|
|
MIPS : Examples |
|
Here we present information relevant to filling out Astronomical Observing
Templates (AOTs) for basic MIPS observations. For more complete
information, consult the Spitzer
Observer's Manual. For more examples, see the Observation Planning Cookbook.
Photometry AOTAssume a photometry observation in the three MIPS bands is desired to determine the far- infrared colors of a point source, or one that is only slightly extended (<2arcmin at 24 and 70 microns, <1.5arcmin at 160 microns), such as a circumstellar disk source. From the observer-predicted flux estimate for the three bands and the MIPS sensitivity charts, it is determined that 200 seconds of on-source integration time is needed at 24 microns, and 100 seconds is required at 70 and 160 microns. This will require two cycles of the basic dither pattern at 24 microns (278 seconds total on-source integration time for the 10 second exposure time), one dither-pattern cycle at 70 microns at the default pixel scale (120 seconds on-source at the 10 second exposure time), and five dither-pattern cycles at 160 microns (giving 100 seconds integration for the 10 second exposure time). The figure shows an example of the AOT frontend for specifying this observation in the MIPS Photometry/Super Resolution section of Spot.Observers are reminded that two cycles of the 160 micron small field photometry AOT provide only a minimum level of data redundancy (4 complete images). Spot will alert you if you have fewer than the recommended number of cycles. For very deep integrations, observers may encounter the photometry/super resolution cycle number limit (20 cycles) before the maximum AOR time (3 hrs) is reached. For such observations, it is recommended that the observer reposition the object or field on the array and repeat cycles until the desired integration time is reached. This can be most easily accomplished by creating cluster offset targets or by doing small raster maps; for example, if a total of 30 cycles is desired, create a cluster target with two positions close together (or do a small raster map) and ask for 15 cycles at each cluster position.
Super-Resolution AOTThere are two example Spot input figures for this example: AOT front end and 160 micron raster map definition dialog box. Note that "columns" are in the cross-scan direction, and "rows" are in the scan direction.To obtain data that can be post-processed for super resolution requires a somewhat different observing strategy from normal photometry observations. The 24 micron pixel sampling provides proper sampling of the PSF using the standard photometry dither sequence, so no modifications are necessary. For super-resolution at 70 microns, the fine pixel scale must be selected, while at 160 microns, a 3x1 raster map should be selected, (see the first Figure above) using 1/8th array stepping in the cross-scan direction, to provide the needed pixel sampling. If, for example, the observer estimates that 600 sec of on-source integration time in 10 sec exposures is required at 70 microns in the fine scale, then at least eight cycles of the standard fine scale observing sequence will be required. If, as above, 100 seconds is required at 160 microns, three cycles of the standard sequence at 10 second exposure times will provide about 180 seconds per pixel on-source integration time in the 3x1 map sequence. Note that by selecting "Calc. Obs. Time" in the AOT window, Spot shows the integration time per pixel for the 160 microns array as about 60 seconds; at this point, Spot does not know where the source is on the array. Since the source is moved 3 times on the array by only 1/8th array steps each time in this example, the total integration time per pixel on source is about 180 sec.
Large-Field Photometry and Raster MappingThere are two example Spot input figures for this example: AOT front end and 24 micron raster map definition dialog box. Note that "columns" are in the cross-scan direction, and "rows" are in the scan direction.Suppose an observer wishes to image a face-on spiral galaxy that is about 2arcmin in diameter, and there is fairly uniform sky around it. The standard photometry dither patterns are designed to image sources less than about 1arcmin in size. The large field patterns are suitable for sources less than about 2 arcmin in size. The large field dither patterns (described in section 8.2.1) assemble images centered on the source coordinates and acquire background data near the source in the in-scan direction. For 24 microns, the background data are acquired with a telescope slew to an observer-defined distance from the source. At 70 and 160 microns, the sky data are acquired through scan mirror motions. Specifying a large field source is no different from standard point source photometry, except that a background chop distance for 24 microns needs to be specified (300 arcsec is the default), and the large field option selected. For regions with complex background, or sources larger than about 2 arcmin, the photometry raster map option can be selected. This option is only available with the small field dither patterns and the default 70 micron pixel scale (10 arcsec). Stepping is done in array coordinates only, in the cross-scan direction and the in-scan direction. Steps in the in-scan direction at 160 microns are defined in fractions of that detector s effective field of view (2x5 arcmin), which is filled through multiple steps of the scan mirror. All other step sizes are defined as fractional array size steps. For regions with complex background, or sources larger than about 4 arcminutes, the photometry raster map option can be selected. This option is only available with the small field dither patterns and the default 70 microns pixel scale (10"). Stepping is done in array coordinates only, in the cross-scan direction and the in-scan direction. Steps in the in-scan direction at 160 microns are defined in fractions of that detector's effective field of view (2' x 5'), which is filled through multiple steps of the scan mirror. All other step sizes are defined as fractional array size steps. Because the small field dither patterns move the pointed position over a relatively large area of the arrays, a MIPS raster map will have large overlaps of exposures. This overlap is required for proper frame-to-frame calibration of these far-infrared Ge:Ga images, especially when complex background structures are present in the images. Because the small field dither patterns move the pointed position over a relatively large area of the arrays, a MIPS raster map will have large overlaps of exposures. This overlap is required for proper frame-to-frame calibration of these far-infrared Ge:Ga images, especially when complex background structures are present in the images. There is no simple formula for determining whether a large region of sky can be observed more efficiently using a raster map or a scan map. Scanning is very efficient for covering large regions on the sky in all three MIPS bands, but does not integrate especially deeply. For deep observations of regions 2 to 10 arcmin in size, it will typically be more efficient to use the raster option. The observer should compare the times required by example inputs to Spot in order to determine the most efficient observing mode. Other considerations should also be taken into account, such as data quality, depth, and full coverage at 160 microns.
