An Efficient Method for Simulating Light Curves of Cosmological Microlensing and Caustic Crossing Events

TL;DR

This paper introduces an efficient recursive method, called Adaptive Boundary Method, for simulating light curves of cosmological microlensing events, improving accuracy and efficiency especially for small sources and high magnifications.

Contribution

The paper presents a novel recursive simulation technique that outperforms traditional methods in modeling microlensing light curves for small sources.

Findings

The method is significantly more efficient than inverse ray shooting.

It maintains high accuracy in high-magnification regimes.

The code is publicly available for future research use.

Abstract

A new window to observing individual stars and other small sources at cosmological distances was opened recently, with the detection of several caustic-crossing events in galaxy cluster fields. Many more such events are expected soon from dedicated campaigns with the \emph{Hubble Space Telescope} and from the \emph{James Webb Space Telescope}. These events can teach us not only about the lensed sources themselves, such as individual high-redshift stars, star clusters, or accretion disks, but through their light-curves they also hold information about the point-mass function of the lens and thus, potentially, the composition of dark matter. We present here a simple method for simulating light curves of such events, i.e., the change in apparent magnitude of the source as it sweeps over the net of caustics generated by microlenses embedded around the critical region of the lens. The method is recursive and so any reasonably sized small source can be accommodated, down to sub-solar scales, in principle. We compare the method, which we dub \emph{Adaptive Boundary Method}, with other common methods such as simple inverse ray shooting, and demonstrate that it is significantly more efficient and accurate in the small-source and high-magnification regime of interest. A \textsc{python} version of the code is made publicly available in an open-source fashion for simulating future events.