SCUBA2 Mapping Techniques Dennis Kelly UKATC Canadian SCUBA2

SCUBA-2: Mapping Techniques Dennis Kelly (UKATC) Canadian SCUBA-2 Consortium

Summary • Scuba 2 will consist of two 5120 pixel TES arrays operating at 850 and 450 microns respectively. • Each array consists of 4 subarrays. • The field-of-view on the JCMT will be 8 arcminutes. • The arrays will be DC-coupled to the data acquisition system. • The brightness of the Earth’s atmospheric emission puts severe demands on flat-fielding for simple imaging. • DREAM avoids this problem by using each bolometer to make a mini-map overlapped with its neighbours and then combining the result. • DREAM was conceived by Rudolf Le. Poole and implemented by Hans van Someren Greve. • SCANMAP avoids the flat-field problem by removing the background by fitting a baseline to the raw data from each bolometer.

Photon-Limited S/N

Transition Edge Sensors Resistance Temperature I = V/R Pohmic = V 2/R Ptotal = Pinput + Pohmic

TES Response Current Input Power

Response of Different Pixels Current Input Power Two bolometers have different responses to the same input power. Assume the measured relative responses are in error by 5%. In the total power measured, this will mean the deduced power is different by 5% of the signal from the Earth’s atmosphere, which is many orders of magnitude greater than an astronomical signal. If each bolometer is doing a differential measurement, then the 5% difference just turns into a 5% error in photometry.

DREAM Pattern Chosen sky positions Secondary mirror is driven so that each bolometer makes a mini-map. The mini-map of each bolometer overlaps with its neighbours.

Sample Points Sky measurement position Vj Solve for sky value Si by least-squares { For j=1, 24: Vj= Σw. S 9 ji i i=1 Where wji is the weight between the positions of Vj and Si

Piston Correction Average bolometer K 1 Bolometer 1 K 2 Bolometer 2 The j-th sky position value deduced from bolometer number b can be written as the sum of the sky flux plus the zero-point offset for that bolometer. This can be expanded to a dependency on all the sky fluxes plus all the bolometer zero-points by introducing weights that are mostly zero, leading to a set of equations solvable by least-squares. An additional equation demands the sum of the bolometer zero-points is zero. Sbj = 0. F 1 + 0. F 2 + ……. . + 1. Fi + 0. Fi+1……. + 0. K 1 + 0. K 2 + …. . + 1. Kb + 0. Kb+1……. . 0 = 0. F 1 + 0. F 2 + ……. . + 0. Fi+1……. + 1. K 1 + 1. K 2 + …. . + 1. Kb+1……. .

Data Processing Factors • The form of the least-squares problems means that most of the data processing (which concerns the geometrical weights in the first stage, and the selection weights in the second stage) can be done in advance. Only the much smaller amount of processing involving the measurements has to be done after data acquisition. • A disadvantage of this is that it is not possible to zero-weight measurements. Any bad values have to be patched-over in some way.

SCANMAP A SCANMAP simulation was based on the following pattern, with full allowance being made for non-linear non-identical bolometers, photon noise and 1/f noise, and a sloping, wind-blown emission from the Earth’s atmosphere. wind direction flyback array scan direction

SCANMAP Noise-free simulation of ten scans of ten identical bolometers with wind-blown sloping sky and a grid of point sources. Same as above, but noise added and bolometers having a spread of responses. As above, but with a linear baseline fitted and removed from each scan of each bolometer.

SCANMAP Pattern Overlapping Scans at 26. 6 degrees to achieve full sampling at 450 microns

SCANMAP Issues • A simple approach to SCANMAP is sufficient for survey work, the best results being achieved when the astronomical sources are small and sparsely distributed relative to the scan length. • Reliable maps of large sources will require an orthogonal pattern of scans, and analysing this will require study.