EELG phys. prop. and ion. efficiency

In this work, we focus on an initial subset of the survey: a 30 square degree area with complete observations in all bands. We combine equivalent width (EW) measurements from J-PAS narrow-band photometry with artificial intelligence tech- niques to identify galaxies with emission lines exceeding 300{AA} in any emission line. We validate our selection using spectroscopic data from DESI/DR1 and characterize the selected sample through spectral energy distribution (SED) fitting with CIGALE.

We identify 917 EELGs up to z=0.8 over 30deg^2^, achieving a purity of 95% and a completeness of 96% for i-SDSS<22.5mag. Importantly, AGN contamination has been carefully considered and is estimated to be around 5%. Furthermore, a cross-match with DESI yields 79 counterparts, whose redshifts are in excellent agreement with our photometric estimates, thereby confirming the reliability of our redshift determination.

In addition, the derived emission line fluxes are in good agreement with spectroscopic measurements, reinforcing the robustness of our methodology. Moreover, the selected sample reveals strong correlations between ionizing photon production efficiency (ksi_ion_) and EW(H{beta}), which are consistent with previous observational studies at low and high redshift and theoretical expectations. Finally, most of the sources surpass the ionizing efficiency threshold required for reionization, highlighting their relevance as local analogues of early-universe galaxies.