Indonesian Students Club

Strong gravitational lensing of active galactic nuclei (AGN) enables measurements of cosmological parameters through their time-delay distances (time-delay cosmography, TDC). With data from the upcoming LSST survey, we anticipate using a sample of O(1000) lensed AGN for TDC. We test two key components of the TDC framework: lens modeling and time-delay measurements.
 
First, I’ll talk about how we can model lenses from ground-based image data. To work with a large dataset, we use Neural Posterior Estimation (NPE), an Simulation Based Inference (SBI) technique, to infer lens mass model parameters. We demonstrate that this method is successful in recovering key model parameters with less than 4% bias per lens and 10% precision per lens overall. We present the impact of physical assumptions included in the training data, distribution shift in model parameters, impact of different data optimization methods (such as perfect deconvolution, lens light subtraction etc) prior to modeling. Under every experiment, we infer the population model of lens galaxies using Bayesian Hierarchical Inference and demonstrate the need for a more flexible and physically motivated population model.
 
Second, I present ongoing efforts to measure the strong lensing time-delays from multi-band lensed AGN light curves. In this project we simulate realistic multi-band LAGN light curves including microlensing and observed at LSST cadence. We use the Python Curve Shifting Software (PyCS) to independently measure the time-delay in each band. To obtain a multi-band time-delay estimate, we combine the individual likelihoods from each band. We present the improvement in going from single to multi-band estimates, and highlight caveats and future extensions of this method.