Marcel Neeleman

I am an astronomer working at the National Radio Astronomy Observatory (NRAO) in Charlottesville, VA. At NRAO, I am part of the North American ALMA Science Center, which is charged with the North American operations of the Atacama Large Millimeter Array (ALMA). IN particular, I work as part of the pipeline working group, which is charged with providing a (nearly) automated pipeline that takes the raw data from the telescope and converts it into science-ready data products.

My scientific interests are using the world's largest telescopes to study the formation of galaxies in the very early universe. To be more specific, I study quasars and quasar absorption line systems with observations that span a large range of the electromagnetic spectrum from far-ultraviolet observations using the Hubble Space Telescope all the way through radio observations with the Jansky Very Large Array. To read more about my research, please see the 'My Research' section below.

My Research

I study the formation of galaxies in the early universe. In particular, I use quasars both as a way to locate massive galaxies at very high redshift, and as a background source to study the diffuse gas surrounding foreground galaxies through absorption line spectroscopy. While the former allows me to study the formation of some of the most massive galaxies that exists in our universe, the latter provides a unique view of the average galaxy population at high redshift. My research is roughly divided into three main topics: studying atomic gas at high redshift, finding the galaxies that are associated with this gas and studying the host galaxies of high redshift quasars. In the below three sections, I provide a quick overview and highlight some of the work I have contributed to these topics.

Damped Lyman-α Absorbers

Using optical and near-infrared spectrographs, we can study the spectra of high redshift quasars in great detail. A typical quasar spectrum is shown in the figure on the left. Leftwards of the Lyman-α emission of the quasar at 1215Å, you can see the Lyman-α forest. This is a region where part of the light of the quasar gets absorbed by foreground atomic hydrogen. Dense clumps of gas will create more absorption and are known as Lyman limit systems (LLS). The strongest absorbers show damping wings in their Lyman-α absorption and are therefore called Damped Lyman-α Absorbers (DLAs).

DLAs are particularly interesting, because they contain the bulk of the atomic hydrogen in our Universe. DLAs have therefore been used to constrain the evolution of this gas with cosmic time. In my 2016 paper, I used Hubble Space Telescope archival data to provide an estimate of the cosmic density of atomic hydrogen at low redshift, where only space-based telescopes can measure the Lyman-α transition. Combining this result with results at higher redshift reveals a steady decline in the amount of neutral atomic hydrogen as the Universe ages. We have put this result in context with the evolution of several other mass constituents of galaxies (i.e., stars, molecular gas, etc.) in Walter, Carilli, Neeleman et al. (2020).

Using the metal lines associated with the DLA, we can also provide an estimate of the metallicity, kinematics and the physical conditions of the gas. The latter two were the main topics of my PhD thesis and my 2013 and 2015 papers. In my 2013 paper, I showed that the metallicity of the DLA is correlated with both the redshift and width of its absorption lines, and in my 2015 paper, I used the fine-structure lines of singly ionized carbon and silicon to constrain the temperature and density of the gas.

Finding and Characterizing Absorption-Selected Galaxies

We expect that DLAs are closely associated with galaxies. This association is based both on the predictions of simulations as well as observational evidence. In the local universe, we see that atomic hydrogen column densities giving rise to DLAs are solely found in the disks of galaxies, and cosmological simulations predict that DLAs occur around the full range of the simulated galaxy population. DLAs therefore provide a unique opportunity to study normal, high-redshift galaxies without any a-priori knowledge of the emission properties of the galaxy. These absorption-selected galaxies provide a complimentary picture of the galaxy population compared to galaxies found in typical high redshift galaxy surveys.

Over the past three decades, a multitude of studies have attempted to find the optical galaxy counterparts to DLAs, but often with little success. This limited success has been attributed to the expected faintness of absorption-selected galaxies. Since DLAs are associated with the full galaxy population, many DLAs are associated with galaxies that are too faint even for the world's best telescopes. By pre-selecting DLAs by their metallicity, recent studies have increased the detection rate. This is because mass and metallicity are expected to correlate in galaxies and more massive galaxies are easier to detect. Together with better instrumentation, this implies we now have an excellent way forward to discover the connection between galaxies and absorbers.

However, optical telescopes are not the only instrumentation that we can use. Capitalizing on the remarkable sensitivity of the Atacama Large Millimeter Array (ALMA), I have set out to detect the galaxies associated with high-metallicity DLAs in the (sub-)millimeter regime. At these wavelengths, we can detect the dust continuum emission and emission of far-infrared emission lines (in particular [CII] 158 μm emission) from the gas within the galaxy. This research has been extremely successful resulting in the first detections of absorption-selected galaxies in the millimeter regime, as well as the highest redshift absorption-selected galaxies. My 2017 study, published in Science, of two absorption-selected galaxies was covered by a press release from the National Radio Astronomy Observatory (NRAO) and includes a cool video describing this research. Subsequent observations with ALMA have now revealed a sample of over 20 absorption-selected galaxies, more than doubling the previous number of known absorption-selected galaxies. The coming decade therefore promises to finally uncover the connection between DLAs and galaxies.

