The FIR/Radio Correlation at high z

At sub-mJy and microJy flux density levels the radio sky "lights up". As we move towards fainter and fainter flux density levels the radio sky becomes dominated by radio emission associated with star forming galaxies.

The correlation between the far-infrared (FIR) and radio emission is the tightest and most universal correlation known among the global properties of normal star forming and starburst galaxies (see v.d. Kruit 1973, Helou \& Bicay 1993, and references therein, Yun, Reddy \& Condon 2001). The famous "standard" plot (Condon 1992):

It is usually (but not always - see for example Ekers priv comm) explained in ``plausible'' terms - the dependence of both the thermal FIR dust emission (reprocessed uv radiation from young dusty stars) and the (largely) non-thermal radio emission (shocked electrons accelerated by SNR) on processes associated with massive star formation. The tightness of the relation is even more remarkable when one realises that the relation applies over four orders of magnitude in luminosity and is independent of galaxy type. In the past, complicated scenarios (such as coupling between the magnetic field and radiation denisty) have been invoked to explain the tightness of the correlation. Problems associated with the differing time scales associated with the rise in the FIR and radio emission also exist but these are largely ignored. Some useful papers in this regard include (Lisenfeld, Volk & Xu 1996; Dunne et al. 2000; Rengarajan & Takeuchi 2001). I have made things worse(!) by demonstrating that the correlation continues to apply to distant star forming galaxies in the HDF (RED DOTS shown below) and extending the relation (in terms of luminosity) by almost 2 orders of magnitude....

Its often convenient to look for departures in this linear correlation by plotting "q" (Condon 1992). Note that for these distant system its essential to k-correct the data - both in the radio but especially the Far-Mid IR. Here is how "q" evolves with z assuming an Arp 220 SED....

Here is "q" for the HDF (dont worry about the absolute value - that gets set by the assumed SED - in this case M82, for an Arp 220 SED "q" comes out around 2.3). The important point is that there is no obvious departure in q (apart from one radio loud AGN in the sample) that would imply any evolution in the correlation out to z=1.3.

The fact that the FIR-RC holds out to high-z is happy news for those who argue for z ~ 2-4 for the SCUBA source population (via the Carilli and Yun 2000 radio-submm photometric/spectral index measurements) and global SFRs based on (unbiased - no dust...) radio continuum measurements.

For more details about this read the papers: (A&A Letters - in press) and an older conference paper .