The Tenerife Experiment

Over the past ten years the Tenerife Cosmic Microwave Background experiment has grown from a single 10 GHz radiometer working at $8^\circ$ resolution using parametric amplifiers to a ``fleet'' of radiometers working at 10, 15 and 33 GHz at $5^\circ$ resolution using the latest state-of-the-art HEMT technology for their receivers. Just installed is a very short baseline interferometer working at 33 GHz and $2^\circ$ resolution. In table  are shown the current system parameters of the instruments comprising the Tenerife experiment.

 

Table: Current system parameters of the Tenerife CMB Experiment.

Centre Frequency	Band-width	System Temperature	Sensitivity
(GHz)	(MHz)	(K)	(mK/Hz1/2
10.45	470	65	6
14.9	1000	75	5
32.5 Rad	3000	85	3
32.5 Int	3000	100	1

Each instrument has two channels A and B. In the case of the radiometer they are switched in anti-phase between the two conic feed horns which have a Full Width Half Maximum FWHM of $5^\circ$ separated by $8.1^\circ$ at a rate of 32Hz. In the interferometer the two channels are the two orthogonal outputs of the broad-band complex correlator, which mixes and correlates the signal from the two pyramid feed horns separated by 11 cm. The radiometers have a further set of switching on 8 seconds due to a ``wagging'' mirror to remove long term drifts. This hierarchy of switching results in a characteristic ``triple-beam'' response for the radiometer, while the interferometer produces a ``triple-beam'' in one channel and a ``double-beam'' in the other. Observations are made at a fixed elevation and the Earths rotation is used to made a scan in Right Ascension. Declinations are changed on a time-scale of a month in order to scan out an area between declinations $+30.0^\circ$ to $+45.0^\circ$.

The result of the observations are scans of 2 to 4 sidereal days from which a selection is made of those unaffected by atmospheric effects. These are then edited to remove parts which could be influenced by the Sun or the Moon. The largest remaining schematic effect is due to long term (<8hours) small amplitude baseline drifts due to variations in the water vapour content over the site. A Maximum entropy method (MEM) is used to remove these trends while not affecting any repeating astronomical signal. and to co-add the scans into a single low noise scan to be used in the statistical tests for the presence of CMB anisotropy.