Besides detecting the galaxies, ALMA in its more extended configuration can also resolve the emission arising from the interstellar medium of absorption-selected galaxies. In my 2020 paper, published in Nature, I discuss ALMA observations of the ionized carbon emission ([CII] 158 μm) from one of these galaxies. In this paper, I show that the emission arises from a cold disk rotating at a similar speed as our own Milky Way. The existence of a dynamically cold disk galaxy at these redshifts is exciting, because it challenges our understanding of how gas can accrete onto a galaxy. The accreted gas must have been cool or cooled quickly in order to form a dynamically cold disk galaxy. The association of this disk galaxy to the DLA gives a unique insight into the kinematics of the atomic gas probed by the absorber, and it reveals that for this single system the atomic gas is co-rotating with the disk. We named this disk galaxy the 'Wolfe' disk, in honor of my thesis advisor, who argued for the existence of these galaxies, and this article was covered extensively by the popular press as well as a press release by the NRAO as well as another cool video. Future resolved observations of absorption-selected galaxies can provide an exciting opportunity to compare the flow of gas within and outside the galaxy in order to explore the role this gas plays in the formation of galaxies.

Quasar Host Galaxies

Quasars are some of the brightest non-transient objects in the Universe, and they have been detected well into the epoch of reionization, when the Universe was less than one billion years old. Optical and near-infrared spectroscopy reveal that the most luminous quasars are powered by billion-solar-mass black holes. Locally, black hole mass correlate with the mass of the galaxy and if this correlation holds at high redshift, then we expect luminous quasars to trace massive quasar host galaxies. Quasars therefore provide a unique opportunity to study massive galaxies in the early Universe.

Studying the stellar light from the galaxies associated with quasars remains difficult, because the light from the quasar outshines the light from the stars in the quasar host galaxies. Directly detecting this stellar light will have to await observations with the next generation near-infrared telescopes like the James Webb Space Telescope and 30-m class telescopes such as the European Extremely Large Telescope. Fortunately the quasar is much dimmer at millimeter wavelengths; allowing us to search for the continuum emission emitted by the dust as well as emission lines emitted by the metals within the interstellar medium of the galaxy. Using the excellent capabilities of ALMA, we can now detect emission lines from quasar host galaxies well into the epoch of reionization. Remarkably, these observations reveal that a fraction of these quasar host galaxies have massive, dusty companions. This is an indication that these quasar reside in large overdensities, as predicted by simulations. In Neeleman et al. (2019), I describe higher resolution observations of the fields surrounding five quasars with known companions. These observations reveal little distinction between the quasar host galaxy and their companion, which indicates that the quasar is not significantly affecting the rest-frame far-infrared properties of its host galaxy.

To further explore the host properties of high redshift quasar galaxies. We have built a sample of 27 high redshift quasar host galaxies that have been imaged at 0.25" resolution with ALMA. In the first paper in a series of three papers, we show that the optical position of the quasar is centrally located in the quasar host galaxy, and none of the quasar properties correlates with the rest-frame far-infrared properties. This further supports the assertion that the quasar does not affect the ALMA observations. In the second paper, we show by carefully stacking the observations that the far-infrared emission is extended beyond the emission we see in individual observations, but this emission is consistent with simulations that predict emission at these scales from an extended component. Finally in the last paper, we model the kinematics of the galaxy using the fitting code Qubefit. We find that about a third of the galaxies have kinematics consistent with a disk and a third show evidence of ongoing merger activity. The final third of galaxies show turbulent motions consistent with the expectations of a bulge. Using these kinematics we can estimate the dynamical mass of the galaxies. This reveals that these galaxies systematically fall above the local black hole - gas mass relation, as shown in the figure. This could mean that for these luminous quasars, the black hole masses grew faster than their host galaxies.

It is exciting to consider that the typical resolution in the above sample of 0.25" is not the best resolution that ALMA can achieve. We have already obtained spectacular imaging of a quasar host galaxy at 0.04" resolution (Venemans, Neeleman et al. 2019). More imaging at this resolution are on the horizon allowing us to explore the interstellar medium at exquisite resolution, and we might even be able to resolve the sphere of influence of the black hole, which would provide an independent estimate of the black hole mass.

Code

Over the years, I have written code in IDL and Python to help with the analysis of observational data. Here I post some of the more developed packages that might be of use to others.

Qubefit

Qubefit is a fully documented kinematic fitting code written in Python. It fits simple user-defined models to spectral data cubes in order to extract kinematic information from the cube using a Bayesian framework. The code was originally writted to deal with spectral data from interferometers (such as ALMA), but has been succesfully applied to data cubes from integral field units in the optical regime (i.e., MUSE and KCWI). The documentation for the code can be found here.

Contact Me

National Radio Astronomy Observatory
520 Edgemont Road
Charlottesville, VA 22903
United States

Office: 328
Phone: +1 (434)244-6819
email: mneeleman-at-nrao.edu