Adaptive Uniform Weighting: Pre-conditioning to Improve Image Fidelity

Abstract: The "dirty" image made by direct Fourier inversion of visibility data is an important first step in inteferometric imaging. This is where the "deconvolution problem" is defined and the degree to which that problem is either well- or ill-conditioned has direct consequences for the ultimate image fidelity that is achieved in practise. An under-utilised degree of freedom during Fourier imaging is the relative weights that are assigned to the visibility data. We explore the circumstances under which some adjustment of the relative weights might provide improvements to the "dirty" image, and consequently the ultimate post-deconvolution image fidelity. We develop a method to calculate a distinct effective local density estimate for each data point. When used in conjunction with a "uniform" weight correction and the desired clean beam (eg. Gaussian) tapering, it provides a significant improvement in the image quality over that provided by the current pixel-based density estimate. In many cases, particularly spectral-line observations and those with only limited sidereal tracking, this adaptive approach improves the beam quality by a factor of 2 to 10, as measured by the RMS residual relative to the best-fitting clean beam, providing an improvement in final image fidelity that is similar in magnitude.