Euclid. Predicting scaling relations for emission-line galaxies using simulated data
The Euclid spectroscopic surveys will substantially increase the number of observed emission-line galaxies. We assess the ability of the Euclid Wide Survey (EWS) and Euclid Deep Survey (EDS), at the final DR3 depth, to recover the star-forming main sequence (MS), dust-mass relation (DMR), and mass-metallicity relation (MZR) from emission-line measurements. We use MAMBO (mocks with abundance matching in Bologna) spectroscopic simulations processed with the official Euclid pipelines. The MS is traced over $0\leq z\leq 4$ using Pa$β$, H$α$, and [OII] as SFR tracers. The shallow EWS flux limit biases the MS towards higher normalisation and a flatter slope by selecting the brightest, most strongly star-forming galaxies. The deeper EDS enables almost complete sampling with Pa$β$ and H$α$ at log$_{10}(M_*/M_\odot)\geq10.3$ over $0\leq z\leq 1.88$, with mild biases at both redshift limits. The [OII]-based MS is increasingly biased towards higher SFRs at the highest redshifts. The DMR is traced through the Balmer decrement over $1.45\leq z\leq 1.88$ and $0.9\leq z\leq 1.88$ in the EWS and EDS, respectively. The derived colour excess agree well with the input values, although detecting the faint H$β$ line restricts the sample to less obscured galaxies at fixed $M_*$, particularly in the EWS; the DMR is better sampled in the EDS. Finally, a golden sample detected in multiple optical rest-frame emission lines at $1.49\leq z\leq 1.88$ recovers gas-phase metallicities with typical accuracy of 0.1 dex at log$_{10}(M_*/M_\odot)\geq10$, while lower-mass galaxies show biases up to 0.15 dex. Relative to H-band limited sample, this sample preferentially traces the upper MS and lower MZR. Despite the simplified simulations lacking effects such as mutual spectral contamination and detector persistence, these forecasts provide a benchmark for future Euclid spectroscopic analyses.