ALMA Memo #291

First Simulations of Imaging Performance of a Spiral Zoom Array; Comparisons with a Single Ring Array

Imaging simulations have been carried out with a version of the spiral zoom array presented in Conway(2000) (Memo 283). Two test images were used, one based on a VLA image of Cygnus A and another based on an OVRO CO map of M51. Both snapshot and long track CLEAN simulations were carried out. The results were compared with those obtained for a single ring array with similar resolution. The single ring array has a fairly uniform coverage out to its maximum baseline and therefore provides a useful constrast to the spiral zoom array which has a strongly tapered (almost gaussian) coverage. For both test images the spiral array gave better imaging performance. The difference was particularly large for the case of the long-track observations when the rms errors were >10 times larger for the ring than for the spiral arrays. This large difference arises because the large near-in sidelobes of ring arrays, caused by the sharp edge to the uv-coverage, persists even after a long track.

These simulations illustrate that it is important to consider that deconvolution requires both extrapolation of the uv data as well as interpolation between uv points. A naturally tapered uv coverage appears to provide more constraints on the necessary extrapolation than for the case of ring arrays. More imaging simulations must be carried out, paricularly comparing zoom arrays to the somewhat tapered 'donut' or 'double ring' arrays that have been proposed. However, these intial tests show that spiral zoom arrays have good imaging performance as well as high observing efficiency (Conway 2000, Memo 283) and the ability to have finely adjustable resolution (Conway 1998, Memo 216).

From here, you can:

View an HTML version of ALMA Memo #291. (Original format)

Download a .62 MB gzipped postscript version of ALMA Memo #291 from

here

Last modified: March 01, 2000