2025MNRAS.540.3693S

Determining the dust properties of high-redshift galaxies from their far-infrared continuum emission is challenging due to limited multifrequency data. As a result, the dust spectral energy distribution (SED) is often modelled as a single-temperature modified blackbody. We assess the accuracy of the single-temperature approximation by constructing realistic dust SEDs using a physically motivated prescription where the dust temperature probability distribution function (PDF) is described by a skewed normal distribution. This approach captures the complexity of the mass-weighted and luminosity-weighted temperature PDFs of simulated galaxies and quasars, and yields far-infrared SEDs that match high-redshift observations. We explore how varying the mean temperature ($\bar{T}_\mathrm{d}$), width, and skewness of the temperature PDF affects the recovery of the dust mass, infrared (IR) luminosity, and dust emissivity index ($\beta _\mathrm{d}$) at $z=7$. Fitting the dust SEDs with a single-temperature approximation, we find that dust masses are generally well recovered, although they may be underestimated by up to $0.6\, \mathrm{dex}$ for broad temperature distributions with a low $\bar{T}_\mathrm{d}\lesssim 40\, \mathrm{K}$, as seen in some high-redshift quasars and/or evolved galaxies. IR luminosities are generally recovered within the $1\sigma$ uncertainty (${\lesssim} 0.3$ dex), except at $\bar{T}_\mathrm{d}\gtrsim 80\,$ K, where the peak shifts well beyond ALMA's wavelength coverage. The inferred dust emissivity index is consistently shallower than the input one ($\beta _\mathrm{d}=2$) due to the effect of multitemperature dust, suggesting that a steep $\beta _\mathrm{d}$ may probe dust composition and grain size variations. With larger galaxy samples and well-sampled dust SEDs, systematic errors from multitemperature dust may dominate over fitting uncertainties and should thus be considered.