2024A&A...682A.146R

Strong gravitational lensing, a phenomenon rooted in the principles of general relativity, grants us a unique window into the distant cosmos by offering a direct probe into dark matter and providing independent constraints on the Hubble constant. These research objectives call for the utmost precision in the estimation of the lens mass and the source brightness distributions. Recent strides in telescope technology promise to provide an abundance of yet undiscovered strong-lensing systems, presenting observations of unprecedented quality. Realizing the full potential of these advancements hinges on achieving the highest fidelity in both source and lens reconstruction. In this study, we introduce LensCharm, a novel Bayesian approach for strong-lensing signal reconstruction. Unlike more prevalent methods, LensCharm enables the nonparametric reconstruction of both the source and lens concurrently, along with their associated uncertainties. We showcase the distinctive strengths of our approach through comprehensive analyses of both real-world and simulated astronomical data, underscoring its superiority in achieving precise reconstructions. We have made LensCharm publicly accessible, envisioning its empowerment of the next generation of astronomical observation reconstructions and cosmological constraints derived from strong gravitational lensing.