It should be a goal, if not a spec, to provide reasonable
total power continuum performance; I would suggest the quantitative goal that the RMS of 1-second integrations with a detection
bandwidth of 500 MHz be within a factor of two of the radiometer equation prediction under excellent conditions, or when looking at a stable cold load (in either case taking out a common mode between beams or low order polynomial from each timestream).
If this is very expensive to accomadate, it would be reasonable to sacrifice this target. However, if the cost is say 3 Kelvin in Trx, we should pay the price considering that that 3 K -- in light of the stringency statistics around 20 GHz-- will have a very small consequence on other performance metrics in practice.
It is clear that this is possible because the old 300' 7-beam total-power receiver acheived excellent stability, and did a huge amount of good work.
The science drivers for stokes I continuum would be: moderately high resolution imaging of SZ clusters; blind extragalactic surveys (for GPS sources, or to characterize CMB foregrounds); high-resolution, zero-spacing maps for Galactic free free or spinning dust regions.
- 03 Oct 2007