SpringerLink
Forum Springer Astron. Astrophys.
Forum Whats New Search Orders


Astron. Astrophys. 328, 565-570 (1997)

Previous Section Next Section Title Page Table of Contents

2. Observations

2.1. The HU Virginis system

HU Vir (K0 III-IV, [FORMULA] [FORMULA] 10.4 days, [FORMULA] = 25 km s-1, V = 8.7 mag) is a rapidly rotating K0 star in a close binary system with an unseen (presumably late type) secondary component. It shows all classical signs of an "active" RS CVn star: light and color variability (Fekel et al. 1986), strong Ca II H and K emission (Strassmeier et al. 1990), H [FORMULA] and ultraviolet emission (Fekel et al. 1986), coronal X-ray emission ([FORMULA] = 2.5 [FORMULA] erg s-1 from RASS; Dempsey et al. 1993) and radio emission (Drake et al. 1989) as well as spectral line variations (Strassmeier 1994).

Our goal for the X-ray observations of HU Virginis was to detect the rotational modulation due to a proposed large coronal loop. An optical study by Strassmeier (1994) produced a pseudo 3D-Doppler map of the lower atmosphere of HU Virginis from rotational mapping. The photospheric and chromospheric maps each revealed two distinctive features [FORMULA] apart in longitude. Observations of the periodically varying H [FORMULA] profiles led Strassmeier to the conclusion that there is significant mass outflow in the middle chromosphere when the larger of the two active regions is visible and inflow when the smaller one is visible. He tentatively proposed that the two active regions are connected by a large coronal loop and that the mass flow structure is confined by a strong magnetic field and possibly is similar to a siphon-type flow. Because it is very likely that the X-ray flux is confined to localized regions in the corona (e.g. oversized flux tubes) the signal should be modulated when these active regions rotate in and out of sight. As mentioned before the X-ray luminosity of late F- to M stars is correlated with rotational velocity as [FORMULA] (Pallavicini et al. 1981), but shows a relatively large scatter: part of it is presumably due to rotational modulation. According to the [FORMULA] relation we expected an X-ray luminosity of 8 [FORMULA] erg s-1 for HU Virginis. However, the ROSAT all-sky survey already found HU Virginis to be much brighter: 2.5 [FORMULA] 1031 erg s-1 according to Dempsey et al. (1993).

2.2. The X-ray flare in 1994

In 1994 HU Virginis was observed from June 15, 20:00 UT, to July 12, 15:00 UT, with a total exposure time of 32.09 ksec, using the ROSAT satellite (Trümper 1983) and the HRI detector (David et al. 1993). All observations were performed in pointing mode. The point spread function of the HRI is such that 99% of the source photons are located within a circular area of 150". The data was binned into 200 sec intervals and the background subtracted. Data reduction was performed by using the EXSAS/MIDAS software (Zimmermann et al. 1994). Due to technical problems of the satellite the initial duration of the observation was cut to one third. The resulting poor phase coverage of the data did not allow the detection of any rotational modulation of the quiescent coronal emission. The count rate varied from 0.1 to 0.4 cts sec-1 which is equivalent to an energy release rate of [FORMULA] 0.4-1.4 [FORMULA] erg s-1.

However, on JD 2,449,544 a large X-ray flare was detected and observed for 1.5 days. Fig. 1 shows the light curve of the X-ray emission at that period of time. Flare onset occured at JD 2,449,544.63 corresponding to rotational phase 0.57 using the ephemeris of Strassmeier (1994). The energy output was rising from 0.16 cts sec-1 ([FORMULA] 5.7 [FORMULA] erg s-1) at flare onset to 3.85 cts sec-1 ([FORMULA] 1.4 [FORMULA] erg s-1) at flare peak. At the end of the observation the emission was still enhanced compared to the quiescent level before the flare, suggesting that the flare was still in progress at that time.

[FIGURE] Fig. 1. Background subtracted X-ray lightcurve of the long duration flare on HU Virginis (ROSAT-HRI observation in pointing mode). The X-ray luminosity rises from 5.7 [FORMULA] erg s-1 at the onset to 1.4 [FORMULA] erg s-1 at flare maximum. During 1.5 days the flare released a total energy of [FORMULA] erg in the 0.1-2.4 keV bandpass.

Fig. 2 compares an X-ray image of the quiescent emission of the star to a "snapshot" during the X-ray flare. By using the best-fit two-ribbon model light curve (see Sec. 3.3 and Fig. 5) we obtained a value for the total energy released by the flare according to the count rate to flux conversion given by Schmitt (1997) and the ratio of PSPC and HRI count rate of 3.2 (Kürster et al., 1997). For a given distance for HU Virginis of [FORMULA] pc (taken from the Hipparcos Catalogue, ESA 1997) the total energy output in the 0.1-2.4 keV bandpass was equivalent to [FORMULA] erg.

[FIGURE] Fig. 2. a (left) ROSAT HRI image of HU Virginis during its quiescent stage, and b (right) during the flare. The increase in X-ray brightness is obvious when compared to the constant second source in the field of view.

2.3. Second observations in 1995

Again, from June 18, 20:59 UT, to June 26, 13:00 UT 1995, HU Virginis was re-observed by the ROSAT satellite. The observation was performed in pointing mode using the HRI detector and a total exposure time of 69 ksecs. This time we obtained a far more continuous X-ray light curve that covers almost one complete rotational/orbital cycle. Fig. 3 shows the full background-subtracted light curve of HU Virginis for that time. We distinguish two "levels" of variability: first, significant variations on a timescale of [FORMULA] 2 days and second, a small flare at the end of the observation. Apparently the "quiescent" emission of the corona of HU Virginis still varied between 0.5-1.4 [FORMULA] erg s-1 and reached 2 [FORMULA] erg s-1 during the small flare at around 2,449,894.3. Compared to the 1994 light curve the mean emission level remained the same though.

[FIGURE] Fig. 3. Coronal X-ray emission of HU Vir during the 8-day observational period in 1995 showing short-term variability. The X-ray luminosity varies between 0.5 and 2 [FORMULA] erg s-1.

One question raised by this light curve is whether the apparent "flickering" is just noise or real short-term variability, e.g. microflares. Following the natural grouping of the measurements we subdivided the data into the individual observing slots: JD 2449887.0-88.0, 88.0-89.0, 89.0-90.0, 90.0-91.0, 91.0-92.0, 92.0-92.5, 92.5-93.0, 93.0-94.0, 94.0-94.4, 94.4-95.2. Examination of the data distribution in these "slots" showed that the standard deviation [FORMULA] of the data from their mean value is of the same size as the typical error. Therefore no evidence of variability on timescales shorter than two days is present.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: March 26, 1998

helpdesk.link@springer.de