[mmaimcal] alma OTF

[Harvey Liszt]

When doing OTF, or indeed any mapping, it should be kept
in mind that it is hard to interpolate or regrid without
increasing the effective resolution unless the sampling
is very good.  So if the increased data rate of a higher
sampling rate can be tolerated, it is preferable.

regards, Harvey
-------------

> 
> "Min" == Min Yun <myun at aoc.nrao.edu> writes:
> 
> Min> Reading thru the ASAC report, I noticed the OTF spec of 
> Min> 30' in 1 second with 1 second turn around time.  Since I am
> Min> doing OTF observations with the 12m telescope right now,
> Min> I did a quick sanity check.
> 
> Min> The 850 GHz beam for ALMA is 6", so the spec calls for
> Min> 300 beams/sec scan speed.  For continuum, the data dump rate
> Min> of 1 msec means we can achieve 3.3 measurements per beam,
> Min> which is very good.  In fact, we can get two measurements
> Min> per beam up to 1.4 THz, so this OTF spec has the continuum
> Min> OTF imaging in mind.
> 
> Actually, 3.3 measurements per beam is close to marginal for the total
> power OTF case.  When taper is taken into consideration, the minimum
> sampling rate becomes something like 2.5 points per beam.  Remember
> too that it is good to have some margin for error given that
> telescopes do not track a given course perfectly (due to wind, for
> example), so the true sampling obtained may be far from ideal.  This
> is not to say that we need to touch up the OTF spec, if the ALMA
> antennas perform as designed.
> 
> Min> For the line work, the top slew speed has to be reduced by
> Min> the minimum data dump rate.  A 16 msec dump time corresponds
> Min> 62.5 measurements per second, and thus 31.25 beam measurements
> Min> per second at best (top slew speed of 4.125 arcmin per second).
> Min> This does not sound too fast, but perhaps we are not limited
> Min> by the atmosphere nearly as much as the continuum case.
> Min> If the minimum dump time gets longer (i.e. phase switching
> Min> cycle), this will significantly slow down the required
> Min> slew rate -- something to keep in mind.  At the 12m, I am using
> Min> 30"/sec slew rate to map CO emission in galaxies.  
> 
> Min> This is not nearly fast enough, and my images are background
> Min> limited rather than thermal noise limited -- not clear if this
> Min> is due to the atmosphere or stability in the system, but we
> Min> will certainly be fighting similar effects.  Oh, the top
> Min> slew speed above should be 3.125'/sec rather than 4.125'/sec.
> 
> You could have gone faster as you are well under the minimum sampling
> for an observing frequency of 115 GHz.
> 
> Cheers,
> 
> Jeff
>