2 RASS data and detection of late-type stars

2.1 The ROSAT all-sky survey (RASS)

During its first half year of operations, the ROSAT observatory carried out the first all-sky survey with an imaging X-ray telescope between July 1990 and January 1991. Further survey observations were carried out in February 1991 (2 days) and August 1991 (10 days). The whole sky was scanned along great circles perpendicular to the direction to the Sun. Because of the Earth's motion around the Sun, the plane of these circles slowly (1$^\circ$/d) rotated around an axis through the ecliptic poles, thus covering the whole celestial sphere within 6 months. Each point of the sky was observed several times as the scan paths of 2 degrees width (i.e., the field of view of the PSPC detector) progressed along the ecliptic. Therefore, the data of any particular source consist of a number of "snapshots'' of up to 30 s duration, separated by the orbital period of the satellite ($\approx 90$ min) and distributed over an interval of at least 2 days. Towards the ecliptic poles, the cumulative exposure time increases due to the larger number of scans covering a particular celestial position. Depending on the ecliptic latitude (and down-time due to passages through the radiation belts of the Earth), the effective exposure time varies between $\sim \!100\,{\rm s}$ and $\sim\! 40\,000\,{\rm s}$ (at the poles), with typical values of $\sim \!400\,{\rm s}$ at the ecliptic. Typical limiting RASS count rates are $\approx 0.015$ ctss^-1; given a typical energy-conversion factor for soft sources of $6\ 10^{-12}$ ergcts^-1 cm^-2 (cf. Sect. 2.3) the typical detection limit of RASS observations amounts to a limiting flux of $f_{\rm x}\approx 10^{-13}$ergcm^-2 s^-1. At the distance limit of the Gliese catalogue, i.e., at 25 pc, this corresponds to an X-ray luminosity of $L_{\rm x} \approx 7.5 \ 10^{27}$ ergs^-1. Note that this is not an average value for the detection limit for the Gliese stars since many stars are closer and the exposure time increases towards the ecliptic poles.

For a more detailed description of the RASS we refer to Voges (1992) and Belloni et al. (1994). Details of the ROSAT observatory in general can be found in Trümper (1983) and Trümper et al. (1991). The PSPC detector used during the RASS is described by Pfeffermann et al. (1986). In February 1997 the remaining gaps left in the all-sky survey were filled with a sequence of more than 500 pointed, partially overlapping PSPC observations so that with the exception of a small region around the strong X-ray source Sco X-1 the whole sky has been imaged with the ROSAT PSPC. In the catalogue presented in this paper we include sources detected in this "survey repair'' pointed observations; they are marked with an asterisk.

The source detection was performed by means of a maximum likelihood algorithm (Cruddace et al. 1988) in the course of the standard analysis software system (SASS; Voges et al. 1992). The significance of an X-ray source is expressed by the likelihood Li = -ln(1-P), where P is the probability of existence; e.g., a likelihood of Li = 7 corresponds to a source existence probability of 99.9%. The result of the SASS is a comprehensive list of approximately 10⁵ sources, each described by the sky position in right ascension and declination, its source detection likelihood, count rate, hardness ratios, extent, and corresponding errors. The data for the brighter X-ray sources have been released as the ROSAT All-sky Survey Bright Source Catalogue (Voges et al. 1996b), which contains sources with Likelihood $\ge$ 15, count rate larger than 0.05 s^-1, and with at least 15 detected photons.

2.2 Identification of X-ray sources with nearby stars

We used the Third Catalogue of Nearby Stars (Gliese & Jahreiß 1991) as input sample for our search of X-ray bright nearby stars. That input sample consists of 3802 stars.

