|Space Telescope Science Institute|
|COS Instrument Handbook|
In ACQ/SEARCH mode, the telescope is moved in a spiral pattern on the sky to cover a square grid up to 5 × 5 steps in size. At each scan point, the telescope stops and data are collected. A two-dimensional array containing the total counts measured at each dwell point is constructed. After completion of the full n × n pattern, the target position is calculated as described below, and the telescope is moved to center the target. Figure 8.1 illustrates the spiral search patterns that are used for ACQ/SEARCH.Figure 8.1: The Spiral Search Patterns Used for ACQ/SEARCHThis Figure shows the 2 × 2, 3 × 3, 4 × 4, and 5 × 5 spiral search patterns executed with the default STEP-SIZE of 1.767 arcsec. The black circles represent the positions of the 2.5 arcsec-diameter aperture, and the numbers show the sequence of steps on the sky. The large outer circle in red has a radius of ~ 3 arcsec.For an ACQ/SEARCH, the user must specify
• The spectral element (grating or mirror) and the central-wavelength setting (if applicable). For a spectroscopic ACQ/SEARCH these will generally be the grating and central wavelength of the initial science observation. However, an observer may specify ACQ/SEARCH with a different grating and central-wavelength setting if there are advantages to doing so.
• The SCAN-SIZE, which is 2, 3, 4, or 5, corresponding to spiral patterns of 2 × 2, 3 × 3, etc.
• The STEP-SIZE, or spacing between grid points. It may be any value from 0.2 to 2.0 arcsec, but we strongly recommend using the default value of 1.767 arcsec. This value has been chosen so that no part of the sky is missed, given the aperture diameter of 2.5 arcsec (2.5 / = 1.767).
• Once the scan is complete, the flight software computes the centroid of the array and thus the position of the target. There are three centroiding options:
1. The first option is CENTER=FLUX-WT, which uses a flux-weighted centroiding algorithm to determine the center of light. It is the default for SCANSIZE=2.
2. A variation on CENTER=FLUX-WT is CENTER=FLUX-WT-FLR. In this case, a floor is subtracted from the counts at each dwell point before the centroid is computed. The floor is taken as the minimum number of counts seen at any one dwell point. FLUX-WT-FLR has the advantage of removing background counts, but leaves one or more points in the array with zero counts. As it can cause computational problems, FLUX-WT-FLR should not be used with SCAN-SIZE=2. CENTER-FLUX-WT-FLR should be used for SCAN-SIZE > 2 × ACQ/SEARCHs.
3. The last option for centering is CENTER=BRIGHTEST, which simply centers the dwell point with the most counts. This is straightforward, but not as accurate as the other centroiding methods. CENTER=BRIGHTEST is appropriate if coordinates are uncertain and the ACQ/SEARCH is followed by either a second ACQ/SEARCH using flux-weighted centering or an ACQ/IMAGE, or if the source is extended and it is only desired that the brightest point be in the aperture.Table 8.2 presents the recommended ACQ/SEARCH parameters as a function of coordinate uncertainty. For all values of SCAN-SIZE > 2 we recommend CENTER=FLUX-WT-FLR as it is more accurate due to better sky and detector background suppression. The CENTER values in Table 8.2 are the current default values for each SCAN-SIZE. Note that even SCAN-SIZE values (2 or 4) trigger additional overhead because of the telescope motion required to displace the aperture by half of a STEP-SIZE in both the dispersion and cross-dispersion directions, so that the overall pattern remains centered on the initial pointing.Analysis of SMOV/Cycle 17 acquisitions indicate that a single ACQ/SEARCH acquisition, whether spectroscopic or imaging, provides a centering accuracy of 0.3 arcsec only 75% of the time. Additional TA stages are thus necessary to achieve the centering accuracy necessary to meet wavelength and photometric requirements.
σ ≤ 0.4 0.4 < σ < 0.7 0.7 < σ < 1.0 1.0 < σ < 1.3 1.3 < σ ≤ 1.6 This is the default STEP-SIZE value and the largest to cover the search area without holes or gaps.If target coordinate uncertainty is on the lower edge of the given range, the STEP-SIZE may be reduced slightly (e.g., 1.5 arcsec) to improve centering accuracy at the expense of total area covered by the search.