[Top] [Prev] [Next] [Bottom]

9.8 Astrometry with FGS 3

Among the three FGSs on HST, only FGS 3 has been calibrated and used as an astrometer because the performance of FGS 1 and FGS 2 is not adequate for astrometric science (see "S-curves" on page 9-8). This section will discuss the overall performance characteristics of FGS 3 in both POSITION mode and TRANSFER mode.

POSITION Mode Characteristics

When FGS 3 is used as an astrometer, the "full" F583W wide bandpass spectral filter is used in POSITION mode for stars of V > 8 to maximize the sensitivity. The F5ND attenuator is used for stars with V < 8. The instrumental properties affecting POSITION mode observation planning and data calibration are: the spatial dependence of the S-curve, the optical field angle distortions across the pickle, and the slow temporal changes in the plate scale, probably due to outgassing of the graphite epoxy. In those areas of the pickle with lower S/N in the S-curve, the limiting magnitude rises from V = 17 to V = 15. A large area in the central region of the pickle has adequate signal to noise so this problem affects observations only at the edges.

The calibration of optical field angle distortion (OFAD; see "Processing Individual Observations" on page 11-4) involves a fifth-order two-dimensional polynomial fit to observations of an astrometric standard field. The measured coordinates of a target are corrected in the pipeline using the transformation defined by these polynomial coefficients. The temporal changes to the plate scale and the OFAD are monitored on a monthly or bi-monthly basis via observations of a standard astrometric field and are provided as updates to the calibration database. In spite of the field dependent behavior of FGS 3, the repeatability and hence the accuracy of individual POSITION mode measurements of stars distributed throughout the pickle is typically about 1.5 mas for V < 14.5 and about 2 mas for fainter objects. These numbers are determined from the residuals of plate overlays in which the pickle's orientation on the sky is constant from plate to plate, so that field-dependent (OFAD) corrections do not enter. However, when the OFAD correction is required, the overall pickle-wide error budget increases to about 2.7 mas.

TRANSFER Mode Characteristics

The S-curves of FGS 3 display strong field dependency on both the x and y axis. Each of these S-curves is nearly ideal at one location in the pickle, but the optimal position for the x axis does not coincide with that for the y axis. This lack of coincidence has affected the astrometric performance of FGS 3 because science observations in TRANSFER mode are executed at only one position, pickle center. Any other position could be used, but pickle center was chosen as an optimal compromise and is the only position in FGS 3 supported by the TRANSFER mode calibration database. The y-axis sports a nearly ideal S-curve, while that on the x-axis suffers from aberrations.

Because the x-axis has the degraded S-curve, it is the limiting factor in the TRANSFER mode performance of FGS 3. When the projected angular separation of a binary system along the x-axis is less than about 20 mas, it is not resolved on the x-axis. On the other hand, the y-axis has been shown to resolve systems with separations as small as 10 mas. (The success at resolving the individual components depends, of course, on the magnitude difference: the larger the delta magnitude, the more difficult the observation.) This suggests that FGS 1R, with its nearly ideal S-curves on both axis at the same place in the pickle (at the center), should be able to resolve binary systems with separations on the order of 10 mas, and perhaps to detect structure in an object at the 5 mas level.

[Top] [Prev] [Next] [Bottom]

stevens@stsci.edu