A multiwavelength campaign on gamma cassiopeiae. I. The case for surface X-ray flaring

In 1996 March we obtained simultaneous Rossi X-Ray Timing Explorer RXTE Pro portional Counter Array (PCA) and Hubble Space Telescope (HST) Goddard High -Resolution Spectrograph (GHRS) light curves for the B0.5e star gamma Cas i n order to compare its X-ray and ultraviolet continuum flux behaviors. The GHRS data set consisted of a nearly continuous sequence of UV spectra cover ing a 21+ hr interval. Each 40 Angstrom spectrum was centered on the Si IV lambda lambda 1394-1403 lines and registered 8100 counts in each 1 s exposu re. Combining spectra and integrating over greater-than 100 continuum pixel s allowed us to define a UV continuum light curve binned to 1 minute with a signal-to-noise ratio of a few thousand pixel(-1). We found that the light curve exhibited variations over a time comparable to the rotation period o f the star, showing two broad minima 10 hr apart, which had depths of 0.8% and 1.8%. The long-term trends in the UV are anticorrelated with the X-ray fluxes, with the X-rays exhibiting increases of similar to 10% and similar to 40% during times of UV flux minima. The stability of the long-term X-ray variations on gamma Cas is supported by phasing our March data with contem poraneous ASCA data, suggesting a possible period of 1.125 days (or a close alias). We also get agreement of dip patterns for an assumed 1.123 day per iod by phasing the GHRS continuum flux curve with IUE light curves in vario us wavelengths from 2 months earlier. We take this as an estimate of the st ar's rotational period. We conclude that the X-ray emission from gamma Cas probably consists of two components. The first is a slowly varying "basal" flux representing the minimum level seen during any given phase. Superimpos ed on this are rapid fluctuations ("shots") that have lifetimes ranging fro m greater-than 10 a to greater-than-or-equal-to 10 minutes. The character o f these components varies from one spacecraft orbit to the next, indicating that the emissions are not produced in a truly "stationary" chaotic enviro nment. Moreover, both the number and amplitude of the shots increase during UV minima. The shot profiles are typically symmetric and can have decay ti mes of a few seconds or less. The shots also have a slightly harder flux di stribution than the basal component, suggesting that the two emission regio ns are not cospatial. The time-averaged X-ray spectrum indicates a quasi te mperature of similar to 10(8) K, in agreement with earlier studies. We present a picture in which magnetically generated structures on and over the star's surface are responsible for the basal and shot X-ray components . The energies and luminosities of the shots are so high that even the weak est events we measure are comparable in strength to the most luminous flare s on cool active stars. Using general cooling relations for a thermal plasm a, one finds that the source region for the shots probably have a size scal e of less-than-or-equal-to 10(4) km and densities of similar to 10(13-14) c m(-3). From a simple flare model, we find that generally only a small fract ion of the shot energy is radiated during the event itself. The remainder o f the hot plasma expands to fill a confined volume, possibly a magnetic loo p, connected to the original hare site. A collection of these loops may the n account for the basal emission. With this model, we estimate that the ind ividual loops have a characteristic density of similar to 10(11) cm(-3) and dimensions of greater than or equal to 0.1R(*). We note that the magnetic interpretation for the shot and basal emission poses several theoretical qu estions, such as how complex, dynamic fields can exist on a star that does not have a convective envelope. These results suggest that gamma Cas is a m ember of an arguably new group of hot stars that flare continuously in X-ra ys. This group may represent an extension of the hotter Bp stars to high va lues of rotation.