Spitzer Documentation & Tools
MIPS Instrument Handbook


3.1.3        Photometry - 70 µm Compact Source (Small Field), Default Pixel Scale

Typical photometry observations at 70 µm are made using the nominal 9.8´´ pixel scale.  The super-resolution mode at 70 µm also provides excellent data for photometry, but is considerably more expensive in terms of observing time overhead and is also about 4 times less sensitive.  In the normal photometry observation of a compact (< 1´) target (see Figure 3.3), a pattern of observations is made that is similar in concept to the 24 µm compact source photometry observation.  As a result of the on-orbit realities, we made changes to our pre-launch plans for this mode; it resembles the pre-launch AOT, but shifted to the less noisy side (side A) and with the amplitudes of the offsets correspondingly scaled down in the cross-scan direction.


To start the observation sequence, the source is centered to the left of the array center by 9 pixels (98´´), 1 pixel to the left of the center of side A (the good side), and the scanning mirror is chopped in a symmetrical fashion, above and below the array center to obtain a column of 6 images — the first and last are both at the central position.  The spacecraft is then slewed to position the source 7.5 pixels to the left of center (74´´), 0.5 pixel to the right of the center of side A, and another column of images is obtained.  The resulting approximate image positions are given in Table 3.3, where the coordinates are in pixels and the origin is taken to be the center of the array.  Array distortion and twist have not been taken into account in this table.  In reality, the images will not fall exactly on the positions given in this table, but these positions are good enough for observation planning purposes.



Figure 3.3: Photometry of a compact source with the 70 µm array.  The positions of the target in each of the 12 frames obtained in the first cycle of the standard photometry observation at 70 µm are shown by the circles, which are the size of the PSF at the FWHM.  The entire 32x32 array is shown with pixels represented schematically by the grid.  The dither pattern involves half-integer pixel offsets in both directions.  Repeat cycles of the photometry AOT will provide just images 2-6 and 8-12, not all 12 shown here.

A stimulator flash frame follows the first and the sixth exposures in the sequence.  A telescope nod then moves the target (see the figure) and the pattern of frames and stim flashes is repeated, completing a single cycle of the AOT and producing 10 target images, with an extra pair of images on the first cycle (and perhaps a few later cycles as well).  The observer specifies the number of times to repeat the complete cycle; if multiple cycles are requested, several sets of frames may be taken at one nod position before switching to the other position.  The observer has no control over the relative placement of the stim flashes or the frames.  The relative positions of these images are chosen to provide 1/2 pixel sampling of the PSF in both the scan and cross-scan directions.  The positioning of the target on the array is illustrated in Figure 3.3 and Figure 3.4; in the first of these figures, the circles show the three consecutive pairs of positions for the image with their diameters being the FWHM of the Airy disk.



Figure 3.4: Simulated source detections on the array during the compact source photometry AOT in Figure 3.3.  Time runs across the top row from left to right, then across the second row from left to right, and so forth.  The observation starts with a source frame and stimulator flash, then several scan-mirror dithered frames.  The sequence is completed with another stimulator flash.  The spacecraft is then offset, and the above sequence of dithers and images is repeated.  In repeat cycles of the photometry AOT frames 1 and 7 are omitted, so the AOT basically provides 10 source images per cycle, not all 12 shown here.  Note that the source stays on side A of the array.

These photometry data provide multiple independent images with sampling of various positions on the pixels and a level of oversampling that is useful in extracting diffraction-limited images.  Note that careful combination is required to gain all the benefits of this sampling, although for simple photometry, integer pixel shifts are adequate.  Furthermore, the scan mirror motion is not perfectly parallel to the columns of the array.  Nonetheless, the pattern provides a redundant level of oversampling that is useful in extracting diffraction-limited images.

Table 3.3: Source positions for 70 μm default pixel scale compact source photometry (in units of pixels, the origin being the center of the array).

Frame # X offset Y offset
1 -9.0 0.0
2 -9.0 -7.4
3 -9.0 3.7
4 -9.0 -3.7
5 -9.0 7.4
6 -9.0 0.0
7 -7.5 0.0
8 -7.5 -7.4
9 -7.5 3.7
10 -7.5 -3.7
11 -7.5 7.4
12 -7.5 0.0


Given the characteristics of the MIPS detectors, it is best to build up long integrations with long exposure times, e.g., large total integration times built from 3 s exposures will not produce as high-quality a final product as fewer cycles of 10 s exposures.