Space Telescope Science Institute
Table of Contents Previous Next Index Print

Near Infrared Camera and Multi-Object Spectrometer Instrument Handbook for Cycle 17 > Appendix D: Techniques for Dithering, Background Measurement andMapping > D.3 Chop and Dither Patterns

D.3 Chop and Dither Patterns
There are a set of fifteen pre-designed patterns available for NICMOS observations. Users may define their own pattern specifications as well in APT, during Phase II development. The pre-defined patterns include four dither patterns, four chop patterns, five dither-chop patterns, and two map patterns. For each of these, the observer will be able to specify the number of positions desired (1 to 50), the dither size (0 to 40 arcsec), the chop size (0 to 1440 arcsec, also used for mapping), and the orientation of the pattern with respect to either the detector or the sky. The POS-TARG special requirement is still available for offsetting the telescope and creating custom-design patterns as well, but there are a number of advantages to using the pre-designed patterns:
All the observations pertaining to an exposure specification line in a pattern result in one association and are simultaneously calibrated and combined in the data calibration pipeline, including background calibration, cosmic ray removal, and flat fielding. Observations obtained with POS-TARG do not result in associations, and will have to be combined manually by the observer.
Patterns permit the observation of a region on the sky with a fixed position angle without fixing spacecraft roll, which increases the number of opportunities to schedule the observations.
Multiple exposures may be obtained at each position by the use of the Phase II exposure level parameter for Iterations. This may be useful for cosmic ray removal. In addition, exposures in different filters at each pattern position can be obtained by linking together exposure lines as a pattern group.
The fifteen NICMOS pre-designed patterns are listed in Table D.1, together with applicable parameters, such as the allowed values for the number of steps in the pattern, the dither size, or the chop size. In addition, the figure number where the pattern is graphically shown is given in the last column of Table D.1. Offset sizes and number of steps in a pattern affect the amount of overhead time required to perform an observation (see Chapter 10). The effects of dithering or chopping on an astronomical image are shown in a set of examples in the next section.
of Dithers
Orient frame
Note on Orientation:
The pattern parameter syntax requires additional input on orientation. Specifically, the pattern must be defined in either the POS-TARG (camera) frame or the CELESTIAL (sky) frame. Dithering to remove detector characteristics should always be performed in the POS-TARG frame of reference. A pattern orientation angle may be specified as well. In the POS-TARG frame, this is the angle of the motion of the target from the first point of the pattern to the second, counterclockwise from the x detector axis (the directions are defined in Figure 6.1). In the CELESTIAL frame, the angle is measured from North through East.
Note that some of the pattern names in Table D.1 are doubled except for an additional -SKY-. The chop can be specified either as POS-TARG or CELESTIAL (default—see below for details).
Move the sky or the telescope?
The pattern syntax attempts to resolve the confusing dichotomy in the old pattern implementation, as to whether the pattern moves the telescope or the target. It does this by providing the two reference frames described above.
Patterns done in the POS-TARG reference frame will move the target, just as the “POS-TARG” special requirement does.
The telescope is slewed in small angle maneuvers (SAMs) so that the target moves within the detector frame of reference as specified by the pattern. When NICMOS images are displayed with IRAF, the POS-TARG x- and y-axis are as shown in Figure 6.1.
Patterns done in the CELESTIAL frame will move the telescope relative to the sky reference frame.
The target will always move in the opposite direction on the detector to the motion of the telescope.
The dither patterns are recommended for measuring the background adjacent to point sources (longward of 1.7 microns), and for the reduction of sensitivity variations and bad pixel effects. The four types of canned dither routines are NIC-XSTRIP-DITH, NIC-YSTRIP-DITH, NIC-SPIRAL-DITH, and NIC-SQUARE-WAVE-DITH. Most of the names are self-explanatory: the NIC-SPIRAL-DITH pattern produces a spiral around the first pointing; the NIC-SQUARE-WAVE-DITH pattern covers extended regions by moving along a square-wave shape; the NIC-XSTRIP-DITH and NIC-YSTRIP-DITH patterns move the target along the x and y directions of the detector, respectively. The difference between the NIC-XSTRIP-DITH and the NIC-YSTRIP-DITH patterns is that the first moves by default along the grism dispersion (orient default = 0), while the second moves orthogonal to the grism dispersion axis (orient default = 90). These patterns are illustrated in Figure D.3, and the direction of the x- and y-axis are the same as in Figure 6.1.
Note that there is an additional parameter for dithering patterns, to center the pattern on the target. The default is to start the dithering at the target position.
Figure D.3: Dither Patterns. Numbers represent sequence of positions where the target will be on the detector.
