Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 345, 121-126 (1999)

Previous Section Next Section Title Page Table of Contents

1. Introduction

The flare phenomenon as a short term variability of brightness of stars is thought to be produced by rapid energy release and is observed in almost all types of stars. The presence of flares in low-mass stars is supported by many observational facts at many wavelengths and is well documented. Stellar flares were discovered at optical wavelengths in the 1940 s on the so called UV Cet type variables. The quiescent radiation of these stars is interrupted by randomly distributed flares. UV Cet type stars have very low luminosities and are seen in the vicinity of the Sun (Gershberg et al. 1998). Further, systematical investigations of stellar flares in nearby stellar clusters (e.g., Pleiades) and associations (e.g., Orion) have revealed a fundamental significance of this phenomenon for the physics and evolution of the young low-mass stars and protostars (see, e.g. Mirzoyan 1995, Montmerle 1997). At present, the observations of stellar flares at different wavelengths of the electromagnetic spectrum has become a scientifically productive area of astrophysics.

In the 1970 s, the first X-ray studies of stellar flares were made, in particular with the Einstein Observatory (EO ) (1978-1981) and the EXOSAT Observatory (1983-1985). Comprehensive reviews of the EO observations of stellar X-ray flares are presented by Haisch (1983) and of the EXOSAT observations by Pallavicini et al. (1990). X-ray observations of stellar flares obtained during the ROSAT All-Sky Survey as well as in ROSAT pointings are discussed by Schmitt (1994).

However, to date no systematic investigation of flare stars and stellar flares based on ROSAT observations has been done. X-ray observations by ROSAT are very well suited for statistical examination: Several physical quantities of individual X-ray photons are registered, namely the time of photon arrival, the direction of incidence, and the photon energy.

On the other hand, the wide spread timing analysis methods of datasets based on time binning technique give rise to certain difficulties: First of all, one has no a priori knowledge of the relevant timescales of a given flare to be detected, and therefore many different binnings of the data have to be considered; second, the bins must be large enough so that there will be enough photons to provide a good statistical sample, while larger bins will dilute short flares; moreover, the common practice of binning data overlooks a considerable amount of information and introduces a dependency of results on the sizes and locations of the bins. Meanwhile, it has been known for some time (Cash 1979), that fitting models to unbinned data is quite straightforward (see, for discussion, Scargle & Bapu 1998).

Hence, it is worthwhile to examine the available observational data of ROSAT in a straightforward and systematic way for variability testing, in particular, for UV Cet type flare stars.

For the flare event detection, we shall use the method described by Scargle (1998) based on Bayesian statistics due to the nearly ideal Poisson nature of photon registration by the ROSAT detectors (Trümper 1982, Pfeffermann et al. 1986).

In this article we will try to address various items relevant to detections and studies of flares with ROSAT . First, we will describe the flare event detection algorithm applied to a typical ROSAT observational data set; second, we will demonstrate this technique with an example, a flare event detected in ROSAT PSPC pointing observations on the nearby flare star No. 3 included in the Catalogue of UV Cet type stars and related objects (Gershberg et al. 1998).

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: April 12, 1999