Working Group 4  "Galaxy Evolution"

The WG4 has identified two level 1 science goals in the Bologna SKA meeting.

These are HI surveys to trace the evolution of gas content in galaxies
from the present to z~5 and deep continuum survyes which trace the star
formation in galaxies in an unbiased way out to similar redshifts as
faint starforming galaxies at intermediate redshift dominate the counts
at faint levels

The requirements for these are summarized in a spreadsheet, along with
the specs taken from the white paper design documents.  Highlighted in
red are the items which are discrepant.  Most designs meet the specs for
the HI sureys as they have a large fraction of the collecting area in a
condensed configuration for the required Tb sensitivity.  Main problem
is the small FOV of the KARST design which is not supplemented with a
focal plane array to get multiple beams on the sky to get a larger FOV. 

The continuum surveys require sufficient resolution to avoid being
severely limited by confusion.  Here 0.1" is not quite enough at 1.4 GHz
and really 0.01" is required at the nanoJy level.  Pushing to higher
frequencies will help in terms of resolution and confusion, but for the
population of star forming objects beyond z = 5 we then are looking at a
restfrequency of 30 GHz and heavily discriminate against non-thermal
sources with typical spectral indices of -0.8.  So surveys in the 1 GHz
range appear better suited and if continuum surveys piggyback on the HI
surveys then definitely 0.01" at 1.4 ghz is required.  Question still to
be addressed is the distribution of stations between 100 and 3000 km. 
The ~0.01" resolution will also help separating nuclear activity from
star forming disks at redshifts between 0 and 5.  In addition, if
hierachical galaxy formation scenarios are correct we expect smaller
entities at high redshifts so for the population beyond z = 5 higher
resolution is very important for testing these scenarios. 
           
For the continuum surveys the main discrepancies with the white paper
desings are the limited FOV of the KARST design and perhaps the 1.5 GHz
frequency cut off of the European design. Furthermore we need to look
closely at the dynamic range. 

In addition to these level 1 science goals attention was drawn at the
Bologna meeting to the possibility of redshifted CO 1-0 emission
entering the high frequency part of the band.  Two papers of interest in
this area are Silk and Spaans (ApJ 488, L97, 1997) and Carilli and Blain
(astro-ph/0201004).  Silk and Spaans demonstrate that the strength of
the CO emission (thanks to the increasing background temparature with
redshift) does not change much with redshift, while the emission from
dust does.  Star forming galaxies with 3x10^5 Orion type regions will
produce emission lines with a strenght of about 0.5 - 1 mJy.  In other
words the muJy regime probes regions with only 300 Orion type regions. 
Carilli and Blain demonstrate the superior speed of SKA for surveys,
thanks to the larger FOV compared to ALMA.  ALMA can detect CO lines at
high redshift, but is limited here to the higher transitions which only
probe the denser regions.  The regime beyond z = 10 is most uncertain:
no observations yet exist and the predictions are heavily dependent on
the adopted cosmology and galaxy formation scenarios.  CO 1-0 beyond z =
10 will be observable at frequencies of 10 GHz and below.  This area is
uniquely accessible to SKA and argues for level 1 science status as it
is a direct probe of galaxy formation scenarios.  Pushing the high
frequency limit to 20 GHz will expand the horizon to CO 2-1 beyond z =
10 and allow CO 1-0 to be observed beyond z = 4.7.  Interestingly this
is about the redshift where the HI surveys will only detect the most
massive galaxies, so the CO regime becomes highly complemenatry to the
redshift range probed by the HI surveys! This argues for 20 GHz as upper
cut-off for the SKA designs.  Main point is the superior speed for
surveys as demonstrated by Carilli and Blain.  Only 4 designs meet the
10 GHz criterion, whereas two (the Canadian and the US design) go as
high as 20 GHz. 

The table below illustrates the redshift space covered by ALMA and SKA:

CO transition     ALMA      SKA (10 GHZ)  SKA (20 GHz)   SKA (30 GHZ)  

1-0              z < 2.8     z > 10.5      z > 4.7        z > 2.8
2-1                < 6.7       > 22          > 10.5         > 6.7
3-2                < 10.5      > 33.5        > 16.5         > 10.5
4-3                < 14.3      > 45          > 22           > 14.3