Michael Jones

When galaxies first began forming stars en masse the UV radiation that they emitted heated and ionized most of the neutral gas in the universe in an era known as the epoch of reionization (re- because immediately after the big bang the universe was entirely ionized, then cooled and became neutral). For the lowest mass galaxies reionization eliminated their entire cold gas reservoir and prevented them from forming new star ever again. At just marginally higher mass are the lowest mas star-forming galaxies in the universe, that were just large enough to hang on to some of their gas through reionization. However, these galaxies are extraordinarily faint and difficult to identify, so very few are known. Furthermore, if they fall into the gravitational well of a massive galaxy, like our own Milky Way, then they will rapidly be stripped of their gas and stop forming stars. Thus, to find these galaxies you have to look beyond our Local Group of galaxies.

I led a novel search effort (Jones et al. 2023) to identify such galaxies in the DESI Legacy Imaging Surveys that combined established low surface brightness galaxy search algorithms and a machine learning candidate classifier. In this search we identified Pavo, one of the lowest mass star-forming galaxies known. Pavo is only 2 Mpc (6.5 Mly) away from us, but is remarkably isolated with absolutely no apparent neighbours. This relative proximity to us, combined with its isolation, make Pavo an excellent target to study how star formation proceeds in the lowest mass galaxies, and to study the impact that reionization had on star-forming in the early universe. Did it only just survive reionization, or was it more of a speed bump?

DESI Legacy Imaging Survey cutout of Pavo. The blue light tracers only the youngest stars, but Pavo extends much further as a faint, almost imperceptible distribution of redder stars. These red stars are the ancient population formed in the early universe.

Galaxy clusters are filled with an intra-cluster medium of hot gas that makes them an extremely hostile environment for any galaxy containing the cool gas needed to form stars. Even fairly massive galaxies like our own Milky Way would be rapidly stripped of gas when they fall into a galaxy cluster. Hence our surprise when in Jones et al. 2022 we identified a new class of isolated stellar system in the Virgo cluster, which are typically 1 million times less massive that the Milky Way, but are made up entirely of recently formed stars. These objects are also very metal-rich for their mass. Generally elements heavier than helium are built up in galaxies through successive episodes of star formation. Some of these blue stellar systems have metallicities comparable to the Milky Way, indicating that they formed from gas stripped from galaxies similar to our own (as they could never have built up some many metals on their own).

Even if these "blue blobs" did form from stripped gas, the question remains, what process stripped the gas and how did they get so isolated? Gas is stripped from galaxies in two ways, tidal stripping and ram pressure stripping. Tidal stripping occurs due to the gravity of another galaxy passing nearby that pulls the outskirts of the first galaxy towards it. Ram pressure stripping occurs when a gas-rich galaxy is travelling at high speed through a hot gas medium (like in a galaxy cluster) and the pressure of this gas forces out its own gas. This is like running into the wind or belly flopping into a swimming pool, except your body is solid, not made of gas. Although either of these mechanisms could be responisble for creating the blue stellar systems we discovered, only ram pressure stripping can explain their isolation because it occurs at such high speeds. This way these objects can become separated from their parent galaxies by great distances while they are still very young.

Left: A Hubble Space Telescope image of one of the young, isolated stellar systems in Virgo, named BC5. It has a clumpy and irregular structure and appears to consistent of only young blue stars. Right: An image of the same field but in UV (from GALEX), indicating where star formation has occurred recently.

Ultra-diffuse galaxies, or UDGs, are a class of recently identified extremely low surface brightness (LSB) galaxies which have been principally found in galaxy clusters. While astronomers have been studying LSB galaxies for decades, including some which are now classified as UDGs, the striking results of van Dokkum et al. 2015 and Koda et al. 2015 show just how prevalent they are, even at the lowest surface brightnesses and in dense cluster environments. There are many open questions about the formation and survival of these galaxies and their connection to the other LSB galaxies and normal galaxies.

While the vast majority of UDGs are red, spheroidal objects located in or around groups and clusters, Leisman et al. 2017 identified a sample of isolated, blue, HI-bearing UDGs (HUDs) that were strongly detected as HI sources in ALFALFA, but have almost invisible optical counterparts in the Sloan Digital Sky Survey. In Jones et al. 2018c we measured the cosmic number density of these source by comparison with the global ALFALFA population, and found that although many fewer HUDs than red UDGs have been found to date, their overall abundances are likely quite similar. However, the connection between HUDs and red UDGs remains uncertain; they may be two phases of one population or two completely separate populations.

I am the PI of a Hubble Space Telescope snapshot program to investigate the the globular systems of HUDs and compare them to both those of normal galaxies and UDGs in denser environments. As globular clusters are thought to form early on in a galaxy's life cycle a comparison of these systems should help to reveal whether there is a deeper connection between HUDs and other UDGs, or if they really are separate populations with separate formation pathways.

Left: Two examples of HUDs from Leisman et al. 2017, showing that they are almost invisible in SDSS images (top row), but are confidently detected in CFHT imaging and show more complex structure. Right: Figure from Jones et al. 2018c showing that the cosmic abundance of HUDs (green band) in similar to that of red UDGs found in groups and clusters (grey band). The latter was calculated by integrating the relationship between halo mass and number of UDGs (measured by van der Burg et al. 2017) over the halo mass function (from high to low masses).

The AMIGA project (Analysis of the interstellar Medium of Isolated GAlaxies) focuses on a population of carefully selected, isolated galaxies and has collected a rich multi-wavelength dataset to characterise their properties. This sample defines a benchmark for galaxies that reside in an almost nurture-free environment and thus acts as a control sample for studies of the impact on interactions on galaxy properties, whether it be star formation rates, gas content, AGN activity, or numerous others. In addition, these galaxies form an ideal laboratory for investigating secular evolution in galaxies, as the influence of neighbours is minimised.

