1. Star Formation in the Present Day
Observations of our own Galaxy (the Milky Way) as well as the nearby galaxies that surround it, show us that new star formation happens slowly at present times. In spiral galaxies, the Milky Way included, the star-formation rate is of only a few stars like our Sun per year. This rate is very modest compared with the amount of stars that these galaxies have already assembled over cosmic history, which is typically tens or hundreds of billions of suns. This fact suggests that star formation activity must have been more important in the past, when the Universe was younger, because at the current formation rates galaxies would have never managed to build their current stellar content.
In the majority of star-forming galaxies, the production rate of new stars is directly related to the already assembled stellar mass, forming the so-called main sequence of star formation (e.g., Speagle et al. 2014; Fig. 1). This is simply a scaling relation: galaxies with large amounts of assembled stars also have larger cold gas reservoirs, allowing for higher star formation rates. Galaxies are expected to stay on the main sequence until their gas reservoirs are exhausted, or until flows of hot gas (coming from supernovae or active black-hole winds) prevent the gas from cooling down to form new stars. When this happens, galaxies become passive, as it is the case for red giant ellipticals in the nearby Universe.
On the contrary, galaxies with enhanced star formation activities, the so-called starbursts, are rare in the local Universe. Starbursts are defined as those galaxies whose birth rate parameter, i.e. the ratio between the ongoing star formation rate and the past average star formation rate, is large, namely b ≡ SFR/ < SFR >past≫ 1 (see Kennicutt 1983; Bergvall et al. 2016). This implies that their specific star formation rates (i.e., the star formation rates per unit of assembled stellar mass) are also large, which in turn means that their stellar-mass doubling times are short, of the order of 10 to 50 million years. The elevated star formation activity is expected to be temporary in these galaxies and can be triggered by galaxy-galaxy interactions (mergers) or internal instabilities. Starburst galaxies typically have irregular morphologies in the local Universe.
2. Star Formation in the First Half of Cosmic Time
Over the past two decades, a number of studies investigated the existence of distant starburst galaxies, in order to understand the importance of this phenomenon in the first 6-7 billion years of cosmic time (about a half of the Universe’s current age). Most of these studies were conducted with far-infrared telescopes, as the most vigorous star formation was hidden behind dust clouds in the past (Sanders & Mirabel 1996; Dole et al. 2006). Contrary to the expectations, these studies found a small fraction of starburst galaxies (Rodighiero et al. 2011; Sargeant et al. 2012). The most intense dusty star-forming galaxies are quite massive and, thus, follow the relation described by the main sequence of star formation.
More recently, other works extended the search for starbursts to smaller sources and the more distant Universe, probing galaxies 10-100 times smaller than the Milky Way, whose light was emitted in the first few billion years of cosmic time. This was possible thanks to the availability of deep astronomical images taken with the Hubble Space Telescope (HST) and the Spitzer Space Telescope, over blank and gravitationally lensed fields of the sky showing thousands of distant galaxies of different sizes. These studies recognised, for the first time, that starbursts did make a significant fraction of the star-forming galaxies present in the past Universe (Caputi et al. 2017), accounting for more than 50% of the cosmic star formation rate density in the first few billion years (Rinaldi et al. 2022). Since then, many other works found examples of starburst galaxies in the distant Universe (e.g., Vanzella et al. 2018; Caputi et al. 2021; Casey et al. 2021; Finkelstein et al. 2022).
In the vast majority of these studies, the starbursts were identified amongst line-emitting galaxies. This is perhaps not surprising, as star-forming galaxies are typically characterised by line emission in their spectra and the presence of such lines facilitates the detection of these sources (Sun et al. 2023). A more systematic investigation of the link between the starburst phenomenon, star formation and line emission in the young Universe has not been done until recently, in the JWST era (Caputi et al. 2024).
3. New Results from JWST
The unprecedented sensitivity and spectral coverage of JWST images are allowing us to investigate star formation activity and other properties in galaxies as small – or even smaller – than the Milky Way satellites, up to very large distances. In such a way, we can now study, for the first time, how the seeds of today’s bigger galaxies were formed in the young Universe.
