Extragalactic astronomer · Caltech/IPAC

Michael G. Jones

I study gas and star formation in low-mass galaxies.

Email CV (PDF) Publications (SciX) ORCID GitHub

mgjones@ipac.caltech.edu

About

Staff Scientist, NASA/IPAC Infrared Science Archive (IRSA)

Portrait of Mike Jones

I'm an extragalactic astronomer and a staff scientist in the NASA/IPAC Infrared Science Archive (IRSA) at Caltech/IPAC. My research focuses on the lowest-mass galaxies in the nearby Universe: how we find them, how they form stars, and how they lose their gas. Small galaxies are among the most sensitive means we have to test models of galaxy formation and cosmology. At IRSA I also work on making the archive easier to use, including prototype large language model assistants.

Before joining IPAC in 2025 I was a postdoc with David Sand at the University of Arizona, and before that a Juan de la Cierva fellow at the Instituto de Astrofísica de Andalucía in Granada, where I worked with Lourdes Verdes-Montenegro. I received my PhD from Cornell University in 2016, working with Martha Haynes and Riccardo Giovanelli on the ALFALFA survey, and my undergraduate degree in Natural Sciences (Astrophysics) from the University of Cambridge. Outside work I hike, cycle, and take photographs, including some astrophotography.

Current research

Pavo in DESI Legacy Surveys imaging. The blue knots are young stars; the faint red haze that extends much further is the ancient stellar population.
Pavo in DESI Legacy Surveys imaging. Its brightest stars are almost resolved into individual point sources, giving it the characteristic mottled, semi-resolved appearance that SEAMLESS searches for.

SEAMLESS

SEmi-Automated Machine LEarning Search for Semi-resolved galaxies

A machine-learning search of the DESI Legacy Surveys for faint dwarf galaxies just beyond the Local Group.

Our current knowledge of the faintest galaxies is still dominated by the Milky Way's own satellites. Those galaxies have all lived inside a massive halo, so their gas, star formation, and structure have been shaped by their host. To learn how the smallest galaxies evolve on their own, we need to find them in isolation, and that means looking beyond the Local Group.

Read more

Searching for dwarf galaxies through concentrations of resolved stars work well within about 1–1.5 Mpc, and searches for diffuse, low surface brightness light work well at much larger distances. In between, at roughly 1.5–4 Mpc, a faint dwarf is semi-resolved: its brightest stars appear as a scatter of point-like sources sitting on top of a faint haze of unresolved light. Both classic methods tend to miss or overlook these objects.

SEAMLESS is built for this semi-resolved regime. We run a low surface brightness detection pipeline, adapted from SMUDGes, over the public DR10 imaging of the DESI Legacy Imaging Surveys. A convolutional neural network, retrained specifically to identify semi-resolved objects, then scores each candidate, and the team visually inspects the most promising ones before follow-up with Gemini, Magellan, HST, GBT, VLA, and MeerKAT.

The first discovery was Pavo, one of the lowest-mass star-forming galaxies known, at distance of about 2 Mpc and with no known neighbours. Its proximity and isolation make it an ideal place to study how cosmic reionization and stellar feedback regulate star formation in the smallest galaxies. SEAMLESS has since discovered other isolated, star-forming dwarfs such as Corvus A, and quenched dwarfs such as Hydrus A.

All SEAMLESS papers on SciX →

149 new HST orbits
111 archival satellites
≈260 satellites
22 Milky Way-like hosts
10 Mpc distance limit

Hubble Census of Nearby Satellites

An HST Treasury program to image every satellite brighter than MV = −9 around a Milky Way-like galaxy within 10 Mpc.

Until recently the Local Group was the only place where the census of the satellite populations was relatively complete down to faint satellites. Ground-based surveys, in particular the ELVES project, have now pushed that completeness out to a few dozen Milky Way-like systems within the Local Volume, but robust confirmation of these satellites and measuring their properties needs the resolution of Hubble.

Read more

The Hubble Census of Nearby Satellites (HCNS) is a 149-orbit Treasury program that is building on the ELVES legacy data set by imagining 149 satellites with WFC3/UVIS and ACS/WFC in F606W and F814W. Combined with 111 satellites that already have archival HST imaging, it forms a volume-limited sample of about 260 satellite candidates around 22 Milky Way-like hosts.

