2.2. The three observational techniques

Single dish radio telescopes

Pionneered by Birkinshaw and collaborators, this was the first standard technique to successfully measure the SZ effect. It was and remains the best adapted technique for local clusters with a very large extent (e.g. Coma: Herbig et al. [15])

Radio interferometers

The most sensitive detections have been provided by radio interferometers which are less affected by sidelobe effects and can remove point sources by using the long baseline measurements. Moreover they benefit from large integration times (several months per year). The Ryle telescope (Saunders and Jones, this conference) in England and the OVRO-BIMA interferometer fitted (perverted) with radio receivers (Carlstrom et al. [10]) have obtained very sensitive maps of now tens of clusters with arcminute resolution at 15 and 30 GHz frequencies.

Bolometer photometers

The SuZie experiment provided the first detection of the SZ effect in the millimeter domain at the CSO 10m telescope (Wilbanks et al. [9], Holzapfel et al. [8]). It is made of 6 bolometers at a wavelength of 2 mm. By using a drift scan technique whereby fixing the telescope in local coordinates, the cluster drifts through the detectors with the Earth's diurnal motion, hence avoiding sidelobe effects and microphonic noise. The cluster A 2163 (the second X-ray brightest known cluster) was later detected at 1mm using the SuZie experiment with different filters and in the submillimeter domain by the balloon borne photometer PRONAOS-SPM (Lamarre et al. [7]) showing for the first time the change of sign of the SZ effect (see Fig. 1). Although in the (sub)millimeter domain, the point sources should not contaminate the SZ measurements so much, interstellar dust thermal emission must be dealt with to correct the measurements. Another limitation comes from sky noise: the water vapour is inhomogeneous in the atmosphere. Its emission in the telescope is variable in time and angle and frequency (see Fig. 1). The Diabolo experiment (Benoit et al. [4]) uses a dichroic beam splitter with six 0.1 K bolometers in order to simultaneously measure and hence subtract the water vapour emission at 1.2 mm (where the SZ effect is almost vanishing) from the SZ measurement at 2.1 mm. At the IRAM 30 m telescope, it provided the highest angular resolution of the SZ effect on several clusters with 30 arcsecond beam in 1995 (Désert et al. [6]) and 22 arcsecond beam since 1997 (Pointecouteau et al. [5] and this conference).