2.9. Microwave background temperature
The ratio of the thermal SZ effect of a cluster at two different frequencies is a function of the temperature of the MBR, with some slight dependence on the temperature of gas in the cluster and the cluster radial velocity. Thus a precise measurement of the SZ effect spectrum can be used to measure the MBR temperature at distant locations, and over a wide range of redshifts. This would allow a test of the MBR temperature evolution
expected from our normal model of radiation in the Universe. Battistelli (2002) found no deviation from eq. (36), but the errors on the SZ effect amplitudes they used were substantial, as were the errors on the MBR temperatures at high redshift (which were obtained from molecular excitation studies). At present this method is relatively insensitive to deviations from eq. (36).
Clearly this technique relies on accurate knowledge of the spectrum of the thermal SZ effect, and therefore on the absence of any significant kinematic SZ effect. Filtering of the effects requires precise knowledge of the relative calibration of the SZ effect data at widely separated frequencies (Sec. 2.1.1), and the absence of temperature substructures in the cluster (Sec. 2.5.3) large enough to cause significant spectral changes.