Extracting H$\alpha$ flux from photometric data in the J-PLUS survey

Abstract

We present the main steps that will be taken to extract H$\alpha$ emission flux from Javalambre Photometric Local Universe Survey (J-PLUS) photometric data. For galaxies with $z\lesssim0.015$, the H$\alpha$+[NII] emission is covered by the J-PLUS narrow-band filter $F660$. We explore three different methods to extract the H$\alpha$ + [NII] flux from J-PLUS photometric data: a combination of a broad-band and a narrow-band filter ($r'$ and $F660$), two broad-band and a narrow-band one ($r'$, $i'$ and $F660$), and a SED-fitting based method using 8 photometric points. To test these methodologies, we simulated J-PLUS data from a sample of 7511 SDSS spectra with measured H$\alpha$ flux. Based on the same sample, we derive two empirical relations to correct the derived H$\alpha$+[NII] flux from dust extinction and [NII] contamination. We find that the only unbiased method is the SED fitting based one. The combination of two filters underestimates the measurements of the H$\alpha$ + [NII] flux by a 28%, while the three filters method by a 9%. We study the error budget of the SED-fitting based method and find that, in addition to the photometric error, our measurements have a systematic uncertainty of a 4.3%. Several sources contribute to this uncertainty: differences between our measurement procedure and the one used to derive the spectroscopic values, the use of simple stellar populations as templates, and the intrinsic errors of the spectra, which were not taken into account. Apart from that, the empirical corrections for dust extinction and [NII] contamination add an extra uncertainty of 14%. Given the J-PLUS photometric system, the best methodology to extract H$\alpha$ + [NII] flux is the SED-fitting based one. Using this method, we are able to recover reliable H$\alpha$ fluxes for thousands of nearby galaxies in a robust and homogeneous way.