[Paper Review] Short-Spacings Correction from the Single-Dish Perspective

Snežana Stanimirović|arXiv (Cornell University)|May 20, 2002

Surface Roughness and Optical Measurements4 citations

TL;DR

This paper presents a comprehensive framework for correcting missing short-spacings in interferometric radio astronomy by combining single-dish and interferometer data. It demonstrates that combining data in the Fourier domain (e.g., via the immerge method) or during deconvolution yields high-fidelity images with accurate flux densities and high spatial resolution, effectively recovering total power and extending dynamic range.

ABSTRACT

While, in general, interferometers provide high spatial resolution for imaging small-scale structure (corresponding to high spatial frequencies in the Fourier plane), single-dishes can be used to image the largest spatial scales (corresponding to the lowest spatial frequencies), including the total power (corresponding to zero spatial frequency). For many astrophysical studies, it is essential to bring `both worlds' together by combining information over a wide range of spatial frequencies. This article demonstrates the effects of missing short-spacings, and discusses two main issues: (a) how to provide missing short-spacings to interferometric data, and (b) how to combine short-spacing single-dish data with those from an interferometer.

Motivation & Objective

• To address the limitation of interferometers in recovering large-scale emission due to missing short-spacings in the u-v plane.
• To enable accurate flux density measurements by incorporating total power information from single-dish telescopes.
• To develop and compare robust data combination techniques that preserve interferometric resolution while recovering extended structure.
• To provide practical cross-calibration and combination strategies for heterogeneous radio astronomy datasets.
• To demonstrate that short-spacings correction is essential for faithful imaging of extended astrophysical objects such as galaxies and nebulae.

Proposed method

• Uses the Fourier domain method (e.g., immerge) to combine single-dish and interferometric visibility data by inserting short-spacings into the u-v plane.
• Applies a 'linear combination' method that blends data in the image plane without requiring Fourier transformation or deconvolution of single-dish data.
• Employs a 'default image' approach where single-dish data are used as a prior in deconvolution to stabilize the solution.
• Uses joint deconvolution of both datasets simultaneously, treating them as a single visibility set with different noise characteristics.
• Relies on cross-calibration between interferometer and single-dish data using overlapping spatial frequency regions.
• Validates methods using HI observations of the SMC at 169 km s⁻¹, comparing results from ATCA and Parkes telescope data.

Experimental results

Research questions

• RQ1 How do missing short-spacings in interferometric data affect the fidelity of extended source imaging?
• RQ2 What are the most effective and robust methods for combining single-dish and interferometer data to recover large-scale emission?
• RQ3 How does the choice of data combination method impact the total flux density, dynamic range, and noise in the final image?
• RQ4 What role does cross-calibration between interferometric and single-dish data play in ensuring consistency and accuracy?
• RQ5 Which method—Fourier domain, linear combination, or joint deconvolution—provides the most reliable short-spacings correction with minimal artifacts?

Key findings

• The immerge method (Fourier domain combination) produced a total flux density of 5600 Jy, slightly below the Parkes-only value of 6100 Jy, indicating a minor underestimation due to over-weighting of short interferometric spacings.
• The 'linear combination' method yielded a total flux of 6500 Jy, exceeding the Parkes value by 7%, likely due to over-weighting of single-dish data in the overlap region.
• The 'default image' method produced a flux of 6300 Jy, within 3% of the Parkes value, showing reliable performance when a large single-dish is used.
• The joint deconvolution method achieved 5900 Jy, within 3% of the Parkes value, and is theoretically optimal but sensitive to accurate noise variance estimates.
• All four methods produced comparable noise levels (28–32 mJy beam⁻¹) and dynamic ranges, with minima around -0.3 Jy beam⁻¹ and maxima near 2.0–2.2 Jy beam⁻¹.
• The Fourier domain method (immerge) is the fastest and most robust, avoiding non-linear deconvolution and edge effects common in single-dish data transformation.