The environmental dependence of the mass-metallicity relation
May. 15, 2023
Gas-phase metallicity is a crucial probe of various physical processes during the formation and evolution of galaxies. Here we studied the environmental dependence of the mass-metallicity relation using state-of-the-art cosmological hydrodynamical simulations of EAGLE and IllustrisTNG. We found that central and satellite galaxies have the opposite environmental dependence of their mass-metallicity relation, which is due to different underlying physical processes.
This blog is based on arXiv:2305.08161.
A galaxy is defined as the assembly of gravitationally bound visible stars. It may also contain multiphase gas and a dark matter halo, but these components are not necessary. The most fundamental quantity to describe a galaxy is its stellar mass, which is the "mass" in the title.
Metal, different from our daily usage, is defined as all the elements heavier than helium. This definition is motivated by the fact that most of the hydrogen and helium in our Universe are formed during the primordial nuclear synthesis process, while all other elements are primarily formed during the stellar evolution process. Metal is stored in galaxies either in the stars, or in the gas phase. The former can be observed through the absorption line in the stellar atmosphere, while the latter can be observed through the emission line emitted by these metal elements after they are excited to a higher energy level and decayed to the basic level. Here we only focus on the gas-phase metallicity, although stellar metallicity is also very interesting. For each galaxy, we can define a quantity called metallicity to describe the fraction of metal within the gas.
The gas-phase metallicity is regulated by many different processes. First of all, the formation and evolution of stars will consume gas and return metal-enriched gas, thus this process increases the metallicity. Secondly, most of the galaxies are accepted fresh gas from the inter-galactic space, and these gas is mostly metal-poor and able to dilute the gas-phase metallicity. Finally, there are various processes that drive gas out of the galaxy, like the feedback process and the gas-stripping processes. These processes can alter the metallicity depending on the spatial distribution of metallicity within the galaxy and the preference affected locations. For example, if the galaxy center is more metal-rich than the outskirt, the AGN feedback process will decrease the global metallicity since it preferentially ejects gas in the galaxy center, which is higher than the average metallicity in the galaxy, while the gas-stripping process will increase the global metallicity since the metal-poor gas on the galaxy outskirt is stripped.
On the opposite, the measurement of galaxy metallicity can help us constrain the above processes and understand the physics behind galaxy formation and evolution. In the previous decades, many models are proposed to explain the observed mass-metallicity scaling relation (MZR) in our local and distant Universe. However, the environmental dependence of this scaling relation is poorly studied, and this is our motivation to initiate this project.
Using the state-of-the-art cosmological hydrodynamical simulation of EAGLE, we found that, at a given stellar mass, low-mass galaxies living in massive halos have higher metallicity than their counterparts in low-mass halos. However, this trend is reversed for high-mass galaxies, where those in massive halos have lower metallicity. So for the IllustrisTNG simulation. This seemingly complicated environmental dependence becomes much clearer when galaxies are separated into centrals and satellites. Central galaxies, which dominate the high-mass end, are more metal-poor when they are living in more massive halos, while satellite galaxies, which dominate the low-mass end, are more metal-rich when found in massive halos. These trends preserve through $z\sim 2$ to our local Universe.
We further explored the underlying physics that drives this environmental dependence of MZR for central and satellite galaxies separately. For central galaxies, we found that the environmental dependence of MZR is primarily due to the excessive gas accretion at high-$z$. Those living in massive halos tend to accrete more metal-poor and pristine gas, thanks to the efficient code-mode accretion, and these gas can effectively dilute the metal content of the galaxy. At low $z$, the effect of this excessive accretion becomes subdominant, while the AGN feedback plays a crucial role here. Due to the negative metallicity gradient possessed by most of the central galaxies in the EAGLE simulation, AGN feedback preferentially ejects those metal-rich gas in the galaxy center and decreases the global metallicity consequently.
Satellite galaxies suffer from two additional environmental effects, compared with centrals. The first one is called strangulation, which means the gas accretion is terminated after a central galaxy becomes a satellite. Without the dilution of these accreted metal-poor gas, the metallicity of these satellite galaxies increases accordingly. The other effect is called ram-pressure stripping, which originates from the interaction between the cold interstellar medium (ISM) in satellite galaxies and hot intra-cluster medium (ICM) in the host halo. When the interaction becomes strong enough, the ISM can be stripped out of the satellite. Since the gas on the galaxy outskirt are the most loosely bound, they are preferentially stripped. Besides, since these gas are relatively metal-poor, their stripping can effectively increase the global metallicity.
The above analysis is based on hydrodynamical simulations and cannot be trusted until its prediction is verified observationally. At low $z$, the environmental dependence of satellite galaxies is consistent with the observation from the SDSS survey, while the environmental dependence for central galaxies requires more accurate and unbiased halo mass calibration. At high $z$, the predictions of the EAGLE and IllustrisTNG simulations are consistent with the observational result in Wang et al. 2022. More statistically robust test of these model predictions become possible along with the next-generation galaxy surveys, like MOONS and PFS.