C.2 WFC3 Patterns

A number of different types of patterns are available to support dithered and mosaicked WFC3 observations. The pre-defined patterns that have been implemented in APT are described in the Phase II Proposal Instructions, which are updated when the selection of a new cycle of proposals is announced. The pre-defined WFC3 patterns in effect in APT at the time of publication of this handbook are summarized here. Patterns not predefined in APT can still be executed by specifying POS TARGs on exposures in the Phase II proposal. WFC3 ISR 2016-14 and WFC3 ISR 2020-07 tabulate the necessary POS TARGs for compact patterns with up to 9 steps for WFC3/IR and WFC3/UVIS, respectively. WFC3 ISR 2023-05 tabulates the necessary POS TARGs for optimal patterns when using WFC3 and ACS simultaneously (i.e. one instrument as prime and the other in parallel). Those patterns are designed to preserve sub-pixel sampling as much as possible over the face of the detector, given the scale changes introduced by geometric distortion. Alternatively, observers can also develop custom Patterns within APT by starting with a Generic pattern and setting the required parameters as desired or by adjusting specific parameters within a pre-defined pattern. 

Pre-defined WFC3 dither patterns designed to subsample pixels can optionally be selected as secondary patterns when WFC3 patterns with larger steps are selected as primary patterns. WFC3 patterns can also be added as secondary patterns to any of the generic pattern types (BOX, LINE, SPIRAL). When combining patterns, the smaller dither pattern should be the secondary pattern to minimize the time spent moving the telescope. Due to geometric distortion (Appendix B), a large mosaic step shifts some objects by an integer number of rows (or columns), and others by an integer plus some fraction of a pixel. The PSF is thus not evenly sampled in the overlap region of the two exposures, so a PSF-sampling dither should be added if spatial resolution is important.

Sets of exposures with offsets executed using patterns or POS TARGs are associated and combined automatically during pipeline processing, as long as the same guide stars have been used for all exposures. Pointings must be contained within a diameter ~130 arcsec or less (depending on the availability of guide stars in the region) to use the same guide stars. Note that the rms pointing repeatability is significantly less accurate if the same guide stars are not used for all exposures (Appendix B of the DrizzlePac Handbook).

The names and purposes of the patterns in effect in APT at the time of publication are given in Table C.1. Note that the initially-adopted names of patterns have been preserved for continuity, although they do not always correspond to the distinction between dither steps and mosaic steps outlined above. The small BOX dither patterns are designed to optimally sample the PSF when 4 steps are used. Since time constraints do not always permit visits to be broken into multiples of 4 steps, LINE dither patterns that optimally sample the PSF in 2 or 3 steps are also given. The BOX and LINE dither patterns are illustrated in WFC3 ISR 2010-09. A full discussion and illustrations of patterns that optimally sample the PSF for different numbers of steps are available in Section C.2 of the DrizzlePac Handbook. Note that PSF sampling generally produces a more significant improvement for IR images than for UVIS images (Section 6.6.1 and Section 7.6.1). The remainder of the patterns in Table C.1 are special-purpose mosaic patterns that are expected to be commonly needed. There are no pre-defined patterns to deal with specific features in flats—notably, the circular dead spot on the IR detector (WFC3 ISR 2008-08) and the UVIS “droplets” (WFC3 ISR 2008-10); however, patterns that can be used to mitigate the effects of these artifacts are discussed in WFC3 ISR 2010-09.

Table C.1: Dithering and Mosaicking Patterns in APT for WFC3.

The default specifications of the patterns are summarized in Table C.2. The equivalent POS TARG moves are summarized in Table C.3, along with the approximate number of pixels corresponding to these moves. The number of pixels was computed using only the linear distortion terms with coefficients measured at the center of each detector. This is an excellent approximation for small moves and for objects that remain in the central region of the detector (Figure B.1 and B.3 in Appendix B).

Note that you can easily scale up the patterns in APT to make them larger; e.g., multiply the Point Spacing and Line Spacing of patterns with half-pixel sampling (WFC3-IR-DITHER-BOX-MIN, WFC3-IR-DITHER-LINE, WFC3-UVIS-DITHER-BOX, WFC3-UVIS-DITHER-LINE) by an odd number to preserve the half-pixel sampling (this is equivalent to multiplying the POS TARGs and steps in pixels by that number). You may want to do this in the IR, for example, to achieve the recommended spacing of at least 10 pixels between positions for photometric repeatability (WFC3 ISR 2019-07) or to move a saturated persistence-generating core of a target by a greater distance than the minimal default distance.

As an example, for the WFC3-IR-DITHER-LINE, one would scale the 0.636" spacing x3 to preserve half-pixel sampling, making this ~15 pixels along the diagonal instead of the standard ~5 pixels.  This will help minimize the effect of self-persistence from bright stars taken in prior exposures from the same visit.

Be aware, however, that larger steps result in larger variations in sub-pixel sampling of the PSF over the face of the detector due to non-linear geometric distortion (WFC3 ISR 2016-14).

Table C.2: Default values of the parameters that define the WFC3 APT convenience patterns

Table C.3: Steps in arcsec in the POS TARG frame and in detector pixels for the WFC3 APT convenience patterns.

For the IR detector, the linear relation between POS TARGs and pixels is simply

POS TARG X = a₁₁ * x
POS TARG Y = b₁₀ * y

where a₁₁ ~ 0.1355 arcsec/pixel and b₁₀ ~ 0.1211 arcsec/pixel near the center of the detector. For the UVIS detector, there is a cross-term that takes into account the fact that the projected axes are not perpendicular:

POS TARG X = a₁₁ * x
POS TARG Y = b₁₁ * x + b₁₀ * y

where a₁₁ ~ 0.0396 arcsec/pixel, b₁₁ ~ 0.0027 arcsec/pixel, and b₁₀ ~ 0.0395 arcsec/ pixel near the center of the detector. This relationship is illustrated in Figure C.1.

The values of these coefficients were derived using optical models and apply to the centers of the detectors. On-orbit geometric distortion solutions give marginally different coefficients (WFC3 ISR 2010-09). The corresponding changes in pixel steps in small dithers are insignificant. The corresponding changes in pixel steps in large dithers or mosaic steps are inconsequential, since non-linear distortion makes the step size in pixels variable over the detector.