MIGHTEE-HI: The star-forming properties of HI selected galaxies

Technical Summary: MIGHTEE-HI: The star-forming properties of Hi selected galaxies

Problem Statement
The population of galaxies exhibits a bimodality between blue, star-forming systems and red, quenched galaxies. While the processes driving galaxies from the star-forming main sequence (MS) to the red sequence are complex, involving feedback, mergers, and environmental effects, the specific role of atomic hydrogen (Hi) reservoirs and large-scale cosmic environments (specifically filaments) in shaping star formation histories (SFHs) remains debated. Previous studies have yielded conflicting results regarding the correlation between Hi content and proximity to cosmic filaments, and the mechanisms driving the scatter in the star-forming MS are not fully understood. This paper aims to decouple these dependencies by analyzing the SFHs of Hi-selected galaxies to understand the interplay between gas content, stellar mass, and the large-scale environment.

Methodology
The study utilizes the Early Science Release data from the MIGHTEE-Hi survey, covering the COSMOS and XMM-LSS fields. The sample consists of 187 Hi-selected galaxies (reduced from an initial 276 due to photometric coverage limits) with redshifts $0.004 < z < 0.084$.

1. Data Integration: The Hi spectral line data (L-band, 1320–1410 MHz) was cross-matched with the Deep Extragalactic VIsible Legacy Survey (DEVILS) photometric catalogue. This provided consistent ultraviolet to far-infrared data (GALEX, CFHT, HSC, VISTA, Spitzer, Herschel) necessary for robust Spectral Energy Distribution (SED) fitting.
2. SED Fitting: The authors employed Bagpipes, a Bayesian inference tool, to fit model SEDs to the observed photometry. They utilized the Bruzual & Charlot (2003) stellar population models with a Chabrier (2003) IMF and the Calzetti et al. (2000) attenuation law.
3. SFH Modeling: Six parametric SFH models were tested (exponential, constant, log-normal, delayed exponential, burst, and double-power law). Using Bayes factors and the Jeffreys scale, the authors determined that the broadband data could not significantly differentiate between the exponential, log-normal, and exponentially-delayed models. Consequently, the exponentially-delayed SFH model was adopted for the primary analysis.
4. Environmental Analysis: The large-scale environment was characterized using the Disperse algorithm to identify the cosmic web skeleton (filaments) based on galaxy distributions. The study examined the alignment between galaxy spin vectors and filament axes, as well as the distance of galaxies to the nearest filament.

Key Contributions and Results

• Star Formation Histories and Mass Relations:

  • A strong positive correlation was found between the Hi-to-stellar mass ratio ($M_{\text{Hi}}/M_*$) and the time of formation ($t_{\text{form}}$), defined as the cosmic time when the galaxy reached its peak star formation.
  • Conversely, a strong anti-correlation exists between stellar mass ($M_*$) and $t_{\text{form}}$. This indicates that more massive galaxies formed the bulk of their stars earlier in the universe's history, consistent with the "cosmic downsizing" scenario.
  • No significant correlation was found between the absolute Hi mass and the time of formation.

• Gas Depletion Timescales ($t_{\text{dep}}$):

  • The atomic gas depletion time ($t_{\text{dep}} = M_{\text{Hi}}/\text{SFR}$) shows a positive correlation with the Hi-to-stellar mass ratio and a negative correlation with stellar mass.
  • Lower-mass, gas-rich galaxies exhibit significantly longer depletion times compared to massive galaxies. The authors interpret this as evidence that smaller galaxies have shallower potential wells, leading to less efficient conversion of Hi to stars and greater susceptibility to gas loss via feedback.
  • Partial correlation tests identified stellar mass as the primary driver of gas depletion time, rather than the star formation rate (SFR) or Hi mass alone.
  • No statistically significant link was found between $t_{\text{dep}}$ and specific star formation rate (sSFR).

• Main Sequence Scatter:

  • The Hi-selected sample predominantly lies on or above the star-forming main sequence.
  • The authors argue that the observed scatter in the main sequence cannot be explained solely by short-term fluctuations in gas depletion, given the long timescales involved ($\sim$Gyr). Instead, the scatter likely arises from long-term environmental effects or processes that alter the efficiency of gas conversion over extended periods.

• Environmental and Spin-Filament Alignment:

  • General Trends: No significant correlation was found between the distance to the nearest cosmic filament and either the formation time ($t_{\text{form}}$) or the gas depletion time ($t_{\text{dep}}$).
  • Spin-Filament Misalignment: The study revisited the finding that galaxies with spin axes misaligned with filaments have higher Hi-to-stellar mass ratios.
  • Merger Hypothesis: While gas-rich mergers are a theoretical explanation for spin misalignment and increased Hi mass, visual inspection of the sample revealed only 7 galaxies with morphological merger signatures, none of which showed a clear link between misalignment and recent starburst activity in the SEDs.
  • Filament Proximity Anomaly: A specific sub-sample of two galaxies located within 1 Mpc of a filament spine exhibited very low gas depletion times and significant spin-filament misalignment. The authors suggest this may indicate a link between proximity to filaments, spin disruption, and enhanced star formation, potentially driven by gas-rich mergers, though the small sample size prevents definitive conclusions.

Significance and Claims
The paper claims to provide a detailed characterization of the star formation properties of Hi-selected galaxies using unprecedented multi-wavelength data. Its primary significance lies in:

1. Decoupling Dependencies: Establishing that stellar mass is the dominant factor driving gas depletion times and formation epochs in Hi-selected samples, independent of the specific SFH model used.
2. Efficiency Differences: Demonstrating that massive galaxies are more efficient at converting gas into stars and have already depleted a larger fraction of their reservoirs compared to low-mass, gas-rich systems.
3. Environmental Context: While finding no broad correlation between filament proximity and global star formation timescales, the paper highlights a potential, albeit tentative, connection between extreme filament proximity, spin misalignment, and rapid gas consumption in specific cases.

The authors conclude that the scatter in the star-forming main sequence is likely driven by long-term environmental effects rather than short-term gas fluctuations, and that while Hi-selected samples offer a unique view of gas-rich systems, they are biased against low-mass Hi-deficient galaxies and dense environments where gas stripping is prevalent. Future work with larger samples and spectroscopic data (e.g., from ORCHIDSS) is required to confirm the link between spin-filament alignment and merger-driven star formation.