The chop patterns are recommended for measuring the background adjacent to extended targets. For each chop pattern, half of the exposures are taken on the target (position 1). There are two basic patterns, NIC-ONE-CHOP and NIC-TWO-CHOP. The NIC-ONE-CHOP pattern produces one image of the target and one image of the background. The NIC-TWO-CHOP pattern produces two images of the target and two background images, with the background fields positioned on opposite sides of the target. These patterns may be repeated by specifying the number of points in the primary pattern. For example, calling the NIC-TWO-CHOP pattern in an exposure with number of Iterations = 1 will produce four images, one on the target, one off to one side (+x detector direction), one back on the target, and one off to the other side (–x detector direction). If the number of Iterations = 2, the observer gets eight images, two images at each position of the pattern. If the primary pattern has number of Points = 2, the pattern will repeat (1,2,3,4,1,2,3,4), and the observer will get eight images. Chop patterns are illustrated in Figure D.4, and the direction of the x- and y-axis are the same as in Figure 6.1.
Because chopping is best done to empty regions of the sky, we provide a set of chopping patterns that are in the CELESTIAL coordinate system, as well as the standard set (that are in the POS-TARG frame). These have the word SKY in their name, and must have a pattern orientation angle (degrees E from N for the first motion of the pattern) supplied. These should be used when the region around the target contains some objects that should be avoided when measuring the background. SKY patterns are illustrated in Figure D.6, and the direction of the x- and y-axis are the same as in Figure 6.1.
Figure D.4: Chop Patterns.
The combined patterns permit dithering interleaved with chops to measure the background. They are recommended for simultaneous minimization of detector artifacts and background subtraction, for observations beyond 1.7 microns. Three types of combined patterns are implemented: NIC-SPIRAL-DITH-CHOP, NIC-XSTRIP-DITH-CHOP, and NIC-YSTRIP-DITH-CHOP. Their characteristics are analogous to the dither patterns NIC-SPIRAL-DITH, NIC-XSTRIP-DITH, and NIC-YSTRIP-DITH, respectively, with the addition that each dither step is coupled with a background image obtained by chopping. These combined patterns are shown in Figure D.5, and the direction of the x- and y-axis are the same as in Figure 6.1.
In a manner similar to the regular chop patterns, the combined patterns have “SKY” versions implemented in the CELESTIAL frame. The chop patterns require an pattern orientation angle which is defaulted to 0.0 (North). The angle is measured from North through East. These are illustrated in Figure D.6.
Figure D.5: Combined Patterns.
There are two MAP sequences. These allow the telescope to be pointed at a regular grid of points, doing a series of exposures at each point. These are done in the CELESTIAL frame, so a pattern orientation angle must be supplied, and the telescope motion on the sky is specified (rather than the target motion relative to the detector, see note above). The NIC-SPIRAL-MAP sequence is basically the NIC-SPIRAL-DITH sequence in the CELESTIAL frame, and automatically maps the (square or rectangular) region around the target. The NIC-MAP sequence defines an arbitrary parallelogram on the sky. The observer may specify the number of points in each of two directions, and the position angle (E of N) of each direction.
As with the dithering patterns, the observer has the option of specifying whether the target is centered in the pattern or not. The target will be centered in the NIC-SPIRAL-MAP pattern if there are 9, 25, 49,... points in the pattern, but will not necessarily be centered otherwise. The observer can specify if the target should be centered along one axis or the other, or both, of the parallelogram defined by the sequence. These are illustrated in Figure D.6.
Figure D.6: Patterns on the sky. Numbers represent sequence of aperture pointings on the sky.
On occasion, it may be advantageous to specify a POS-TARG on the exposure line to move the target to a different position than the aperture reference point. For this situation, the POS-TARG offset is always performed first to change the telescope pointing. For example, a user wants to position a target in each of the four quadrants in Camera 2. The user specifies the NIC2-FIX aperture for which the aperture reference point is at the center of the array (128128 pixels) and specifies a POS-TARG –4.8,–4.8. A four point dither pattern using the NIC-SPIRAL-DITH pattern with point spacing = 9.6 arcseconds and pattern orient = 0 would achieve the desired results. (See the following example.) The target will be in the lower left quadrant of the array for the first position of the pattern, the lower right for the second position, the upper right for the third position, and in the upper left quadrant for the fourth position of the pattern.
Predefined convenience patterns are recommended for NICMOS observations. These predefined patterns can be selected using the APT pattern editor. Observers can specify their own pattern by using a generic pattern form. Patterns are not supported by calnicb which combines dithered observations into a mosaic. The IRAF/STSDAS task drizzle and MultiDrizzle (Koekemoer et al., 2002 HST Calibration Workshop, p337) can also be used to combine images into a mosaic. See also the new Multidrizzle Handbook (Fruchter et al 2009) for examples on how to combine NICMOS data with Multidrizzle.

Table of Contents Previous Next Index Print