KIPAC Tea Talk: QUELLS: Detecting Strong-Lensing Echoes in Unresolved Quasar Light Curves at Survey Scale

Abstract: Strongly lensed quasars provide valuable probes of cosmology and dark matter, but conventional discovery methods often rely on spatially resolving multiple images, leaving compact systems difficult to identify in wide-field surveys. Unresolved systems, however, can still be identified through their variability: the combined light curve contains delayed copies of the same intrinsic quasar signal. I will present QUELLS, a scalable Gaussian-process framework for detecting these strong-lensing echoes directly in irregularly sampled quasar light curves. The method models intrinsic variability, measurement noise and survey cadence to distinguish delayed signals from ordinary stochastic variability, assigning detection scores to rank lens candidates and estimate their preferred time delays. I will discuss its performance on survey-realistic multi-band simulations, including the benefits of combining photometric bands, and present first results from its application to ZTF DR24 multi-band light curves of 299,151 SDSS and DESI DR1 quasars. Designed for population-scale searches, QUELLS offers a complementary route to discovering compact strong lenses that may be missed by imaging-based methods, with applications to ZTF, Gaia DR3/DR4 and future Rubin/LSST observations.