The PAU Survey: Background light estimation with deep learning techniques

TL;DR

This paper presents BKGnet, a deep learning model designed to accurately estimate and correct background light in astronomical images, significantly improving photometry and reducing systematic errors in the PAUS survey.

Contribution

The paper introduces BKGnet, a novel deep neural network for background light estimation in astronomical imaging, tailored for PAUS data, enhancing photometric accuracy and error correction.

Findings

BKGnet improves photometric flux measurements by up to 20%.

It reduces systematic background error trends with magnitude.

It decreases the photometric redshift outlier rate.

Abstract

In any imaging survey, measuring accurately the astronomical background light is crucial to obtain good photometry. This paper introduces BKGnet, a deep neural network to predict the background and its associated error. BKGnet has been developed for data from the Physics of the Accelerating Universe Survey (PAUS), an imaging survey using a 40 narrow-band filter camera (PAUCam). Images obtained with PAUCam are affected by scattered light: an optical effect consisting of light multiply that deposits energy in specific detector regions contaminating the science measurements. Fortunately, scattered light is not a random effect, but it can be predicted and corrected for. We have found that BKGnet background predictions are very robust to distorting effects, while still being statistically accurate. On average, the use of BKGnet improves the photometric flux measurements by 7% and up to 20% at the bright end. BKGnet also removes a systematic trend in the background error estimation with magnitude in the i-band that is present with the current PAU data management method. With BKGnet, we reduce the photometric redshift outlier rate