Research

As a postdoctoral fellow at the Max-Planck-Institut für Extraterrestrische Physik (MPE), I'm excited about using new observations with the NOEMA interferometer to probe dust-obscured bulge growth in Milky Way and Andromeda progenitors. Merging these observations with kinematic measurements from the KMOS3D survey, I hope to investigate the physical processes driving the assembly of galaxies.

As part of my thesis at Yale, I pioneered a new method to map the emergence of galactic structure using the Wide Field Camera 3 grism on the Hubble Space Telescope. By combining information on the spatial distribution of ionized gas from the 3D-HST survey and dynamics from Keck spectroscopy, I shed new light on where star formation occurs in galaxies. Some of my major results are described below. 


Forming Massive Galaxy Cores 

A cauldron of star birth in the center of a young galaxy. This illustration reveals the celestial fireworks deep inside the crowded core of a developing galaxy, as seen from a hypothetical planetary system. The sky is ablaze with the glow from nebulae, fledgling star clusters, and stars exploding as supernovae. The rapidly forming core may eventually become the heart of a mammoth galaxy similar to one of the giant elliptical galaxies seen today. Credit: NASA, Z. Levay, and G. Bacon (Space Telescope Science Institute).

A cauldron of star birth in the center of a young galaxy. This illustration reveals the celestial fireworks deep inside the crowded core of a developing galaxy, as seen from a hypothetical planetary system. The sky is ablaze with the glow from nebulae, fledgling star clusters, and stars exploding as supernovae. The rapidly forming core may eventually become the heart of a mammoth galaxy similar to one of the giant elliptical galaxies seen today. Credit: NASA, Z. Levay, and G. Bacon (Space Telescope Science Institute).

Since galaxies grow inside-out (see below), most massive galaxies are thought to have formed their dense stellar cores at early cosmic epochs. However, this formation phase has been notoriously difficult to observe. Undertaking a large program with MOSFIRE and NIRSPEC on the Keck telescopes we uncovered one of these massive galaxy cores in it's turbulent formation phase in the early universe. This result was published in Nature (Nelson et al. 2014, arXiv) with accompanying press releases (NASAYale, Keck, ESA) and articles (The Washington Post, Christian Science Monitor, and TIME – plus a pretty hilarious one in The Week). We also showed that the population of progenitors likely followed a simple inside-out growth trajectory (van Dokkum, Nelson et al 2015). In the future, I hope to use the ability to map Hα, Hβ, and IR emission at high spatial resolution with JWST to fully uncover the core formation phase of massive galaxies. 


sizeGrowth.png

Star formation builds galaxies inside-out

{Nelson et al. 2012} A galaxy’s Hα emission gives us information on where it is growing via star formation at the epoch of observation; its near-infrared (NIR) continuum emission tells us where it grew in the past. By mapping the distribution of Hα and NIR continuum emission in galaxies at 0.7 < z < 1.5, I showed that the Hα has a larger half-light radius. Taken at face value, this means that during this epoch when a third of the total star formation in the history of the Universe took place, most galaxies are increasing their radii due to star formation, assembling from the inside-out.

Looking ahead, to fully understand this size evolution in its physical context, requires determining whether the integrated size growth we observed in van Dokkum et al. 2013 and van Dokkum, Nelson et al. 2015 is explained by this star formation or if the Universe requires other processes to dissipate angular momentum.


Star formation is spatially coherent

{Nelson et al. 2016b} The star forming main sequence describes the derivative (star formation rate) - integral (stellar mass) pair serving as an organizing principle for galaxy growth. Key to understanding the physical drivers of the star forming main sequence is where star formation is located in normal star forming galaxies, as well as where it is enhanced in galaxies above the main sequence and where it is suppressed in galaxies below the main sequence. With the first systematic census of star formation across the SFR-M∗ plane at z ∼ 1, I found that star formation is spatially coherent: star formation is enhanced at all radii in galaxies above the main sequence and suppressed at all radii in galaxies below it. This provides strong constraints on the physical processes driving the enhancement and suppression of star formation in models of galaxy formation. 

coherentSF.png

Most galaxies form most of their stars in disks

{Nelson et al. 20013} I marshaled a number of lines of evidence to buttress this claim. Stacking Hα maps of galaxies homogenized by size, I showed that the ionized gas displays nearly exponential surface brightness profiles. By comparing with expectations from model axis ratio distributions, I determined that the geometries of the ionized gas distributions were consistent with disks under different viewing angles. Conducting a multi-night follow-up campaign with the NIRSPEC spectrograph on the Keck telescope, I showed that a representative sample of normal star-forming galaxies all exhibited rotation-dominated kinematics. These lines of evidence converged to show that the Hα emission we observe originates predominantly from the exponential disks of galaxies. This result thus implies that most star formation occurs in galactic disks, at least in the regimes in which Hα is relatively unobscured and can be taken as a proxy for star formation (but see below).


Dust and In-Situ Bulge Building

{Nelson et al. 2016a} Probably the single greatest uncertainty in the interpretation of the maps and dynamics of Hα emission as star formation is radial gradients in dust extinction within galaxies. Using the WFC3 grism to map the distribution of Hα and Hβ emission, I made the first spatially resolved measurements of the Balmer decrement at z > 1. Systematic measurements of the Balmer decrement provide stringent constraints on the radial gradients in dust attenuating emission form star formation in galaxies. This allows us to make fully dust-corrected measurements of the radial distribution of star formation. With this information I showed that dust is unimportant at all radii in galaxies with M < 10 10 M . Galaxies with M* > 10 10 M , on the other hand, have significant attenuation toward their centers. This has significant implications for how galaxies grow: low mass galaxies form most of their stars in exponential disks while higher mass galaxies have significant fraction of their star formation in their centers. This means that bulge growth can occur at least in part due to in situ star formation.

radialDustAttenuation.png