Statistical microlensing towards magnified high-redshift star clusters

We study light variability of gravitationally magnified high-redshift star clusters induced by a foreground population microlenses. This arises as the incoherent superposition variations from large number source stars traversing random magnification pattern on plane. The curve resembles scale-invariant, Gaussian process timescales years to decades, while exhibits rapid and frequent micro-caustic crossing flares larger amplitudes days months. For concrete example, we young Lyman-continuum-leaking cluster recently discovered in lensed Sunburst Arc at $z=2.37$. show that one image happens be intervened faint galaxy, hence should exhibit variable flux $1$--$2\%$ level, which is measurable space with $\sim 1$--$3\,$ks exposures Hubble Space Telescope more easily James Webb Telescope, or ground using $\sim$4-meter telescope without adaptive optics. Detailed measurement this will enable us determine absolute macro intrinsic mass length scales cluster, test synthetic models stellar population, probe multiplicity massive stars. furthermore suggest monitoring other images are free significant intervention microlenses, allow planetary compact objects constituting little just few percent dark matter. Given typical surface density intracluster stars, expect phenomenon relevant for many Cosmic Noon behind galaxy lenses.