2025A&A...699A.299M

The Flashlights program with the Hubble Space Telescope imaged the six Hubble Frontier Fields galaxy clusters in two epochs and detected twenty transients. These are primarily expected to be caustic-crossing events (CCEs) where bright stars in distant lensed galaxies, typically at redshift z≈1-3, get temporarily magnified close to cluster caustics. Since CCEs are generally biased toward more massive and luminous stars, they offer a unique route for probing the high end of the stellar mass function. We take advantage of the Flashlights event statistics to place preliminary constraints on the stellar initial mass function (IMF) around cosmic noon. The photometry (along with spectral information) of lensed arcs is used to infer their various stellar properties, and stellar synthesis models are used to evolve a recent stellar population in them. We estimate the microlens surface density near each arc and, together with existing lens models and simple formalism for CCEs, calculate the expected rate for a given IMF. We find that, on average, a Salpeter-like IMF (α = 2.35) underpredicts the number of observed CCEs by a factor of ∼0.7, and a top-heavy IMF (α = 1.00) overpredicts by a factor of ∼1.7, suggesting that the average IMF slope may lie somewhere in between. However, given the large uncertainties associated with estimating the stellar populations, these results are strongly model-dependent. Nevertheless, we introduce a useful framework for constraining the IMF using CCEs. Observations with James Webb Space Telescope are already yielding many more CCEs and will soon enable more stringent constraints on the IMF at a range of redshifts.