Table of Contents
Fetching ...

Episodic Star Formation -- I. Overview and Scatter of the Star-Forming Main Sequence

Yuqian Gui, Dandan Xu, Haoyi Wang, Xuelun Mei, Enci Wang, Cheng Li, Stijn Wuyts

TL;DR

This study investigates episodic star formation as a key source of scatter in the star-formation main sequence (SFMS) by tracing z=0 star-forming centrals in the TNG100 simulation back to z~1. It reveals a two-branch SF pattern within each episode, with an inner, high-metallicity central branch and an outer, fresh-gas branch that forms at large radii and retreats inward as activity peaks, while outer-regions exhibit stronger SFR variability. The authors quantify that ~0.2 dex temporal SFR fluctuations within individual galaxies plus ~0.15 dex inter-galaxy differences can together explain the ~0.25 dex SFMS scatter observed today. They also document distinct metallicity and structural signatures between SFR peaks and valleys, including bimodal distributions of young-star metallicity and gas-phase metallicity linked to the two branches. The findings imply that episodic, radially propagating star formation coupled to cold-gas replenishment and feedback can account for current SFMS scatter and motivates further exploration of the fundamental metallicity relation and cold-gas dynamics in subsequent papers.

Abstract

Episodic star formation cycles in both high- and low-redshift galaxies have gained more and more evidence. This paper aims to understand the detailed physical processes behind such behaviors and investigate how such an episodic star-forming scenario can explain the scatter in star-formation rate (SFR) of star-forming main-sequence galaxies. This is achieved through tracing back in time the history of z=0 star-forming central galaxies in the TNG100 simulation over the past 7-8 Gyrs. As the first paper in this series, we provide an overview of the episodic star formation history. We find that two branches of star formation typically develop during each episode: while one branch happens in heavily metal-enriched gas in the centers of galaxies, a secondary branch starts in lower-metallicity regions at galaxy outskirts where fresh gas first arrives, and gradually progresses to inner regions of galaxies. Additionally, the temporal variation in the SFR at galaxy outskirts is more significant than that at centers. As a consequence, the metallicities in both gas and young stars exhibit remarkably different distributions between SFR peaks and valleys. The resulting temporal SFR fluctuation within individual galaxies has an average of ~ 0.2 dex, while the intrinsic differentiation between (the historical mean of) galaxies is ~ 0.15 dex. These two together can well account for the scatter in SFR of ~ 0.25 dex as observed for z=0 star-forming main-sequence galaxies.

Episodic Star Formation -- I. Overview and Scatter of the Star-Forming Main Sequence

TL;DR

This study investigates episodic star formation as a key source of scatter in the star-formation main sequence (SFMS) by tracing z=0 star-forming centrals in the TNG100 simulation back to z~1. It reveals a two-branch SF pattern within each episode, with an inner, high-metallicity central branch and an outer, fresh-gas branch that forms at large radii and retreats inward as activity peaks, while outer-regions exhibit stronger SFR variability. The authors quantify that ~0.2 dex temporal SFR fluctuations within individual galaxies plus ~0.15 dex inter-galaxy differences can together explain the ~0.25 dex SFMS scatter observed today. They also document distinct metallicity and structural signatures between SFR peaks and valleys, including bimodal distributions of young-star metallicity and gas-phase metallicity linked to the two branches. The findings imply that episodic, radially propagating star formation coupled to cold-gas replenishment and feedback can account for current SFMS scatter and motivates further exploration of the fundamental metallicity relation and cold-gas dynamics in subsequent papers.

