New Approximation of Magnification Statistics for Random Microlensing of Magnified Sources

Gravitationally lensed extragalactic sources are often subject to statistical microlensing by stars in the galaxy or cluster lens. Accurate models of the flux statistics are required for inferring source and lens properties from flux observations. We derive an accurate semi-analytic approximation for calculating the mean and variance of the magnification factor, which are applicable to Gaussian source profiles and arbitrary non-uniform macro lens models, and hence can save the need to perform expensive numerical simulations. The results are given as single and double lens-plane integrals with simple, non-oscillatory integrands, and hence are fast computable using common Monte Carlo integrators. Employing numerical ray-shooting experiments, we examine the case of a highly magnified source near a macro fold caustic, and demonstrate the excellent accuracy of this semi-analytic approximation in the regime of multiple micro images. Additionally, we point out how the maximum persistent magnification achievable near a macro caustic is fundamentally limited by the masses and number density of the foreground microlenses, in addition to the source's physical size.