6 Discussion

We have presented data from three observing campaigns carried out at the VLA to investigate intraday variability of compact extragalactic radio sources. In total, we observed 14 flat-spectrum sources (some repeatedly); out of these only 3 (namely 0153+744, 0836+710, and 1928+738) did not show IDV in any of the observations. This confirms that this phenomenon occurs consistently in a significant number of compact radio sources (with a flat spectrum), although clear statistical conclusions are precluded by our target selection, which favored objects that had previously shown fast variability. Also, the detailed properties vary strongly between sources. In most cases, the variations in total flux density are accompanied by similar variability of the linear polarization. Very fast variations (i.e., on timescales $\leq$ 2days, type II) occur in about 40% of all cases. One should keep in mind, however, that this is a small, by no means complete sample of sources. More detailed statistics, including observations of a larger number of sources with the Effelsberg 100m telescope, will be presented separately (Kraus et al., in preparation).

The size of a variable source can be derived from typical timescales assuming an intrinsic origin of the variability. In this case the linear size cannot be much larger than $c \cdot \Delta t$; with the usual synchrotron theory this leads to an expression for the brightness temperature of the source (cf. Wagner & Witzel 1995)

$$T_{\rm B}\,[{\rm K}] = 4.5~10^{10} \cdot S\,[{\rm Jy}] \cdot ... ... L}\,[{\rm Mpc}]} {\Delta t\,[{\rm d}] \cdot (1+z)} \right)^2.$$ (9)

Therefore, intraday variations imply brightness temperatures of 10¹⁶ - 10¹⁹K, or even 10²¹K for the very rapid variations in PKS0405-385 observed by Kedziora-Chudczer et al. (1997). This is a severe violation of the inverse Compton limit of 10¹²K (Kellermann & Pauliny-Toth 1969). We note that for the standard shock-in-jet model the observed (variability) brightness temperatures should be reduced by a factor ${\cal D}^3$ where ${\cal D}$ is the Doppler factor of the source (Blandford 1990). However, to bring down the brightness temperatures to the inverse Compton limit, Doppler factors of order 20-200 are needed. Such high Doppler factors are not supported by other observations (e.g. Witzel et al. 1988; Ghisellini et al. 1993), and they pose severe theoretical problems (e.g. Begelman et al. 1994). Since the discovery of IDV a large number of models have been developed to explain this phenomenon. Intrinsic explanations have used e.g. the motion of a compact structure (shock) in an underlying relativistic jet (Blandford & Königl 1979; Marscher & Gear 1985; Qian et al. 1991; Qian et al., in preparation) or the reconnection of magnetic field lines and coherent emission processes (Benford 1992; Lesch & Pohl 1992). Alternatively, IDV could be caused by extrinsic mechanisms, e.g. scattering in the interstellar medium (ISS, e.g. Rickett et al. 1995). Since the strength of ISS depends strongly on the galactic latitude (Rickett 1990), the measurement of a complete sample with a broad distribution in galactic latitude can be used for a crucial test for scattering models. However, none of the models discussed is fully capable of explaining all observed properties of IDV.

In forthcoming papers (Kraus et al., in preparation) we will present the IDV observations carried out at the 100m telescope of the MPIfR in Effelsberg and a complete statistical analysis of all available observations.

Acknowledgements

We thank the NRAO staff for their expert assistance and advice, and A. Patnaik and K. Otterbein for help during the data analysis. The VLA is a facility of the NRAO, which is operated by Associated Universities Inc., under cooperative agreement with the NSF.