Speaker
Description
A ubiquitous problem in astronomy is correctly assigning absorption or emission features in a spectrum to the multiple processes occurring along the line of sight within the spectroscopic field of view. We introduce a new data processing pipeline to decompose APOGEE spectra into components associated with the target star, terrestrial atmosphere, and dust along the line of sight. In this model, the sum of the components is exactly the data, meaning unexpected signals are exactly retained. This decomposition is obtained by modeling each component as a draw from a high-dimensional Gaussian distribution in the data-space (the observed spectrum)---a method we call ``Marginalized Analytic Data-space Gaussian Inference for Component Separation'' (MADGICS). This technique provides statistically rigorous uncertainties and detection thresholds, which allows better leveraging of low signal-to-noise spectra. We will focus on applications to mapping Galactic dust via the 15273 Å diffuse interstellar band. Possible impacts on other science goals include radial velocity determination, spectroscopic binary (SB2) modeling, and stellar parameter inference.
