Selected Projects

DECaPS2: The Dark Energy Camera Plane Survey DR2

I led the data reduction for the second and final release of the Dark Energy Camera Plane Survey (DECaPS2). We provide a catalog of 3.32 billion sources in our survey footprint that covers 6.5% of the sky. We developed new methods well-suited to photometry in the Galactic plane, developed improved measures of photometric depth adapted to crowded-fields, and validate our pipeline extensively with synthetic star tests. For more, see our project website. Tip: See how far you can zoom in! (be patient)

APOGEE Spectral Decomposition

I developed and applied a new pipeline to the reduce SDSS-V APOGEE spectra that separates each visit into a linear combination of the most probable contributions from the sky, star, and ISM. This component separation approach to spectral reduction represents a paradigm shift from commiting to single point estimates (of the telluric absorption, sky emission lines, etc.) to marginalzing over our uncertainty about the contribution of each component to the overall spectrum in a joint fit that produces dense posteriors. The goals of this project are to measure the strong diffuse interestellar band near 15273 Å, discover new DIBs, reduce biases in stellar parameter and abundance measurements, and obtain robust parameter measurements of spectroscopic multiples. I have already demonstrated the power of this approach by improving the precision of the APOGEE stellar radial velocities to 30 m/s (see publications).

Diffuse Interstellar Band in Gaia DR3 RVS Spectra

I developed and applied a new pipeline to the Gaia (DR3) RVS spectra to measure the diffuse interstellar band (DIB) near 862.1 nm. This method marginalized over stellar types, avoiding false detections of stellar residuals as DIB detections, as in the public Gaia catalog. This analysis allowed us to show that there were no statistically significant detections of DIBs in the Local Bubble. For more, see our project website.

Kinetic Tomography: Reconstructing Dust in 3D + 1V

I am leveraging the precision DIB catalogs I produced (primarily from APOGEE) to infer not only where the dust is in 3D, but also how it is moving (its radial velocity). Preliminary results from this collaboration with Philipp Frank show dust densities inferred from the DIB equivalent width out to 6 kpc from the sun. Further, those dust clouds are colored by radial velocities simultaneously inferred from the DIB velocities. Future work will further improve in spatial resolution with the deluge of data from the ongoing SDSS-V survey and by also leveraging what we know about the 3D distribution of dust from photometric extinction dust maps.

High Angular Resolution 3D Dust Mapping

I extended a previous dust mapping methodology using bayestar and combined that with deep, high-quality NIR photometry from DECaPS2 (see above) to achieve an extraordinarily high angular resolution 3D dust map, 5x the resolution of the 2D Planck dust map! Through this careful statistical inference, we achieve 1 pc plane-of-sky resolution for structures 3 kpc away, for example in nearby spiral arms of the Galaxy. The video above shows the dust map integrated to 7 kpc, with zooms and pans over the entire southern Galactic plane to emphasize both the resolution and scale of this dust map! The GIF below shows the improvement in angular resolution acheived by our 3D dust map compared to a previous version (Bayestar19) and high resolution far infrared 2D emission maps from Herschel/SPIRE. The video below shows the 3D nature of the map by integrating to different distances. Both show only a 4° x 4° region in the footprint of the map. The map is publicly available in several forms, including by querying the Python package dustmaps.