Summary

These three radio telescopes illustrate different methods of imaging using dipolar arrays as applied to radioastronomy. GEETEE: One-dimensional image synthesis on the meridian with the entire aperture being present at the same time; CLARK LAKE: A two dimensional image synthesis which gave periods of integration much larger than the meridian transit time. The entire aperture was present during an observation schedule; MOST: Rotational synthesis which is used to synthesise a large two dimensional array, using a linear array. All of them use principles of beam forming. GEETEE and CLARK LAKE use the method of measurement of visibilities in the $(u,v)$ domain, while MOST employs the method of direct fan beam synthesis.

We see that the dipolar arrays are used in the meter wavelength ranges more often than at high frequencies. They have very wide fields of view (GEETEE, almost $100^0$) and are very good workhorses for surveying the sky. They are good imaging instruments also since they combine the phased array techniques with the principles of synthesis imaging to make images. Unfortunately most of the arrays are equipped with a limited number of correlators and cannot measure all the possible $\lq\lq {n (n-1)/2}''$ baselines with $\lq\lq n''$ aperture elements. Thus they are not well suited for applications of self-calibration. Being skeleton telescopes, they have no redundancy in the imaging mode and redundant baseline calibration is not easily applicable. (See Chapter 5 for a discussion on self-calibration and redundant baseline calibration). This has resulted in surveys with limited dynamic range capability. None of these low frequency arrays are equipped with feeds with orthogonal polarisation. So they are not suitable for polarisation studies.

While combining the beam forming techniques with the synthesis techniques, one has to be very careful about the sampling requirement of the spatial frequencies; otherwise one will end up with grating lobes in the synthesised image, even while using linear arrays with contiguous elements spaced ${\lambda}/2$ apart. Since the dipolar arrays are employed generally as correlation telescopes and do not have a common collecting area in the arms used for correlation, they suffer from the ``zero-spacing problem^7.7". Most often today's receivers employ bandpass sampling^7.8 and if the sampling frequency is not properly chosen one will lose signal to noise. While imaging with arrays it is not un-common, one confronts conflicting requirements between surveying sensitivity and the field of view.

A question may arise in your minds at this stage - with a handful of telescopes using the phased array approach, is there any future for them in radio astronomy? In the remainder of this chapter, I will discuss the possible future of dipolar arrays for radio astronomy.

Footnotes

... problem^7.7

The zero spacing problem refers to the difficulty in imaging very large sources, (whose visibilities peak near the origin of the u-v plane) with arrays which provide few to no samples near the u-v plane origin. See Section 11.6 for a more detailed discussion.

... sampling^7.8

See Chapter 1