STIS has a single coronagraphic mask aperture for direct imaging. The aperture
) contains one occulting bar and two intersecting wedges and is shown in Figure 12.6
. This illustration of the coronagraphic aperture is derived from an on-orbit lamp flat. The approximate positions of the predefined aperture locations are marked. The wedges vary in width from 0.5 to 3.0 arcseconds over their 50 arcseconds length, while the rectangular bar measures 10 by 3 arcseconds. The small occulting finger on the right was damaged during the assembly of STIS and is not used. The entire coronagraphic aperture measures 50 ×
50 arcseconds, slightly smaller than the size of the unobstructed CCD aperture. The parallel readout of the CCD is along the AXIS2
direction, and heavily saturated images will bleed in this direction (vertically in this figure). Saturation near the top of the detector can result in serial transfer artifacts that produce tails in the AXIS1 direction; for additional details of this effect see the discussion in the first of the two June 2013 STScI Analysis Newsletters (STAN)
The aperture cannot be combined with a filter and so, when used with the CCD,
yields a bandpass of ~2000–10,300 Å. See Section 5.2.1
for the spectral properties of the images obtained. A number of locations on the occulting masks have been specified, to correspond to widths of 2.75, 2.5, 2.0, 1.75, 1.0, and 0.61
arcseconds on each wedge. The mask is not available for use with the MAMA detectors due to concerns about bright object protection of the MAMAs.
In combination with the option of a coronagraphic mask, there is a limited amount
of apodization via a Lyot stop which masks the outer perimeter of the re-imaged exit pupil. Consequently, diffraction from the secondary mirror assembly and the telescope spider is not apodized. The STIS coronagraphic imaging facility is well suited to imaging problems involving faint material surrounding a relatively bright source. Typical examples include circumstellar disks, such as β Pictoris, and the host galaxies of bright QSOs.
In Figure 12.7
we provide a comparison of the PSF suppression provided by STIS coronagraphic imaging relative to WFPC2 imaging and the Optical Telescope Assembly scatter. It had been hoped that the optical performance of the STIS CCD clear aperture without the coronagraph would be comparable to that with the coronagraph, although, without the coronagraph, the CCD long wavelength halo from the central source and the window reflection ghosts are present. In practice the coronagraph does provide substantial additional suppression of the PSF wings, especially at wavelengths >8000 Å, where the halo of light scattered within the CCD itself dominates the far wings of the PSF.
Due to the very broad bandpass of the unfiltered STIS CCD, the STIS
coronagraphic PSF shape is very strongly dependent on the target’s spectral energy distribution. When using the coronagraph to look for a point source or a localized structure that is strongly asymmetric (such as an edge-on disk), the best approach is to observe the target star at a minimum of two and preferably three different roll angles, and then compare the images to separate the stellar PSF from the real circumstellar structure. When looking for more diffuse or symmetric material, it will be necessary to use a separate comparison star. Here it is important to match the colors of the target and comparison star as closely as possible. We suggest that all the broadband UVBRI color differences be less then 0.08 magnitudes. In either case, it is also essential to compare stars at the same location on the coronagraphic mask.
Breathing and focus differences will also significantly affect the quality of such a
subtraction, but the observer has only limited control over these parameters. The best alignment of STIS PSF images tends to occur when comparing images taken in the same part of adjacent orbits. When observing the same star at multiple roll angles, it is therefore often useful to do a sequence of adjacent one-orbit visits, each at a different roll angle. As large departures from the nominal roll angle can also affect the PSF shape, it may be helpful to keep the roll changes as small as is consistent with the structure to be imaged. When observing a separate comparison star, it is best if possible to observe a star in the same part of the sky, during an adjacent orbit, and at the same angle relative to the nominal spacecraft roll, as the observations of the prime target, but remember that picking a comparison star with a good color match must be the first priority.
An alternative strategy would be to take observations separated by several days, but
constrained so that each observation is done at the nominal spacecraft roll. When very large roll changes or several orbit-long visits are required, this might give better results than doing the observations in adjacent orbits, but there is very little operational experience using this approach.
Attempting to observe multiple coronagraphic targets or the same target at different
roll angles in a single orbit is not recommended. The overheads required to do separate visits in a single orbit are very large, and the PSF alignment between different parts of the same orbit are usually inferior to that obtained between the same part of adjacent orbits.
Coronagraphic images of stars of various colors have been obtained as parts of
calibration programs 7151, 7088, 8419, 8842, and 8844 and are available from the archive. These images may be useful in providing comparison objects or in estimating exposure times. However, for the best PSF subtraction, we still recommend that each coronagraphic program include its own tailored PSF observations.
In planning any observing program with the 50CORON
aperture, observers should carefully consider the required orientation of the target. The telescope’s V2 and V3 axes are at 45° to the STIS AXIS1/AXIS2
coordinate system (see Figure 11.1
) and so diffraction spikes further reduce the unocculted field of view.
A series of apertures has been defined for the coronagraphic mask so that targets
can be placed on the 3 arcseconds wide bar and 5 locations on each of the two wedges. These apertures are summarized in Table 12.6
below. We defined a special coronagraphic acquisition technique for placing stars at these predefined locations. This involves performing a bright-target acquisition with a filtered aperture, followed by a slew to the chosen location on the coronagraphic mask. An example of an acquisition into one of the bars on the 50CORON
aperture is provided in Section 8.5.6