A variation on this procedure is to get the field image two months or so in advance of the time the spectrum will be obtained. This used to be called an early acquisition (EARLY ACQ) but is now handled as a separate Visit. An early acquisition takes more time than an on-board acquisition but, unlike an interactive acquisition, it imposes no burden of real-time contact. The observer must be prepared to analyze the early acquisition image quickly (within a week or two) if the positions from it are to be incorporated into the telescope schedule. Again, we have found that early acquisitions have not been needed in practice in Cycles 4 and 5. Moreover, the shortness of Cycle 6 (see Section on page 11) makes early acquisitions especially impractical.
If you wish to use either WFPC2 or FOC for early- or interactive acquisitions you must refer to documents specific to those instruments. Details on the use of the GHRS' imaging capability are provided in the next chapter.
However, in working with bright stars you need to carefully consider the count rate that will occur for two reasons. First, if the count rate is high and the integration time is longer than about 0.6 sec, you can "wrap" the counters. The image of the star is then an annulus and there is no bright core to center on. As long as the STEP-TIME is 0.6 sec or less this should not happen. (The maximum count rate is about 100,000 per second and the counter will hold ~64,000 counts.) Even if "wrapping" does not occur, an acquisition for a very bright star may not work well if the count rate is extremely high. This is because the counters saturate, leading to a Point Spread Function (PSF) with less contrast than for a properly imaged one. In an extreme case the center of the stellar image can be flattened, leading to the wrong "center" being placed into the aperture. Chapter 7 provides the information you need to determine the acquisition count rate.
But acquisitions with Side 1 of the GHRS can be problematic. In particular, grating G140L can obtain spectra of objects too faint to be acquired with MIRROR-N1. If you find that you cannot get 100 or more counts in the peak pixel with N1, you have two options. First, you can use MIRROR-N2 on Side 2, which has a much broader response. The penalty for doing this is a delay of about 40 minutes in switching from Side 2 to Side 1. Your other option is to acquire your object with the FOS and then ask for your ACCUM in the LSA. An FOS-assisted acquisition will not work for observing in the SSA, but there is no significant time involved in moving the target from the FOS acquisition aperture to the LSA. Note that FOS-assisted acquisitions should be done on the blue side of the FOS because the FOS red side apertures are more distant from the GHRS and it is unlikely to be possible to use the same guide stars throughout. For more information on FOS-assisted acquisitions, see Section on page 45.
The Point Spread Function (PSF) of the GHRS has been restored by the COSTAR mirrors, making it possible to separately observe stars that are very close together. For example, in an Early Release Observation in April, 1994, two stars in R136a separated by only 0.25 arcsec were observed independently. One of these stars was only 0.1 arcsec from a brighter neighbor. This was done by centering on a bright object in the field and then offsetting to the targets of interest. We suggest that you consult with us if you wish to work in crowded fields. Also, see Section on page 44 and Section on page 46.
Very Bright Stars
When is a star too bright for an ONBOARD ACQuisition? In practice we are unaware of any real need to use interactive acquisition just because a star is very bright. GHRS acquisitions are done with ultraviolet light, so it is the UV flux of the star that matters. There are no stars too bright to acquire with the attenuated mirror A1, for example. Even with Side 2 it is not necessary to specify an interactive acquisition for a very bright star if the BRIGHT=RETURN option is used. Since the few very bright stars which could cause problems are always the brightest point sources in their immediate area, there is no apparent reason not to use BRIGHT=RETURN with an ONBOARD ACQuisition. Faint Objects
You can get a spectrum for as long as you wish with the GHRS, but the maximum permissible integration time per dwell point (STEP-TIME) for a GHRS acquisition is 12.75 seconds. As long as the brightest point in the search area has 1,000 counts or more within the STEP-TIME you choose, your acquisition should work fine, and there is little risk of failure even if you only have ~100 counts in the brightest pixel. If you are using Side 2 of the GHRS (G160M, G200M, G270M, or Echelle-B), there are probably no objects you cannot acquire with a STEP-TIME of 12.75 seconds or less if it is possible to also get a spectrum. Extended Objects
For most extended objects, it may be possible to offset from a nearby point source or at the least the pointing can be specified from an early acquisition image. Interactive acquisitions should be necessary only rarely. Another method is to acquire the object with the FOS and then offset to the GHRS' LSA. You may also use a centering option designed expressly for extended objects, especially uniform ones like the Galilean satellites of Jupiter. Chapter 4 contains more information on this option, known as LOCATE = EXTENDED. Variable Objects
Objects whose ultraviolet brightness varied often caused acquisitions to fail when it was necessary to specify both BRIGHT and FAINT count limits. The advent of software that automatically finds the brightest object in the field (BRIGHT=RETURN) obviates that problem in most cases. A variable object in a crowded field might benefit from an early acquisition to determine precise coordinates, but an interactive acquisition should generally be unnecessary. Moving Targets
Sophisticated pointing at moving objects (i.e., objects within the solar system) may require interactive acquisition to be sure the desired portion of the object's surface is centered in the observing aperture. There are cases where an ONBOARD ACQuisition will suffice, especially if the object is small (essentially point-like) and has a well-determined orbit. An ONBOARD ACQuisition can often work well even for a large moving object like Jupiter by first centering on a small object nearby whose relative position is well known (one of Jupiter's moons, for example), and then offsetting to the position of interest on the planetary disk. Solar system astronomers may wish to consult with a moving-target specialist at STScI before specifying the acquisition mode. Crowded Fields
Work in crowded fields can usually be done by obtaining an early acquisition image, so that you have something to work from to specify the object to be observed, an image that has been obtained with HST's full spatial resolution. The camera observation is usually best done at about the same wavelength that the spectroscopic observations will be made.