In Jones et al. 2018a I used the global HI profiles of several hundred isolated galaxies, observed by the AMIGA team or gathered from the literature, to define a new standard for the HI content of galaxies in the absence of interactions. The resulting optical to HI scaling relations had not been updated for the most isolated galaxies since Haynes & Giovanelli 1984, with most of the other recent scaling relations being measured with sources from HI surveys, which are intrinsically biased towards gas-rich galaxies. The increased sample size, morphological diversity, and improved maximum likelihood fitting method also gave more self-consistent relations than found previously.

Hickson Compact Groups (HCGs) are extremely dense groups of 4-10 galaxies in the local Universe. HCG 16 is a prototypical case of a compact group of galaxies in an intermediate stage of evolution, where its neutral gas has been heavily disturbed, but not yet lost or consumed (as hypothesised by Verdes-Montenegro et al. 2001). The group is made up of two spirals with low star formation rates, and two star bursting lenticulars. There is a 5th galaxy (NGC 848) to the south-east which optically might not appear to have interacted with the core group, but in HI is clearly connected to it by a ~160 kpc long tidal tail.

Although most of the extended HI gas resides in the vicinity of the two lenticular galaxies, comparison with scaling relations reveal that neither of these are deficient in HI. In fact the only HI-deficient galaxies in the group are the two spirals, which have both lost over 90% of their expected HI content. This led us to conclude in Jones et al. 2019 that most of the extended HI was likely disrupted by the passage of NGC 848, but was probably already loosely bound due to the ongoing interaction between the two spirals. Furthermore, the tidal tail to the SE is actually considerably longer than it appears and forms a continuous structure from NGC 848 all the way to the NW corner of the group. Finally, there is a strong candidate tidal dwarf galaxy between the two galaxy pairs, that is also right on the boundary of being classified as an ultra-diffuse galaxy.

An interaction 3D representation of the gas in HCG 16 can be found here.

Bottom Left: Map of the integrated HI emission in the group. Bottom Right: Optical r-band image of the group from DECaLS. Top: A segmented position-velocity slice (shown with the red line in the lower panels) through the entire group. A continuous structure can be seen stretching from NGC 848 to the NW corner of the group.

During my PhD I worked with Professors Martha Haynes and Riccardo Giovanelli on the ALFALFA HI survey, a blind 21 cm survey out to a maximum redshift of 0.06 and covering 7,000 square degrees of the sky. ALFALFA represents the cutting edge of large area, blind HI surveys. It used the Arecibo radio telescope in Puerto Rico, the largest single dish telescope in the world at the time, and a 7 pixel feed array (ALFA). The final HI source catalogue contains over 30,000 extragalactic source and can be found here.

My research focused on the global properties of the survey, investigating the statistical trends and biases present in the dataset. This included a quantification of the impact of source confusion on current and future HI surveys (Jones et al. 2015) and the construction of a generic model for the contribution of confusion noise to deep HI stacking experiments (Jones et al. 2016a). My cloud-based tool for predicting the level of source confusion in a generic stacking experiment can be found here. I also investigated the role environment plays in influencing the shape of the HI mass function of galaxies (HIMF), finding that the ‘knee’ mass increases towards denser regions, but not finding any evidence of a change in the low-mass slope (Jones et al. 2016b).

An SDSS DR10 optical image (left) of two HI galaxies detected in ALFALFA. The Elliptical galaxy is not detected. The ALFA beam is shown as a dark circle. The HI spectra of the two galaxies taken by ALFALFA (right). The excess emission present in the offset spectrum, associated with the profile of the lower galaxy, is a telltale sign of confusion.

My work on ALFALFA, and in particular the HIMF, has continued as a post-doc. In Jones et al. 2018b I calculated the HIMF for the final ALFALFA sample (Haynes et al. 2018). This is the most precise measurement of the HIMF to date. However, while the ‘knee’ is determined over a cosmologically fair volume and varies little across the survey footprint, the low-mass slope is only measurable in the relatively nearby Universe (D < 60 Mpc) with ALFALFA. Hence we see large shifts in the gradient of the low-mass slope in different regions of the sky. There is tentative evidence that the low-mass slope steepens within the large scale overdensity surrounding the Virgo cluster, perhaps due to the presence of gas-rich inflowing filaments, and flattens in the void in the foreground of the Pisces-Perseus supercluster. However, more investigation is required.

The ALFALFA HIMF in the Northern Spring and Fall sky regions of the survey (left). There is a clear shift in the low-mass slope of the HIMF in these two halves of the survey, but the 'knee' mass is essentially unchanged. The plot on the right shows the 2-sigma errors bars and ellipses of various measurements of the HIMF. On key result here is that no sample has an low-mass slope similar to that of individual groups that have been studied.

I am the co-PI (along with Rebecca Koopmann) of the Arecibo Pisces-Perseus Supercluster Survey (APPSS). This is an ALFALFA follow-up program to observe dwarf galaxies, below the ALFALFA sensitivity limit, in order to increase the source density of galaxies with known redshifts and velocity widths in the vicinity of PPS. With this dataset we aim to measure with high signal-to-noise the radial inflow of galaxies onto the supercluster filament through the use of the baryonic Tully-Fisher relation. The majority of the work for this project is being conducted by the student and faculty members of the Undergraduate ALFALFA Team (UAT, organised by Martha Haynes and Rebecca Koopmann), a network of undergraduate-focused institutions across the USA that contribute and gain research experience within the wider ALFALFA project.