With this sample we will measure the most robust satellite luminosity function to date, extend the globular cluster luminosity function to lower galaxy masses than ever before, and derive star formation histories and quenching times for an unprecedented number of satellites.

HST image of BC5, one of the blue blobs in Virgo. It is clumpy and irregular, and consists entirely of young blue stars.
HST image of BC5, one of the blue blobs in Virgo. It is clumpy and irregular, and consists entirely of young blue stars.

Blue blobs and dark clouds in Virgo

Young, isolated stellar systems and almost dark gas clouds formed from stripped gas in the Virgo cluster.

Galaxy clusters are a hostile place for cold gas: the hot intracluster medium (ICM) strips gas from galaxies via ram pressure as they fall into the potential well of the cluster. It was therefore remarkable when we identified a new class of object in the Virgo cluster: tiny, stellar systems, made up entirely of young, metal-rich stars, isolated from any nearby galaxy, but embedded in the hot ICM. Their metallicity means they formed from gas stripped out of a massive galaxy, and their isolation points to ram pressure stripping, which can occur at very high velocities. Yet how they survive to form new stars while bathed in the hot ICM is not yet well understood.

Read more

These "blue blobs" are related to the dark HI clouds in Virgo, gas clouds that seem to have no stars. We found the very faint stellar counterpart (BC6) of one such cloud that had been thought to be entirely dark since its discover almost two decades prior, providing a link between the two populations. Volunteers in our Zooniverse project made more than 600,000 classifications to finding new blue blobs candidates, some of which we were able to match to known "dark" clouds through spectroscopic observations with the Hobby-Eberly Telescope. We are now expanding our follow up programs with ALMA (molecular gas), GBT (HI), and a new 26-hour JWST program (GO-10165) for deep NIRCam and MIRI imaging.

All blue blobs papers on SciX →

HI contours over optical imaging of LS-359104-131, a satellite of NGC 7541, showing a long one-sided gas tail.
VLA C+D HI contours overlaid on DESI Legacy Surveys imaging of LS-359104-131, a satellite of NGC 7541. The gas extends into a long, one-sided tail, a likely sign that it is being ram pressure stripped.

HI-SAGA

HI in Satellites Around Galactic Analogs

Measuring the neutral gas in satellites of Milky Way analogs to learn how and when satellites stop forming stars.

Almost all of the Milky Way's satellites are gas-poor and no longer forming stars, but the SAGA (Satellites Around Galactic Analogs) survey found that satellites of other Milky Way-like galaxies are often star-forming. Is our own system an outlier among other systems of similar mass, or are we missing part of the picture?

Read more

HI-SAGA is a spin-off of SAGA that measures the neutral hydrogen (HI) content of SAGA satellites. Gas is the fuel for star formation, so knowing how much HI a satellite still holds, and whether that gas is being disturbed or stripped, gives an indications of how efficiently satellites are quenched after they fall into their host's halo. Mapping of HI with the VLA D-configuration is an efficient means of determining both how much gas each satellite has as well as the large-scale distribution of that gas, thus revealing candidates that are likely undergoing gas stripping and should be followed up with higher resolution observations.

Jones et al. 2024 →

Past research

Gas-rich ultra-diffuse galaxies

Measured the cosmic abundance of HI-bearing UDGs found in ALFALFA, led an HST snapshot program showing they host few globular clusters, and found evidence that some UDGs form through tidal heating of normal dwarfs.

Jones+18b, Jones+21, Jones+23a

Isolated galaxies vs. dense groups

Compared galaxies that have evolved on their own in isolation with those in the densest environments. With AMIGA I defined the HI content of several hundred of the most isolated galaxies known, a nurture-free baseline, and then used it to trace how Hickson Compact Groups lose their gas through repeated interactions, from a detailed study of HCG 16 to a global census of HI in compact groups.

Jones+18a, Jones+19, Jones+23b

ALFALFA and the HI mass function

Measured the HI mass function from the complete ALFALFA survey and its dependence on environment, and quantified source confusion in HI surveys and stacking.

Jones+15, Jones+16a, Jones+16b, Jones+18c, Jones+20