Date: 9 Oct 2001
Time: 4:00 pm EDT (2:00 pm Socorro, 1:00 pm Tucson)
Past minutes, etc on MMA Imaging and Calibration Division Page
Agenda
--------
--------
There was an ACA Tiger Team telecon last Friday.
Wild has posted a summary of the Oxford/RAL meetings.
--------
ASAC meeting tomorrow; agenda at the usual place.
--------
A telecon was held on Friday. Wootten wrote some minutes of this conference here.
Pierre Cox and Ewine van Dishoeck have produced a first draft science summary of ALMA+ACA simulations.
Stephane's group has performed some simulations on a file provided by Crutcher:
1) All images include Single-Dish data, with "typical" errors on it. The SAME single dish data is actually used in the ALMA case than in the ALMA+ACA case, to avoid differences due to noise realisations (which otherwise would not be unsignificant...)
2) With ALMA (+Single Dish) only the fidelity obtained in the error free case on the U and Q images is about 1 to 3 !... That is what I call "completely fails". Using ACA (in the error free case), they are of order 10 to 20.
3) However, speaking of "fidelity" in this case is not quite appropriate, because even a zero value is an interesting signal. When expressed in terms of "absolute error", the peak error is typically 3 times less with ACA than without, and the rms error about 5 times less with ACA than without. The peak and rms errors are very similar on U and Q images (although the signal is quite different). I am not quite sure this is what you would naively expect, but it is re-assuring in some sense. It may also help you in making a simplifying assumption for the analysis:
Assume the typical deconvolution error result in a 6 mJy/beam rms "noise" on U and Q images, and about a factor 10 larger on the I image with ACA, 5 times more with ALMA only... Apply that to the typical value for U (100 mJy, peak 200 mJy) and Q (+/- 60 mJy) and you get an error vector which indicates by how much you changed the polarisation direction...
If I am right, this is about 4 degrees for ALMA+ACA data, and 20 degrees for ALMA... I guess 4 degrees is quite acceptable given the calibration uncertainties, but may be 20 degrees (rms) starts being large.
4) These differences remain similar in presence of "typical" pointing, amplitude and phase errors.
--------
--------
Travel
Oh, always
------