2 to 6. Thus, the X-ray variability displayed by P1724 is quite typical for TTS. The variability cannot be explained as rotational modulation. This means that the X-ray emission is not dominated by a single active region or by a dense group of active regions associated with the predominant spot group. The X-rays probably originate from a multitude of active regions distributed widely on the stellar surface.

We conclude that the X-ray properties of P1724 are extreme in some aspects, but nevertheless are consistent with those of other X-ray bright TTS in the Orion nebula cluster.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1998

Online publication: June 2, 1998