A Simultaneous Stacking and Deblending Algorithm for Astronomical Images

TL;DR

This paper introduces an advanced algorithm that simultaneously stacks and deblends astronomical images, significantly reducing bias and variance in flux estimates for faint sources in confusion-limited surveys.

Contribution

The paper presents a novel simultaneous stacking and deblending algorithm that outperforms standard methods by greatly reducing bias and variance in flux measurements.

Findings

Reduces bias in flux estimates by up to an order of magnitude.

Improves RMS error by at least a factor of three in confusion-dominated regimes.

Applicable to Herschel and similar telescopes in low signal-to-noise, source-confused observations.

Abstract

Stacking analysis is a means of detecting faint sources using a priori position information to estimate an aggregate signal from individually undetected objects. Confusion severely limits the effectiveness of stacking in deep surveys with limited angular resolution, particularly at far infrared to submillimeter wavelengths, and causes a bias in stacking results. Deblending corrects measured fluxes for confusion from adjacent sources; however, we find that standard deblending methods only reduce the bias by roughly a factor of two while tripling the variance. We present an improved algorithm for simultaneous stacking and deblending that greatly reduces bias in the flux estimate with nearly minimum variance. When confusion from neighboring sources is the dominant error, our method improves upon RMS error by at least a factor of three and as much as an order of magnitude compared to other algorithms. This improvement will be useful for Herschel and other telescopes working in a source confused, low signal to noise regime.