3.1 Star formation in Lyman-α emitters
Lyman-α emitters, i.e., galaxies whose spectra show the Lyman-α transition line in emission, are some of the easiest-to-detect distant galaxies. This is because the Lyman-α line is emitted at λ wavelengths λem = 1216 Angstroms, so when it comes from very distant sources it reaches us redshifted to visible frequencies due to the expansion of the Universe (with λ = (1 + z)λem, where z is the redshift of the source). Thus, distant Lyman-α emitters can be identified in spectroscopic surveys conducted with traditional, ground-based telescopes, which operate at optical wavelengths. Tens of thousands of these galaxies were known even before the advent of JWST.
Instead, many of their most important properties were unknown in the pre-JWST era, especially for the most distant sources whose light was emitted in the first 1-2 billion years of cosmic time. In particular, one of the most important questions that astronomers had was if the Lyman-α emitters were really young galaxies forming their first stellar generations, or more evolved galaxies, whose newly formed stars coexisted with older ones formed in previous star-formation episodes.
This problem has recently been tackled by Iani et al. (2024) using JWST images taken with the Near-Infrared Camera (NIRCam; Rieke et al. 2005) and Mid InfraRed Instrument (MIRI; Rieke et al. 2015, Wright et al. 2015). They studied the stellar populations of 182 Lyman-α emitters identified in VLT/MUSE spectra, along with the stellar populations of 450 similarly distant galaxies without Lyman-α emission (the so-called Lyman-break galaxies), as control sample. They found that about 75% of the Lyman-α emitters are young galaxies with ages < 100 Myr, while the remaining 25% have older stellar populations, which suggests that they are experiencing a rejuvenation effect (Rosani et al. 2018). These older Lyman-α emitters are more massive and less dust-extinct than the younger ones, and are typically found on the star-formation main sequence. Instead, most of the young Lyman-α emitters are starburst galaxies. Remarkably, the properties of the control sample are quite similar: about a half of them are very young and classified as starbursts (based on their emitted UV luminosities), in spite of not having Lyman-α in emission, which is due to HI resonant scattering and/or dust-absorption.
3.2 Revealing Dust-obscured Star Formation in Optically Dark Galaxies
The new JWST data are also proving to be fundamental to understand the properties of galaxies discovered with previous infrared telescopes, but which are extremely faint (or completely undetected) at optical wavelengths, even in the deepest HST images. Early studies suggested that most of these sources were dusty star-forming galaxies in the distant Universe, whose light was emitted only a few billion years after the Big Bang, at redshifts z = 3 − 6 (e.g. Caputi et al. 2012, Sun et al. 2021). A number of recent JWST studies confirmed that this is the case (e.g. Barrufet et al. 2023, Williams et al. 2024). In some cases, JWST has enabled a detailed study of their internal dust distribution (e.g. Kokorev et al. 2023). Moreover, JWST has discovered many more, fainter examples of such galaxies (e.g. Nelson et al. 2023).
The dusty star-forming nature of these optically dark galaxies was confirmed in several cases with sub-/millimetre telescopes (e.g. Caputi et al. 2014; Smail et al. 2023). More recently, van Mierlo et al. (2024) studied a more extreme class of such sources, which were discovered in the mid-infrared, but extremely faint even at near-infrared wavelengths. Their JWST data analysis indicates that these sources are mostly prominent Hα emitters (indicative of new star formation activity), including a few which are bright sub-millimetre sources.
3.3 Star Formation around the Epoch of Reionization
With no doubts, one of the most important steps forward in the JWST era is the possibility of investigating star formation activity in galaxies around the Epoch of Reionization, when the atomic hydrogen present in the intergalactic medium became progressively ionised by the ultraviolet photons produced by the earliest stars and galaxies formed. Cosmic Reionization was a process that lasted several hundred million years and occurred within the first billion year of cosmic time. The exact nature of the sources of Reionization is still under debate, but a number of recent studies are now providing clear hints on their properties and their relation to the general population of star-forming galaxies present at those early cosmic times.
Numerous line emitting galaxies have been identified with JWST, in most cases Hβ+[OIII] emitters, towards the Epoch of Reionization, confirming that optical line emission was quite common even in galaxies with a damped Lyman-α line (e.g. Bunker et al. 2023). A major success was the finding of Hα emitters at these cosmic epochs for the first time (Rinaldi et al. 2023), as Hα is one of the most secure tracers of star formation activity. This discovery was possible thanks to the extended wavelength coverage provided by MIRI on JWST. Only MIRI can see the Hα line at such early cosmic times. In general, line emitting galaxies, either identified via spectroscopy or photometry, have been proposed as sources of Reionization. The corresponding line fluxes allow for an estimate of the photon production efficiency which, coupled with constraints on the escape fraction, indicates whether these sources are effective ionizers or not (e.g., Endsley et al. 2023, Rinaldi et al. 2024).
