Difference Imaging of Lensed Quasar Candidates in the SDSS Supernova Survey Region

TL;DR

This study tests a difference imaging method to identify gravitationally lensed quasars in SDSS data, finding it effective but limited by survey resolution, with promising applications for future surveys like LSST and Pan-STARRS.

Contribution

The paper demonstrates the application of difference imaging to detect lensed quasars, providing a new approach and evaluating its effectiveness and limitations in SDSS data.

Findings

21 variable sources identified out of 20768

Only one source showed non-point source structure, rejected as lens candidate

Estimated false positive rate of about one per square degree

Abstract

Difference imaging provides a new way to discover gravitationally lensed quasars because few non-lensed sources will show spatially extended, time variable flux. We test the method on lens candidates in the Sloan Digital Sky Survey (SDSS) Supernova Survey region from the SDSS Quasar Lens Search (SQLS) and their surrounding fields. Starting from 20768 sources, including 49 SDSS quasars and 36 candidate lenses/lensed images, we find that 21 sources including 15 SDSS QSOs and 7 candidate lenses/lensed images are non-periodic variable sources. We can measure the spatial structure of the variable flux for 18 of these sources and identify only one as a non-point source. This source does not display the compelling spatial structure of the variable flux of known lensed quasars, so we reject it as a lens candidate. None of the lens candidates from the SQLS survive our cuts. Given our effective survey area of order 0.71 square degrees, this indicates a false positive rate of order one per square degree for themethod. The fraction of quasars not found to be variable and the false positive rate should both fall if we analyze the full, later data releases for the SDSS fields. While application of the method to the SDSS is limited by the resolution, depth, and sampling of the survey, several future surveys such as Pan-STARRS, LSST, and SNAP will avoid these limitations.