2026ApJ...998...49W

Hα luminosity traces the ionizing-photon output of short-lived, massive stars and is therefore a key diagnostic of star formation in high-redshift galaxies. Recent JWST observations have delivered updated Hα luminosity functions (LFs) of Hα-emitting galaxies out to an unprecedented 3.8 < z < 6.6, yet strong gravitational lensing can systematically skew the bright end and bias the cosmic star formation rate (SFR) density inferred from the LF. To quantify the population of strongly lensed Hα-emitting galaxies at 3.8 < z < 6.6, we use Monte Carlo sampling based on these new LFs and gravitational-lensing ray-tracing simulations with lenstronomy: sources adopt Sérsic surface-brightness profiles; foreground deflectors are drawn from an empirical velocity-dispersion function to build a 900 deg² light cone. In parallel, we convolve the intrinsic LF with the global magnification probability density to obtain an analytic approximation of lensed counts, which efficiently propagates intrinsic LF parameter uncertainties to lensed results. Our simulations show that ≳90% of Hα-emitting galaxies with SFR ≳ 10³M_⊙ yr^–1 are strongly lensed systems at 3.8 < z < 6.6, and the lensed number fraction per SFR bin rises from ∼10% to ∼90% across ${\mathrm{log}}_{10}{\rm{SFR}}\simeq 2.64$-3.05 M_⊙ yr^–1. Given the $\sim 62\,{{\rm{arcmin}}}^{2}$ GOODS-South coverage and a 50%-100% overall completeness, the existing FRESCO and CONGRESS observations should contain ∼1.7-3.4 strongly lensed Hα-emitting galaxies with SFR > 10 M_⊙ yr^–1. These results imply that, in the JWST era, analyses relying on samples with apparent SFR ≳ 10³M_⊙ yr^–1 must correct for strong lensing in order to obtain unbiased inferences.