The procedure whereby the positions of RASS sources were matched with the stars of our input sample is the same as described in HSV98. We accepted sources with a likelihood greater than or equal to 7 within 90 arcsec distance from the input stars. As for the BSC sample, the choice of this cut-off radius is empirically justified by means of a Monte Carlo simulation of about the same number of random positions as input positions. However, since the binary fraction is much larger in the Gliese catalogue than in the BSC, the number of independent input positions is significantly smaller than the total number of catalogue entries; we therefore combined all binaries into one input position for each system, since most of them are too close to be separated with the RASS data. This results in only 3365 independent input positions, for which we determined the distribution of offsets. The same number of random positions results in 112 (artificial) matches with X-ray sources, yielding a mean of $1.1 \ 10^{-3}$matches per square arcsec in the offset distribution plane, i.e., about one third of that of the random sample used for the 9110 BSC stars (as expected). At 90 arcsec offset the number of matches of X-ray sources with real stars exceeds the number of artificial matches by a factor of 2. That means, at 90 arcsec offset between optical and X-ray position the differential probability that the X-ray source can be attributed to the star (and not to a background object) is 50%. This differential probability increases very rapidly for smaller values of positional offset, while for even larger values of offset the chance for obtaining a spurious identification exceeds that of finding the true X-ray counter part.

We note that the accuracy of the input positions in the Gliese catalogue (given only to integers of seconds in RA and tenth of arcminutes in Dec) is less than for the BSC stars, hence resulting in a somewhat broader distribution of the offsets for the real stars. On the other hand, the intrinsic detection probability is larger for the Gliese stars than for the BSC stars because the Gliese stars are closer to us and the content of late-type stars is much larger. This would cause a somewhat steeper distribution of the offsets. Probably, both effects compensate each other, thus leading to a 50% differential probability for a correct identification at essentially the same offset value.

Of the X-ray sources extracted by the match procedure, 469 are rather weak sources that are not included in the Bright Source Catalogue (Voges et al. 1996b). We checked their X-ray images by eye for reality. Specifically, we rejected photon distributions that are significantly contaminated by nearby strong sources or that are obviously extended. In questionable cases, we ran the standard source detection algorithm of EXSAS on the source images in different passbands and decided on the basis of the results which sources to retain in our final catalogue.

Confining now attention to the 3365 (independent) input positions identified with Gliese stars, we detected X-ray emission from 1252 stars, i.e., the average detection rate is 37%. Since the total search area around these 3365 stars is $3365\cdot \pi \cdot (1.5\hbox{$^\prime$})^2 = 6.61\ifmmode\hbox{\rlap{$\sqcap$}... ...}$\sqcup$} \parfillskip=0pt\finalhyphendemerits=0\endgraf}\fi^{\circ} = 0.016\%$ of the sphere, and the total number of RASS sources amounts to $\sim 150\,000$, we would expect 24.0 chance coincidences of Gliese stars with background (or foreground) X-ray sources (i.e., 1.92% of our detected sources).

2.3 Determination of X-ray fluxes and luminosities

The procedure of determining X-ray fluxes has also been described in HSV98. In this paper, we followed the same procedure, except using a slightly different formula for the calculation of individual energy-conversion factors

$${\rm ECF} = (5.30 \cdot HR + 8.31)\; 10^{-12}\,{\rm erg\,cm}^{-2}\,{\rm cts}^{-1}$$ (1)

which was derived by Schmitt et al. (1995) from an X-ray study of a complete sample of main-sequence stars within 7 pc distance; here HR denotes the hardness ratio defined through

$$HR = \frac{{H - S}}{{H + S}},$$ (2)

where H and S denote the source counts in the hard (0.5-2.0 keV) and soft (0.1-0.4 keV) passbands of ROSAT. The hardness ratio is an "X-ray color'' that is influenced by both the plasma temperature and the hydrogen column density.

Since the SASS source detection was separately performed in both passbands and since most of our X-ray sources were detected in both bands, the hardness ratios can be estimated for many stars, although in some cases with quite substantial errors. In a few cases, when the sources were not detected in either the soft or the hard passband, we set HR = +1.0 or -1.0 by definition, respectively. We refrain from estimating individual errors for $f_{\rm x}$since the error in ECF is very difficult to quantify. In general, we estimate this error to be within a factor of two for the weaker sources and less for the brighter sources.

The X-ray luminosities are calculated by the relation

$$L_{\rm x} = 4 \pi d^2 \times f_{\rm x} ,$$ (3)

where d is the distance to the star. We used the distances revised on the basis of the Hipparcos parallaxes (ESA 1997) and kindly made available to us by H. Jahreiß. No X-ray luminosities are computed for those few stars for which no reliable distances exist. Note that the catalogue contains a few stars which obviously do not belong to the solar environment but were erroneously included in the third version of the Gliese catalogue.