In parallel, other studies based on JWST data proposed that the process of Reionization could have been driven by low-stellar-mass, young starburst galaxies (Endsley et al. 2024, Simmonds et al. 2024), as theoretically proposed by Sharma et al. (2017). Starburst galaxies and strong line emitters have some properties in common. However, their relation at the Epoch of Reionization was not laid out until very recently.
In spite of their enhanced detectability, strong line emitters constitute less than a third of all star-forming galaxies at the Epoch of Reionization (Rinaldi et al. 2023, Caputi et al. 2024), which indicates that these emitters alone do not provide a full census of the cosmic star formation activity in the early Universe. Interestingly, Caputi et al. (2024) also showed that a significant fraction of the prominent line emitters are starburst galaxies, but not all of them are. And the same happens with the star-forming galaxies which are not prominent line emitters. So, this suggests that all starbursting galaxies may have had a role in the process of Reionization, but it is unclear when their main contribution happened — either at the beginning of the starburst phase, during the phase of intense line emission, or after, when all the photons associated with the new star formation were emitted in the UV, qualifying as reionizing photons.
4. Conclusions
JWST is revolutionizing our understanding of the process of star formation in galaxies across cosmic times in two main ways: i) by discovering star formation activity down to very small galaxies, like the Milky Way satellites and even smaller units, which may simply be stellar aggregates on the way of galaxy formation; ii) by revealing the link between star formation activity and spectral line emission, and in turn the implications for cosmic Reionization. The pending challenges for the near future are to: establish whether prominent line emitters or non-emitter starbursts were the main sources of reionization; achieve a full understanding of the underlying physics that characterized the star formation process in these small systems.
References
Barrufet, L. et al., 2023, MNRAS, 522, 449.
Bergvall, N. et al., 2016, A&A, 587, A72.
Bunker, A. et al., 2023, ApJ submitted (arXiv:2306.02467).
Caputi et al., 2012, ApJL, 750, L20.
Caputi, K.I. et al., 2014, ApJ, 788, 126.
Caputi, K.I. et al., 2017, ApJ, 849, 45.
Caputi, K.I., 2021, ApJ, 908, 146.
Caputi, K.I., 2024, ApJ, submitted (arXiv: 2311.12691).
Casey, C. et al., 2021, ApJ, 923, 215.
Dole, H. et al., 2006, A&A, 451, 417.
Endsley, R. et al., 2023, MNRAS, 524, 2312.
Endsley, R. et al., 2024, MNRAS submitted (arXiv:2306.05295).
Finkelstein, S.L. et al., 2022, ApJ, 928, 52.
Iani, E. et al., 2024, ApJ submitted (arXiv:2309.08515).
Kennicutt, R.C., Jr., 1983, ApJ, 272, 54.
Kokorev, V. et al., 2023, ApJ, 945, L25.
Nelson, E. et al., 2023, ApJ, 948, L18.
Rieke, G.H. et al., 2015, PASP, 127, 584.
Rieke, M.J. et al., 2005, SPIE, 5904, 1.
Rinaldi, P. et al., 2022, ApJ, 930, 128.
Rinaldi, P. et al., 2023, ApJ, 952, 143.
Rinaldi, P. et al., 2024, ApJ in press (arXiv: 2309.15671).
Rodighiero, G. et al., 2011, ApJ, 739, L40.
Rosani, G. et al., 2020, A&A, 633, A159.
Sanders, D.B. & Mirabel, I.F., 1996, ARA&A, 34, 749.
Sargent, M.T. et al., 2012, ApJ, 747, L31.
Sharma, M. et al., 2017, MNRAS, 468, 2176.
Simmonds, C. et al., 2024, MNRAS, 527, 6139.
Smail, I. et al., 2023, ApJ, 958, 36.
Speagle, J.S. et al., 2014, ApJS, 214, 15.
Sun, F. et al., 2021, ApJ, 922, 114.
Sun, G. et al., 2023, MNRAS, 526, 2665.
van Mierlo, S. et al., 2024, ApJ submitted (arXiv:2311.17764).
Vanzella, E. et al., 2018, MNRAS, 476, L15.
Williams, C. et al., 2024, ApJ submitted (arXiv:2311.07483).
Wright, G.S. et al., 2015, PASP, 127, 595.