SCUBA: The Final Delivery Date

So what is SCUBA anyway? For those readers new to the JCMT or whose head has been firmly buried in the sand for the past few years it is a submillimetre camera and photometer. It has 2 arrays of detectors, one of 37 pixels optimised for 850 micron and one of 91 pixels optimised for 450 micron. These 2 arrays look out simultaneously at the same area of sky with a field-of-view of approximately 2.3 arcminutes. Each pixel is diffraction limited, which corresponds to a resolution of 7 arcseconds at 450 micron and 14 arcseconds at 850 micron. There is a filter mechanism which means that the 850 micron array can also be used at 750 micron or 600 micron and the 450 micron array can be used at 350 micron, with slightly less than optimised sensitivity and resolution. In addition to the arrays there are 3 separate pixels individually optimised for 1100, 1400 and 2000 micron. These pixels look out simultaneously but are offset from each other on the sky. All the detectors in SCUBA will be able to achieve background photon-noise limited sensitivity (it achieves this by cooling the detectors to 0.1K). To enable accurate calibration of data, SCUBA also has a sky transmission calibration system and an internal calibrator to remove variations in detector sensitivity.

Laboratory results: The reason we are all now so confident is that the recent laboratory results are so impressive. Just before Christmas 1995, 93% of the 131 pixels were operating to specification, and have been shown to remain so even with vibration levels on the cryostat 10 times that measured on the Nasmyth platform. Figure 1 shows the noise spectrum of one of the SCUBA 850 micron detectors under three different background conditions, as measured through the SCUBA electronics system. The lowest curve shows the noise measured with the filter drum closed, i.e. system noise, the highest curve shows the noise looking out at the lab, i.e. photon noise from a 300K black-body, and the middle curve shows noise measured with a reflector at the cryostat window so the detectors are looking back at a temperature of approximately 50K, very close to the effective temperature of the sky at Mauna Kea at 850 micron on a good night. The clear separation of these three curves shows that SCUBA is background-limited. The noise performance has also been shown to be completely stable over a 10 hour period. The optical performance has been shown to be very satisfactory. Figure 2 shows a map of a point source with the LW array with contours at 1-10% of the peak.

Figure 1: Noise spectrum from one of the 850 micron detectors under 3 different background conditions.

Remaining work: The remaining laboratory work is predominantly getting the other 7% of the pixels working to spec. These are mainly known faults and errors which are being fixed at time of writing in preparation for a cold run starting mid-February which may be the final laboratory cold run (although we do have enough time to have one more if necessary). There is also still a little work remaining on testing out observing software as far as possible and in completing the overall flatfielding of the arrays.

Expected sensitivities: This is all many of you will be interested in no doubt! SCUBA will be background-noise limited, as promised, however until such time as we actually get onto the telescope we do not know whether the background noise will be pure photon-noise or whether systematic sky variation ("sky-noise") will dominate. It seems highly likely that there will be sky-noise, but it is not known at what level. We do have software schemes for using the fact that we have an array to reduce any sky noise that may be present but again until we actually have real data the effectiveness cannot be guaranteed. Nonetheless we can still make some predictions, erring slightly on the conservative side. The following table shows a predicted noise-equivalent flux density (NEFD) at 450 micron, 850 micron and 1100 micron, along with the equivalent 5-sigma detection level in 5 minutes and 1 hour. However it is essential to bear in mind, that these are only estimates based on very best weather.

	filter	     NEFD	5 sigma in 5 mins		5 sigma in 1 hour

		(mJy Hz-1/2)	      (mJy)	  	            (mJy)

        

	450	     500		150		   	     42

	850	      40		 12			      3

	1100	      40		 12			      3

Figure 2: Fully-sampled jiggle map of a point source at 850 microns, aligned on the central pixel of the LW Array. (There are 10 contours, starting at 1% and increasing in 1% increments).

In all three cases data is obtained simultaneously with the long and short-wave arrays. The 1.1, 1.4 and 2.0 micron "photometric" pixels look out at the sky simultaneously but offset from each other. Thus observations of a point or small source the telescope has to be offset from one to the other but for a large source all three could be scanned over the same area with only a slight offset between the maps at the edges.

So wish us luck for the remaining tests and we will post the first telescope results on the Web as soon as we get them. Elsewhere in this issue there are details of the likely timescales and deadlines for applications for SCUBA. There is also a summary of the results of laboratory polarimetry tests with SCUBA.

Walter K. Gear

SCUBA Project Scientist