GHRS Instrument Handbook
The GHRS is, of course, primarily a spectrograph, but it includes useful imaging capabilities, especially because the detectors of the GHRS are blind to much of the visible light that dominates the flux of most stars. You may wish to request an IMAGE, for example, to confirm that the telescope had your object properly centered in the data-taking aperture before the exposure was taken.
Note the following in using the imaging capability:
- GHRS IMAGEs and MAPs are obtained with the focus diodes (see Section on page 80) at the ends of the array of main science diodes. The focus diodes are smaller and square, making them more useful for focusing, but at the price of a lower count rate. The total count rate over the LSA is, of course, unchanged, and it is that which is predicted with the information in Section on page 85. Multiply the count rate estimated for the regular diodes by approximately 0.3 to get the value appropriate to the focus diodes when they are centered on the star.
- A MAP is obtained as an integral part of an acquisition whereas an IMAGE is a separate observation that may or may not have anything to do with an acquisition. A MAP with SEARCH-SIZE=3 or 5 is made as the acquisition procedure causes the telescope to make small motions in a square spiral pattern, thereby enabling it to record a larger portion of the sky than the LSA itself subtends. An IMAGE can only record the light in the 1.74 1.74 arcsec region of the LSA. A single MAP (SEARCH-SIZE=1) is equivalent to an IMAGE. Note that MAP=ALL-POINTS may not be used with an ONBOARD ACQ.
- A standard IMAGE will have a pixel spacing of 0.109 arcsec and will cover the entire LSA aperture of 1.74 1.74 arcsec. You may also select pixel spacings of 0.055 or 0.027 arcsec, with proportionately smaller regions of the sky covered in a 16 16 (the default) IMAGE. You may also use IMAGE with the SSA.
- The PSF of the GHRS is not simple nor has it been well-studied. If you need high-quality imaging, we recommend that you consider the FOC or WFPC2.
- Either the LSA ("2.0") or SSA ("0.25") may be selected as the aperture. The SSA is so small that it is generally pointless to image it, although there may be special cases where IMAGE mode is of use, particularly for confirming pointing in a crowded field.
- A mirror is the usual choice as optical element. A grating may also be specified--see below.
- The number of pixels in the x and y directions can be chosen separately and can range from 1 to 512 pixels. However, a large number of pixels only oversamples the region of sky subtended by the LSA and does not make the IMAGE include a larger area. The parameters to specify are NX, NY, DELTA-X, and DELTA-Y, for which the defaults are 16, 16, 4, and 4, respectively. The product of NX and NY may not exceed 512. An image that is critically sampled in the x direction may be obtained by specifying NX=32, NY=16, DELTA-X=2, and DELTA-Y=4.
- The PRECISION parameter may be specified as NORMAL (the default) or HIGH. PRECISION may only be specified if DELTA-Y=4. Using PRECISION=HIGH causes the image to be obtained with only one focus diode instead of two, thereby eliminating uncertainty over the relative response of the two.
- There are two focus diodes available to raster over the LSA. Thus the total time needed is the dwell time per pixel (0.2 seconds is the default) times the number of pixels in the x direction (default is NX=16) times the number of y pixels (default is NY=16), all divided by 2. The maximum permissible dwell time per pixel is 12.75 seconds.
In IMAGE mode you may specify a grating instead of a mirror as the spectrum element (note that this may not be done in Acquisition Mode). Doing so for a target that emits primarily in lines can yield the equivalent of using a slitless spectrograph over a very small portion of the sky (the 1.74 arcsec square region of the LSA). Thus the focus diodes would be swept over the image of the line to produce a picture that is resolved spatially in the y direction and spectroscopically in the x direction. This mode of use would be very slow if all you wanted was the spatial structure of a small object (the FOC would probably be better), but there might be interesting uses for obtaining spectrophotometrically pure, spatially resolved images in the ultraviolet. Please consult us if you wish to explore this option.