Acquisition Mode

In this mode the target is first acquired in the Coronographic Acquisition Aperture (see Figure 5.1). The ACQ mode image is now obtained, with an integration time specified by the observer, which should be long enough to determine the centroid of the target image accurately. An ACQ mode exposure actually results in two images. Each image is analogous to an ACCUM mode image (i.e., it has a single reset and read before and after each integration period). The two images are flat fielded and corrected for the "shading" effect (see Chapter 3) using the on-board NICMOS computer, and then compared in order to filter out any CR hits. The brightest object common to both images and located inside the coronographic acquisition aperture is found. This object is assumed to be the desired target. A simple (and of undemonstrated reliability for anything other than a point or circular source) algorithm is now used to determine the position of the centroid of this source, and the telescope is moved such that this position is centered in the coronographic hole. The expected accuracy of this procedure for point sources is better than one quarter of a Camera 2 pixel (i.e., 0.019 arcsec), assuming all cosmic rays are successfully removed on-board. The instrument is now ready to acquire more data, and the ACQ mode procedure finished.

Problems to beware of in the ACQ process include cosmic rays and coronograph centering accuracy. It is expected, as mentioned above, that the target should be centered behind the hole with an accuracy of better than one quarter of a pixel. How much better remains to be seen. Observers should note that in order to compare library PSFs with coronographic images for the purposes of determining source extensions or morphology, an accuracy of one quarter of a pixel may not be sufficient at long wavelengths. The two separate images obtained in ACQ mode will filter out most cosmic rays in short exposures, but if an ACQ mode exposure longer than about 10 minutes (i.e., 5 minutes for each of the CRSPLIT images) is required, observers should seriously consider a reuse target offset or INT-ACQ observation, because the probability of a CR hit not being filtered out on-board becomes larger the longer the exposure. In this case it is likely that the acquisition procedure will fail because the brightest source will be a CR hit, and the intended target will not be hidden behind the coronograph during the subsequent integrations. If the target is bright enough to saturate the detector during the ACQ mode exposure, this is likely to cause poor centering. A failed or unsatisfactory acquisition will probably cause the subsequently acquired images to be useless to the observer. Our initial on-orbit experience suggests that an onboard acquisition is unlikely to succeed for targets with an H magnitude brighter than roughly 4, but this number will be refined with further experience.

The coronographic mode acquisition capability was being characterized during the period this Handbook was revised. An update on its performance will be placed on the STScI NICMOS WWW page on 1 August 1997.