[asac] comments for Nov 13 ASAC telecon

[Stephane Guilloteau]

>
Hi Christine,

    here are my own comments to your question


>(2) Extension of Band 3 to lower frequencies


    SiO is one thing, but not as compelling as it looks. SiO is technically
usefull for the test interferometer, but with ALMA itself, I think the
argument
falls a little bit. Strong quasars offer as much sensitivity, ot even more
for,
all "technical" purpose.

    However, the key issue is Science. If Band 2 is second priority, which
it
may be for FUNDING reasons (even if it is comparatively cheap, it's still
quite
a lot of money), there are a number of critical lines between 86 and 89 GHz.
HCN and its isotopes, idem for N2H+ and so.

    Is there really anything so usefull between 84 and 86 GHz ?

>


>(3) Scientific Merit of the Enhanced Correlator



>In Hasegawa's memo on the scientific merit of the advanced correlator, I
>am a bit confused by his third example, searching for absorption line
>features agains bright nuclei or quasars. This may be because I don't
>work in this field, but I don't understand why we need the enhanced
>correlator to do this science. If the gas to be studied in absorption is
>in the Milky Way itself, then the radial velocity would probably be known
>fairly accurately from the rotation curve of the Milky Way. If the gas
>was in another galaxy between us and the quasar, the radial velocity is
>often known to 100-200 km/s from optical spectroscopy.
   Not true: the intervening galaxy is often not detected in the optical,
because
of the contrast with the quasar spectrum. Blind searches are required...

> Since super-high
>resolution (< 1km/s) would not be required (at least not in an initial
>search), then a bandwidth of, say 400 km/s with 400 channels would be
>sufficient.
    Not true either. We need high spectral resolution here. 1 km/s isn't
sufficient.

> Finally, if the gas was in the actual molecular torus around
>an AGN, again the velocity would still be known to 100-200 km/s accuracy,
>and the line-widths would be expected to be relatively narrow (few 100
>km/s, if I'm remembering some recent IRAM work properly). One way
>that I can see that the enhanced correlator would be helpful for
>absorption line studies is if one was doing a blind search; however, such
>a search could probably be done with the existing correlator, at the
>expense of more integration time.
>
    The last sentence is correct, and the whole question is:
Is it more efficient (money-wise) to build a more expensive correlator which
will
do some science much faster than a simpler, cheaper one, which will be slow
in a few cases ?

    My personal answer to that question follows these guidelines:
- the "enhanced" correlator should NOT degrade the sensitivity for all
"normal"
science cases.
- for other projects, the cost of observing time is going to be of order 200
k$/day
for the "enhanced ALMA" (assuming a 6.7 % /year operating cost and a 30
years
lifetime). So you "regain" 20 M$ investment on about a 100 days of
observations,
which for 30 years lifetime is just 3 days per year.
- The extra cost must include the processing (and storage) cost also.

==> Conclusion: an enhanced correlator pays off quickly, provided it has no
sensitivity
loss in the "normal" modes.

        Stephane