PI: Eddie Schlafly (email, website)
Group Website: ISM*@ST
Project Duration: 12 month rotation project for a graduate student; can evolve into thesis project
Project Abstract
Dust scatters and absorbs light from astronomical sources, changing the observed colors and brightness of stars and galaxies. This has broad impact across astronomy. The Dark Energy Spectroscopic Instrument (DESI) is presently executing a survey of much of the extragalactic sky, and will observe >30M galaxies and >10M stars by 2026; millions have already been observed. This project will develop tools to combine DESI spectroscopy with photometry from surveys like Pan-STARRS and Gaia to measure the dust column to stars observed by galaxies. Initial forecasts suggest that maps of dust derived from this data can dramatically improve on existing dust maps based on dust thermal emission, allowing next-generation surveys to reach their cosmological objectives.
Background and Context
Many large cosmological surveys aim to map the distribution of galaxies over as large a fraction of the observable universe as possible. The clustering of galaxies at different scales as a function of cosmic time provides a powerful probe of the expansion history of the universe through the baryon acoustic oscillation scale. However, because dust reddens light, it also affects the observed clustering of galaxies and reduces the precision with which the baryon acoustic oscillation scale can be measured. Dust-introduced uncertainties are already one of the chief concerns for surveys like the Dark Energy Survey and Dark Energy Spectroscopic Instrument, and will be even more important for next generation surveys like the Legacy Survey of Space and Time. The combination of massive spectroscopic surveys of stars (like that from DESI) and precisely calibrated photometric datasets (like those from Gaia, PS1, DES, and LSST) offers the most promising approach for improving existing dust maps.
DESI is a large, welcoming collaboration where most of the astronomers are focused on extragalactic cosmology and relatively little effort is being invested into the stellar or dust programs. This means that the huge sample of stellar spectroscopy from DESI is largely unexplored, and that new contributions in this area could make a big impact in the collaboration.
Similar approaches have been applied to data from the Sloan Digital Sky Survey (Schlafly et al. 2011, 2016), demonstrating the basic soundness of the approach, though targets from those surveys are too sparse to make a map of the quality needed to mitigate cosmological systematics in next generation surveys. This technique has not been applied to data from DESI, where the number of stars observed may be adequate. Moreover, past work has relied on pre-baked catalogs of stellar parameters from surveys. Improved analysis of the underlying spectra may allow for better prediction of stars' intrinsic colors.
Student work
Planned work as part of project:
- Implement a simple model for the intrinsic colors of stars as a function of their stellar parameters along the lines of Schlafly et al. (2011) using DESI spectroscopy.
- Filter the resulting reddening measurements into maps.
- Determine impact of changing reddening maps on clustering measurements for DESI.
Side projects or future projects:
- Improve prediction of stellar colors based on DESI spectra via direct analysis of spectra. e.g. using a neural net connecting 1D spectra with observed colors.
- Map dust in the Galactic plane using the SDSS-V Galactic Genesis survey (e.g., Schlafly et al. 2016).
- Identify stars with peculiar extinction curves for follow-up spectroscopy
Skill sets (no prior experience necessary):
- Understanding of photometry & spectroscopy
- Understanding of stellar parameters
- Proficiency in analyzing large data sets
- Academic and technical writing
