📝 SummarXiv

Abstract

Interpreting galaxy properties from astronomical surveys relies heavily on spectral energy distribution (SED) modelling, yet uncertainties in key model ingredients are often overlooked. By leveraging a $z\sim0$ galaxy sample from the EAGLE simulation, we generate synthetic SDSS spectral and VISTA photometric observations with controlled assumptions, to assess how variations in stellar spectral library, initial mass function (IMF) and metallicity prescriptions within the BPASS-framework affect inferred galaxy properties. Our analysis combines spectral fitting from 3800 to 9200 A with photometric constraints extending to 2.3 $\mu$m, enabling robust assessment across a broad wavelength baseline. Our findings reveal mass, age and star formation rate vary by $0.27\pm0.09$, $0.19\pm0.11$ and $1.4\pm1.0$ dex, respectively, greater than observational uncertainties reported in surveys. Notably, we find stellar spectral library choice is capable of transforming a galaxy from appearing star-forming to quiescent, while a fixed metallicity assumption yields systematic biases when the chosen metallicity is incorrect. These modelling differences impact the reconstructed total mass assembly history in galaxies by up to $\sim12$ percent and bias the demographic and star formation history conclusions drawn from surveys. As upcoming missions like Euclid, Roman and CASTOR aim to characterise galaxy evolution with unprecedented precision, our results highlight the need for careful propagation of SED modelling uncertainties and transparency in model selection.