Aims. Determining how efficiently gas collapses into stars at high redshifts is key to understanding galaxy evolution in the epoch of reionization (EoR). Globally, this process is quantified by the gas depletion time ( t dep ); on resolved scales, it is quantified by the slope and normalization of the Kennicutt-Schmidt (KS) relation. This work explores the global ( α CII ) and spatially resolved ( W CII ) CII-to-gas conversion factors at high- z and their use when inferring gas masses, surface densities, and t dep in the EoR. Methods. We selected galaxies at 4 < z < 9 from the SERRA cosmological zoom-in simulation, which features on-the-fly radiative transfer and resolves interstellar medium properties down to ≈30 pc. The CII emission modeling from photodissociation regions allows us to derive the global α CII and maps of W CII . We study their dependence on gas metallicity ( Z ), density ( n ), Mach number (ℳ), and burstiness parameter ( κ s ), and provide best-fit relations. Results. The α CII decreases with increasing Z and galaxy compactness, while the resolved W CII shows two regimes: at Z < 0.2 Z ⊙ , it anticorrelates with n and Z but not with κ s ; above this threshold, it also depends on κ s , with burstier regions having lower conversion factors. This implies W CII ∝ Σ CII −0.5 , as dense, metal-rich, and bursty regions exhibit higher CII surface brightnesses. Applying a constant α CII leads to an overestimation of Σ gas in bright Σ CII patches; this in turn flattens the KS slope and leads to overestimations of t dep by up to a factor of 4.
Vallini et al. (Tue,) studied this question.