I have been analysing images of Stardial using part of the TASS pipeline from Michael Richmond. One star I recently found, showed some remarkable behaviour. It was constant from 1999 to 2001, but then in February and March 2002, it had an apparent "outburst" by 1.5 magnitudes in the course of some 50 days. The ascending and descending branch are nearly symmetric (see attached plot). I have found some variables in this way before, and I normally check a couple of images to verify nothing is wrong with them in the area of the variable. I never noticed something suspicious, and most "discoveries" were either confirmed because it concerned existing variables, or are confirmed by other means. So I was rather confident that the thing was "real". Also because the curve was almost perfect, and doesn't show much scatter (by the way, the error bars on the plot are a combination of errors on the photon statistics and the standard deviation for the image in the ensemble photometry). The star is in the overlapping zone of two Stardial regions, and therefore, instead of checking the images first, I reduced the other region (for which the images are taken 15 minutes after the images of the first region). To my astonishment, I found no variation whatsoever of the star. So I finally got to check the images. It turned out that the culprit is a geostationary satellite, leaving a half a degree trail almost exactly crossing the star in question. The xvista programs reject most pixels of the trail as non-starlike, but the star, still being brighter than the trail, was not rejected. So part of the trail within the chosen aperture is included in the star's brightness on the nights in question. Active geostationary satellites have a fixed direction with respect to the Earth. Now, about a month before the spring equinox and a month after the fall equinox (February and October respectively), the sun is at a declination for sunlight to reflect from geostationary satellites (probably from the antenna), directly to observers at mid northern latitudes. So at these times these satellites are brightest, and the light curve attached, is just a record of this phenomenon for one particular satellite. It is an awkward way to observe the changing declination of the sun ! I found three possible candidates for the satellite, one being launched in 2001, which might explain why the phenomenon was not observed before 2002. I checked recent images, and the trail is there again. However, it doesn't cross the star anymore, as the satellite apparently drifted westward somewhat. For anyone wanting to have a look, it's in the 10h00 region of Stardial, and the trail is obvious in the recent images. Patrick __________________________________________________ Do you Yahoo!? Yahoo! Shopping - Send Flowers for Valentine's Day http://shopping.yahoo.com
stardial.gif