A New Technique for Heterodyne Spectroscopy: Least-Squares Frequency Switching (LSFS)

TL;DR

LSFS is a novel heterodyne spectroscopy technique that directly derives spectra and gain profiles without traditional reference spectra, reducing noise and artifacts while maintaining high sensitivity.

Contribution

The paper introduces Least-Squares Frequency Switching (LSFS), a new method that improves heterodyne spectroscopy by eliminating the need for reference spectra and reducing systematic artifacts.

Findings

Retains nearly full theoretical sensitivity

Eliminates baseline wiggles and systematic artifacts

Effective for large bandwidths and complex spectral channels

Abstract

We describe a new technique for heterodyne spectroscopy, which we call Least-Squares Frequency Switching, or LSFS. This technique avoids the need for a traditional reference spectrum, which--when combined with the on-source spectrum--introduces both noise and systematic artifacts such as ``baseline wiggles''. In contrast, LSFS derives the spectrum directly, and in addition the instrumental gain profile. The resulting spectrum retains nearly the full theoretical sensitivity and introduces no systematic artifacts. Here we discuss mathematical details of the technique and use numerical experiments to explore optimum observing schemas. We outline a modification suitable for computationally difficult cases as the number of spectral channels grows beyond several thousand. We illustrate the method with three real-life examples. In one of practical interest, we created a large contiguous bandwidth aligning three smaller bandwidths end-to-end; radio astronomers are often faced with the need for a larger contiguous bandwidth than is provided with the available correlator.