Stellar populations and star formation histories of the most extreme [OIII] emitters at $z=1.3-3.7$

Abstract: As the James Webb Space Telescope approaches scientific operation, there is much interest in exploring the redshift range beyond that accessible with Hubble Space Telescope imaging. Currently, the only means to gauge the presence of such early galaxies is to age-date the stellar population of systems in the reionisation era. As a significant fraction of $z\simeq7-8$ galaxies are inferred from Spitzer photometry to have extremely intense [OIII] emission lines, it is commonly believed these are genuinely young systems that formed at redshifts $z<10$, consistent with a claimed rapid rise in the star formation density at that time. Here we study a spectroscopically-confirmed sample of extreme [OIII] emitters at $z=1.3-3.7$, using both dynamical masses estimated from [OIII] line widths and rest-frame UV to near-infrared photometry to illustrate the dangers of assuming such systems are genuinely young. For the most extreme of our intermediate redshift line emitters, we find dynamical masses $10-100$ times that associated with a young stellar population mass, which are difficult to explain solely by the presence of additional dark matter or gaseous reservoirs. Adopting nonparametric star formation histories, we show how the near-infrared photometry of a subset of our sample reveals an underlying old ($>100$ Myr) population whose stellar mass is $\simeq40$ times that associated with the starburst responsible for the extreme line emission. Without adequate rest-frame near-infrared photometry we argue it may be premature to conclude that extreme line emitters in the reionisation era are low mass systems that formed at redshifts below $z\simeq10$.