What drives the redshift evolution of strong emission line ratios?

Abstract

We study the physical mechanisms that cause the offset between low-redshift and highredshift galaxies on the [O iii]λ5007/Hβ versus [N ii]λ6584/Hα "Baldwin, Phillips & Terlevich" (BPT) diagram using a sample of local analogues of high-redshift galaxies. These high-redshift analogue galaxies are selected from the Sloan Digital Sky Survey. Located in the same region on the BPT diagram as the ultra-violet selected galaxies at z ∼ 2, these high-redshift analogue galaxies provide an ideal local benchmark to study the offset between the local and high-redshift galaxies on the BPT diagram. We compare the nitrogen-to-oxygen ratio (N/O), the shape of the ionising radiation field, and ionisation parameters between the high-redshift analogues and a sample of local reference galaxies. The higher ionisation parameter in the high-redshift analogues is the dominant physical mechanism driving the BPT offset from low-to high-redshift, particularly at high [N ii]λ6584/Hα. Furthermore, the N/O ratio enhancement also plays a minor role to cause the BPT offset. However, the shape of the ionising radiation field is unlikely to cause the BPT offset because the high-redshift analogues have a similar hard ionising radiation field as local reference galaxies. This hard radiation field cannot be produced by the current standard stellar synthesis models. The stellar rotation and binarity may help solve the discrepancy.

Modelling emission lines in star forming galaxies

“…Steidel et al (2014) also argued that the line ratios found in high redshift galaxies were characteristic of harder ionising radiation fields, similar to those in stellar population models that include the effects of binary stars (Stanway et al 2014;Xiao et al 2018) or stellar rotation, which can affect the number and type of Wolf-Rayet stars produced, which in turn changes the hardness of the ionising spectrum (Conroy 2013) 1 . Bian et al (2020) reached a similar conclusion based on their analysis of the local counterparts of high redshift galaxies.…”

“…However, star formation rate is a confounding variable and when samples are compared at the same star formation rate they are found to have the same electron density (see also Liu et al 2008;Kaasinen et al 2018). Several studies have argued that simply selecting the extreme tail of the local star-forming galaxy population is insufficient to explain the evolution of the BPT diagram and that there are other differences that drive this behaviour (Steidel et al 2014;Bian et al 2020). It would be instructive to repeat the analysis carried out by Kashino & Inoue (2019) for the samples of high redshift galaxies used to establish the evolution in the BPT diagram, to see if the connection between HII region properties and galaxy properties evolves in a way not described by the 𝑧 = 0 relations.…”

“…The median [OIII]/H𝛽 line ratio is measured to be 0.1-0.5 dex higher at 𝑧 ∼ 2.3 than it is at 𝑧 = 0. A number of studies have argued that this increase in the [OIII]/H𝛽 line ratio requires HII regions with a higher ionisation parameter (Kewley et al 2001;Steidel et al 2014;Bian et al 2020). Kewley et al (2001) modelled the emission line ratios in a sample of intense local starbursts that occupy the same region of the BPT diagram as high-redshift galaxies.…”

Fitting spectral energy distributions of FMOS-COSMOS emission-line galaxies atz∼ 1.6: Star formation rates, dust attenuation, and [OIII]λ5007 emission-line luminosities

“…Indeed, Steidel et al (2014Steidel et al ( , 2016 applied harder radiation fields using the binary Population and Spectral Synthesis code (BPASS) and these authors were able to produce models covering larger values of log([OIII]λ5007/Hβ), but always below the Kauffmann et al (2003) line. Other authors point to a higher ionization parameter or electron density as the main parameter producing the offset (e.g., Brinchmann et al 2008;Dopita et al 2016;Kojima et al 2017;Bian et al 2020). Nevertheless, current stellar synthesis models are unable to produce such hard radiation fields (Levesque et al 2010).…”

Fitting spectral energy distributions of FMOS-COSMOS emission-line galaxies at z$\sim$1.6: Star formation rates, dust attenuation, and [OIII]$λ$5007 emission-line luminosities

“…Indeed, Steidel et al (2014Steidel et al ( , 2016 applied harder radiation fields using the binary Population and Spectral Synthesis code (BPASS) and these authors were able to produce models covering larger values of log([OIII]λ5007/Hβ), but always below the Kauffmann et al (2003) line. Other authors point to a higher ionization parameter or electron density as the main parameter producing the offset (e.g., Brinchmann et al 2008;Dopita et al 2016;Kojima et al 2017;Bian et al 2020). Nevertheless, current stellar synthesis models are unable to produce such hard radiation fields (Levesque et al 2010).…”