BeppoSAX Ephemeris

Once the spacecraft attitude is known, it is of key importance to know the position of the Earth with respect to the BeppoSAX frame of reference, to exclude that portion of the sky that is hidden by the Earth ($\sim 30$% out of the whole sky), and, if necessary, to discard any occulted direction, among the possible solutions resulted from the localization procedure. Unfortunately, this kind of info cannot be promptly distributed from the SOC: if, on one hand, any ephemeris information is included in every FOT data (usually, at least four or five days are needed for producing them), on the other hand, there is no way to automatically receive the satellite coordinates at the burst time by the SOC. The solution to this problem has been gained by the author of this work, by means of an automatic program calculating the BeppoSAX position. This program reads the NORAD two-line orbital elements of BeppoSAX (TLE format) from a local archive, that is continuously and automatically maintained and updated: the current BeppoSAX TLE elements are retrieved from a web site at the following web page: http://www.celestrak.com/NORAD/elements/science.txt, reporting the TLE elements of several scientific satellites currently orbiting round the Earth. On average, the BeppoSAX elements are updated every two or three days.

At this stage, the BeppoSAX position is calculated at the burst time: the accuracy of the algorithms adopted has been tested with several hundreds bursts from this catalog, for which the BeppoSAX position was already known from the FOT ephemeris files; it resulted that the mean discrepancy between the true and the calculated position of the spacecraft was $\sim$ 30-40 Km, which turns into a $< 1\rm ^{\circ}$ uncertainty in the position of the Earth limb. Since even the smallest error boxes obtained with the GRBM localizations are much greater ( $> 10\rm ^{\circ}$, including systematics) it is straightforward that the error on the spacecraft position can be neglected.

Concerning the error region for every direction: every error box is obtained at 90% confidence level (CL), when possible, otherwise 68% or 50%; furthermore, in addition to this purely statistical box, a $\sim 10\rm ^{\circ}$ systematic error is taken as well. Eventually, after rejecting any position that might be completely Earth-blocked with its entire error box, there are several cases: when the number of distinct and acceptable solutions is greater than two, then no final position is given; otherwise, when the confidence level concerning the error region of each position is meaningful (this condition, of course, is not always verified), then the position(s) is (are) assumed reliable.

For each final position, the following pieces of information are available: coordinates of the centroid position and of the four error box corners; angular distances of these same points (i.e., centroid + corners) from the Sun, in order to reject possible solar hard X-ray flares; elevation angles above the Earth limb of these points, to evaluate whether the error box is hidden by the Earth; mean angular distance from the centroid of each of the four corners, roughly corresponding to a sort of ``mean error radius'' (for a strict definition, see sec. ), though the error box rarely shows a circular shape.