Abstract

Episodic star formation cycles in both high- and low-redshift galaxies have gained more and more evidence. This paper aims to understand the detailed physical processes behind such behaviors and investigate how such an episodic star-forming scenario can explain the scatter in star-formation rate (SFR) of star-forming main-sequence galaxies. This is achieved through tracing back in time the history of z=0 star-forming central galaxies in the TNG100 simulation over the past 7-8 Gyrs. As the first paper in this series, we provide an overview of the episodic star formation history. We find that two branches of star formation typically develop during each episode: while one branch happens in heavily metal-enriched gas in the centers of galaxies, a secondary branch starts in lower-metallicity regions at galaxy outskirts where fresh gas first arrives, and gradually progresses to inner regions of galaxies. Additionally, the temporal variation in the SFR at galaxy outskirts is more significant than that at centers. As a consequence, the metallicities in both gas and young stars exhibit remarkably different distributions between SFR peaks and valleys. The resulting temporal SFR fluctuation within individual galaxies has an average of ~ 0.2 dex, while the intrinsic differentiation between (the historical mean of) galaxies is ~ 0.15 dex. These two together can well account for the scatter in SFR of ~ 0.25 dex as observed for z=0 star-forming main-sequence galaxies.

Paper Structure

This paper contains 16 sections, 11 figures.

Figures (11)

  • Figure 1: The star formation history of one of our sample galaxies (ID at $z=0$ is 504650). The red line represents the SFR of this galaxy evaluated within $2R_{\mathrm{hsm}}$. The green and blue upper triangles represent the valleys and peaks in the upward pairs. The blue and green hollow hexagons represent peaks and valleys in the downward pairs.
  • Figure 2: The $g-r$ color - log $M_\ast$ evolutionary track for one example galaxy (ID-433317). The contours indicate the distribution of color as a function of stellar mass $M_\ast$ of central galaxies in today's universe (red contour) and at $z$ = 0.7 (green contour), respectively. The black solid line marks the evolutionary track of the galaxy, with circles color-coded by redshifts.
  • Figure 3: The episodic star-formation history and property evolution for one sample galaxy (ID-601819). The top panel presents the temporal variation of ${\rm SFR}_{\leqslant 2R_{\mathrm{hsm}}}$ (global, in solid line) and ${\rm SFR}_{\leqslant 0.5R_{\mathrm{hsm}}}$ (central, in dashed line). The second panel presents the evolution of cold gas mass within $2R_{\mathrm{hsm}}$ for the star-forming gas ($M_{\mathrm{SF}}$, solid line) and the cold non-SF gas ($M_{\mathrm{cold,\,non-SF}}$, dashed line). The third panel shows the metallicity evolution for both star-forming gas and cold non-SF gas, denoted as $Z_{\mathrm{SF}}$ (solid) and $Z_{\mathrm{cold,\,non-SF}}$ (dashed). The bottom panel presents the distribution of newly-formed stars at redshift $z$ with stellar metallicity $Z_{\ast}$. The color represents the mass-weighted average distance (from the galaxy center, normalized by $R_{\mathrm{hsm}}$). We only plot those pixels that contain at least three stellar particles to avoid shot noise.
  • Figure 4: The relative mass $\Delta \log M$ within $2R_{\rm hsm}$ of both star-forming gas (blue) and cold non-SF gas (magenta) as a function of time offset around the nearest peak/valley moment. In both panels, the shaded regions represent the central 68th percentile distribution. For clarity of visualization, the red dashed line marks the median SFR within $2R_{\mathrm{hsm}}$.
  • Figure 5: The histograms of the increment in normalized half-stellar-mass radius of young stars $R_{\mathrm{hsm,\,<100\,Myr}}/R_{\mathrm{hsm}}$ from the $(i-1)^{\rm th}$ snapshot to the $(i+1)^{\rm th}$ snapshot, where $i$ is a snapshot at either SFR valley or peak. The blue and red histograms represent the increment in normalized size around peaks and valleys, respectively. The dashed lines and shaded regions represent the corresponding median and central 68th percentile in each histogram. The black dashed line indicates the value of zero (representing no change in $R_{\mathrm{hsm,\,<100\,Myr}}/R_{\mathrm{hsm}}$). Typically, the disk size of young stars grows around the time of the valley and becomes smaller around the time of the peak.
  • ...and 6 more figures