Scan MappingMIPS scan mapping is useful for efficiently imaging large areas of sky in all three MIPS bands. The design of a MIPS scan map will be influenced by several factors; the larger the scan map, the more stringent the constraints resulting from these factors will be. The primary factors to consider are: 1) required integration time (i.e., scan rate), 2) scan leg size, and 3) scan leg orientation. The basic inputs for a MIPS scan map (scan rate, scan leg length, and scan leg offset, each selected from a list in the Scan Map AOT front-end), directly address two of these. The scan leg offsets in the cross-scan direction determine the amount of overlap between scan legs, and can be different for a change from the initial, forward scan direction, than for a change from the reverse to the forward direction.Definitions: A "scan leg" is defined to be in one direction only. An "up and back" scan is made of two scan legs. The scan leg length as input to Spot is the "full coverage" length in all three MIPS bands. The actual distance over which a scan is made is approximately 21 arcminutes longer than the full coverage scan length. This "overscan" area does not provide co-spatial coverage in all three bands, but otherwise overscan data will be of the same quality as in the rest of the scan map. The target coordinates are defined to be the map center location. The cluster option is not available for MIPS scan mapping. The orientation of the scan legs on the sky needs to be taken into account in designing the map. Simple map no. 1: The Figure shows the Spot configuration for a basic medium scan rate, 1deg x 0.5deg map, with 160" (half-array) cross-scan steps. The total time required to obtain this map is very nearly 3 hours, so only one map cycle can be done per AOR. For this configuration, the 160 micron scans will have minimal overlap and data redundancy. It is therefore recommended, for good quality 160 micron data, that this AOR be repeated at least once. Inherent multiple redundancy at 24 and 70 microns removes that need if only these bands are of interest. However, detection of asteroids in these data will require at least a second map obtained at a later time. Simple map no. 2: The Figure shows the Spot inputs to define a 1deg x 0.5deg map using the fast rate. The 160 micron data will contain coverage gaps, but this AOR design fills the gaps because the small cross-scan step for the return legs of the map brings the second leg back along nearly the same path. A full pixel sized offset is made automatically to achieve a fully sampled scan path. The minimum scan leg length for the fast rate is 1 degree. Simple map no. 3: This is a narrow, multi-scan strip. A scan map might provide a very effective and efficient way of acquiring multi-band data on a long, narrow strip of sky. An example might be repeatedly scanning in a narrow strip to build up the desired integration time. An example of a scan strip 5 deg long rescanned 8 times is shown in the figure. This particular map design will fully sample the 160 microns field four times. A quarter-array scan offset is made for enhanced sampling and protection against bad pixels.
Spectral Energy DistributionThe observing sequence for MIPS SED is simple and straight-forward. Consider a relatively small target (<1arcmin) source. In this case, a chop distance of 1arcmin will reach sufficiently far off-source to sample the sky. In the figure, we have selected this chop distance. A ten-second exposure, repeated in 4 cycles, yields on-source integration time of about 250 seconds. Note that in a single cycle, spectra are taken at two positions in the slit; 4 cycles repeats this sequence 4 times, with 4 spectra taken at each position. A simple raster map option similar to that for Photometry/Super Resolution is available as well
Total PowerThe observing sequence for MIPS TP is also simple and straight-forward. Here is a screen grab from Spot of the AOT front-end.For more examples, see the Observation Planning Cookbook.
|
help@spitzer.caltech.edu http://ssc.spitzer.caltech.edu/mips/examples.html
