I identified a sample of ~ 2000 emission-line galaxies at 1.9 < z < 2.35 using the 3D-HST grism survey data. Objects are selected on the basis of strong rest-frame optical emission lines, with [O III] λλ4959, 5007 being the strongest feature in > 90% of the sources. I found that these objects have lower stellar mass and lower dust content than their continuum-selected counterparts. This selection function is important for placing the forthcoming galaxy samples from missions like Euclid, Roman, and JWST into the context of the global galaxy population.

Bowman et al. 2019, ApJ, 875, 152

I am a co-developer of the flexible spectral energy distribution (SED) fitting code MCSED. The code includes several easily-adjustable prescriptions for dust attenuation, star formation history, stellar metallicity, nebular emission, and thermal emission from dust. It can use photometry, emission-line fluxes, and absorption indices to constrain the fits.

I fit the SEDs of the z ~ 2 emission-line galaxy sample using the extensive photometric data from Skelton et al. (2014) which include 20-40 bands of photometry spanning from rest-frame 1200 Å to 8 μm at z ~ 2. The SEDs become redder with increasing stellar mass due to both an increase in dust content and a greater population of old, red stars.

Bowman et al. 2020, ApJ, 899, 7

Please find the current version of MCSED on GitHub.

Lyman-alpha photons are created as a result of active star formation in galaxies, yet the mechanisms by which those photons escape from their host galaxies are not well understood due to the resonant scattering nature of the   n = 2→1 transition in the hydrogen atom. Since Lyman-alpha emitters are used in the HETDEX survey to trace the underlying large scale structure of the Universe, it is important to understand how this selection samples from the underlying galaxy population.

We extracted VIRUS spectra for the ~ 900 optical emission line galaxies (Bowman et al. 2019) that have been observed in the HETDEX survey to measure the Lyman-alpha escape fraction of this sample. We used the physical properties measured for the sample in Bowman et al. (2020) to bin and stack the spectra to examine the escape fraction as a function of stellar mass, star formation rate, size, dust content, and [O III] / Hβ ratio.

I played an advising and support role on this project, which was led by an undergraduate researcher who I mentored for four years at Penn State. Laurel is now a graduate student in the department of astronomy at The University of Texas at Austin.

Weiss et al. 2021, ApJ, 912, 100

Upcoming missions like Euclid and Roman will use emission-line selected galaxies to explore the nature of dark energy through the measurement of baryon acoustic oscillations. The achievable precision of these experiments relies upon knowing the number of galaxies that will be observed. 

The [O III]-selected sample from 3D-HST grism frames presented in Bowman et al. (2019) is directly analogous to the objects that will be identified by the Euclid and Roman missions, which will use grism spectroscopy to find [O III] emitters out to z ~ 2.8. I jointly measured the luminosity function and flux completeness function from this sample of ~ 2000 [O III] emitters at z ~ 2, and used this measurement to predict the number of galaxies that will be found by Euclid and Roman. 

I also used the UV-based star formation rate (SFR) estimates to calibrate the relationship between SFR and dust-corrected [O III] luminosity at z ~ 2, which can be used to estimate the volumetric star formation rate density of [O III] emitters. This calibration is consistent with others that appear in the literature and implies that galaxies selected via strong [O III] emission contain roughly half of the total star formation activity at z ~ 2 (relative to measurements from UV+IR studies at the same epoch).

Bowman et al. 2021, ApJ, 920, 78