Cosmic Origins Spectrograph Instrument Handbook for Cycle 17

7.5 Imaging Acquisitions

---

Most of the time observers will wish to acquire their target using the COS/NUV configuration in ACQ/IMAGE mode. Ordinarily both COS detectors will be available for use, and so there is little time lost in switching from an NUV acquisition to an FUV spectrum (no more than about 2 minutes; see Chapter 9).

In ACQ/IMAGE mode the following steps occur:

1. The shutter is opened and a target acquisition image is obtained. The telescope is not moved, meaning that an acquisition using ACQ/IMAGE will be successful only if the target lies within the aperture at this point. An area of 4 × 4 arcsec, centered on the aperture, is then read out. This image is recorded and downlinked, and becomes part of the data package that is archived. The plate scale in imaging mode is 23.6 mas per pixel, so 4 arcsec is 170 pixels.
2. A 9 × 9 pixel checkbox array is then passed over the 4 × 4 arcsec image. First, the pixel with the most counts is identified. In the unlikely instance that two pixels have equal counts, the first one encountered is used. The 9 × 9 array centered on that brightest pixel is then analyzed using a flux-weighted centroiding algorithm to calculate the expected target position.
3. Finally, HST is moved to place the calculated centroid at the center of the selected aperture. Another exposure is taken and recorded for later downlink as a verification of the centering. Both of the recorded acquisition images are 345 × 816 pixels, and are taken in TIME-TAG mode.

Note that NUV ACQ/IMAGE acquisitions require two minutes plus twice the exposure time specified for the S/N = 40 acquisition image. If the ACQ/IMAGE exposure is the first in a visit, then an additional five minutes of overhead are required as well. (Note that S/N = 40 is a standard value for HST acquisitions and is based on STIS experience.)

7.5.1 Exposure Times and Count Rates

The best way to determine actual count rates, exposure times, and the overall time needed for an acquisition is to use the COS acquisition ETC and APT. Here we provide less accurate information to give you a general idea of what happens.

Figure 7.3: Exposure Time Needed for ACQ/IMAGE Mode.


 
The time is given as a function of target flux. This calculation assumes a flat source spectrum.
 

Figure 7.3 shows acquisition exposure times needed to reach S/N = 40 for various combinations of mirrors and apertures. A flat source spectrum is assumed.

7.5.2 Imaging Acquisitions with Mediocre Coordinates

If you are less certain of your target coordinate accuracy, or wish to be more conservative, it is possible to scan a larger area of sky, also in undispersed light. The procedure is the same as for an NUV acquisition in dispersed light (see Section 7.7 below), except that ACQ/SEARCH is used with the spectral element selected as MIRRORA or MIRRORB. Use of SCAN-SIZE=3, for example, should be adequate to find the object if it falls within 3 arcsec of the aperture center.

7.5.3 Imaging Acquisitions with MIRRORB or the BOA

Both MIRRORB and the BOA produce images that may affect an acquisition.

As noted elsewhere, what is termed "MIRRORB" is not a separate optical element but is instead MIRRORA oriented so that light is reflected from the order-sorting filter in front of MIRRORA. Because of the finite thickness of that filter, MIRRORB forms a secondary peak which has approximately half the intensity of the primary peak and is displaced by 20 pixels (about 0.5 arcsec) in the y direction. There is some overlap between the wings of the primary and secondary peaks, but they are well enough separated to lead to reliable acquisitions (see Figure 7.4).

Figure 7.4: Cross section through the center of an image obtained with MIRRORB.


 
Note the secondary peak, which has an amplitude of about half that of the primary peak and is displaced by 20 pixels (center to center) in the y direction.
 
Figure 7.5: Cross section through an image obtained with the BOA.


 
This is a slice through the center of the profile. Note the secondary peak, which has an amplitude of 44% relative to the primary peak and is displaced by 10 pixels (center to center) in the x direction.
 

In the case of the BOA, the slight wedge shape of the neutral density filter produces a comatic image with a secondary peak with an intensity 44% that of the main peak, displaced by 10 pixels (Figure 7.5). If both the BOA and MIRRORB are used then a series of 4 peaks is created, displaced in both x and y, although the fourth peak is very faint (Figure 7.6 and Figure 7.7). Note that these three illustrations have been made using images from test observations on the ground. These will be updated once COS is in orbit.

In each of these cases the primary peak in the image is significantly brighter than any secondary peak and is well removed from it, and so we anticipate successful acquisitions. Acquisitions using all of these attenuating options will be tested during SMOV.

Figure 7.6: A cross section in the x direction through the center of the image formed when both the BOA and MIRRORB are used.


 
Figure 7.7: A cross section in the y direction through the center of the image formed when both the BOA and MIRRORB are used.