|
|||
| Joint Astronomy Centre |
|
|
|
| Show document only |
SCUBA has three primary observing modes: photometry, jiggle-mapping and scan-mapping. The first two of these can also be used with the polarimeter in place. Below, some common terms are discussed, and then the three observing modes are described. To estimate the observing time required in each mode, please visit the SCUBA Integration Time Calculator. For advice on setting up your MSBs, consult the JCMT observing tool primer and the SCUBA observing tool FAQ pages. The polarimetry options are described in more depth here.
Common Terms
PhotometryPhotometry is the ideal mode to use when an isolated point source with an accuate position is to be observed. The central pixel of each array is used. To account for a ~1 arcsecond offset in the long- and short-wave array centres, the secondary mirror moves so as to observe a small 3×3 grid, with 2 arcsecond spacing, around the source. This also protects the observer against poor pointing. During photometry observations the telescope chops and nods; the chop can be specified to ensure that the "off" position falls perfectly onto another bolometer. This "2-bolometer" chopping mode improves the noise by a factor of approximately 12 percent. Also, a "wide" photometry mode is available which actually observes at 12 positions to make two rings: one 1 arcsecond from the centre and the other 7 arcseconds form the centre. This is a somewhat experimental mode at the moment. The telescope nods once per 9-point integration in standard photometry mode; this means that each integration takes approximately 18 seconds, not including overheads.Multiple photometry data sets can be coadded as necessary in data reduction. Observational overheads for photometry are in practice of the order of 50%; this needs to be added to the calibration overhead to give a total overhead of order 100%. Jiggle-mappingJiggle-mapping is the preferred mode for extended sources that are still smaller than the 2.3 arcminute field of view. This mode gets its name from the "jiggle pattern" that the secondary mirror carries out during the observation. Due to the hexagonal arrangement of the bolometers in each array, simply "staring" at one position leads to undersampling of the sky. To produce a fully-sampled map at both the short and long wavelengths, it is necessary to follow a 64-point pattern with 3-arcsecond spacing (since the beam at 450 microns is around 7 arcseconds). Since the telescope also nods, this leads to a single 64-point integration taking about 128 seconds, not including overheads. To improve the atmospheric subtraction the jiggle is in practice split into 4×16 sets, with a nod between each set. Chopping also takes place, at each individual jiggle position.
It is also possible to use a mode where only a 16-position jiggle-pattern is used - this still provides a full-sampled image at 850 microns but not at 450 microns. However, because of the over-sampling of the 850 micron array in the 64-point jiggle pattern, this mode actually does not provide any sensitivity advantage. It is used however for pointing observations, since we only use the 850 micron data to adjust this. Multiple jiggle-map data sets can be coadded as necessary, including those that are offset but still overlap. Observational overheads for jiggle-mapping are in practice of the order of 20%; this needs to be added to the calibration overhead to give a total overhead of order 70%. Scan-mappingScan-mapping is the most complicated of SCUBA's observing modes but is necessary to image some of the largest structures, such as the Galactic centre. In this mode the telescope scans the array across the sky whilst chopping to produce a differential map of the source. The telescope does not nod during scan-mapping. Our default method is known in-house as "Emerson2" (Emerson 1995, ASP Conf Ser 75, 309). Multiple chop throws in 2 orthogonal directions are used with chop amplitudes chosen so that, except at the origin, the zeroes in the Fourier transform of one do not coincide with the zeroes in the FT of the other up to the spatial frequency limit of the telescope beam. For SCUBA, 6 different chop configurations are typically used: chop throws of 30, 44 and 68 arcsec each with chop position angles of 0 (Dec chopping) and 90 degrees (RA chopping) in a coordinate frame fixed on the sky. In practice this means that the sky must be scanned six times; we typically find that a set of 6 such integrations will take a total time of about 72 minutes (including observational overheads) to cover an area of 10 by 10 arcminutes and reach a noise level of about 35 mJy in good conditions. This assumes the typical slow scanning method, where the telescope scans at a rate of 24 arcseconds per second, with a chop frequency of 7.8 Hz, giving 3 arcsecond sampling along a scan length. A faster scan-map mode, leading to 6 arcsecond sampling, is also available which may be useful if only 850 micron data is required. The maximum possible individual scan length is about 20 arcminutes; since this may correspond to a diagonal scan across the total required area then the actual size of any individual scan map should be smaller than this. We recommend a typical total size of about 700 x 700 arcseconds.Multiple scan-map data sets can be coadded as necessary, including those that are offset but overlap. Observational overheads for scan-mapping are in practice of the order of 50%; this needs to be added to the calibration overhead to give a total overhead of order 100%.
|
