Simulating stellar variability for the Roman sky
| Program ID | 19066 |
|---|---|
| Science Category | Stellar Populations |
| Program Type | Analysis |
| Category | Small |
| Principal Investigator | Adrien Thob |
| PI Institution | University of Pennsylvania |
| Co-Investigators |
|
| Abstract | Stellar variability connects stellar physics to cosmology, Galactic archaeology, and multi-messenger astronomy. Pulsating stars like Cepheids anchor the distance ladder, eclipsing binaries provide model-independent masses and radii, and compact object binaries test late-stage stellar evolution while previewing LISA gravitational wave sources. Roman will transform variability science through its two time-domain surveys of the Galaxy: the Galactic Bulge Time Domain Survey (GBTDS) and Roman Galactic Plane Survey (RGPS). Their high-cadence multi-filter strategies are optimized to discover short-period compact binaries, enable asteroseismology of hundreds of thousands of red giants, and detect thousands of longer-period variables. However, no framework exists to simulate realistic variable star populations tailored to Roman's specific cadence patterns, filter sets, and crowded-field challenges, leaving the community without validated pipelines when data arrive. We propose to build a stellar variability module for the open-source py-ananke framework, which generates synthetic resolved-star surveys from cosmological simulations. We will integrate binary population synthesis via COSMIC, pulsation models from state-of-the-art isochrones (PARSEC/MIST), and light curve generators to produce realistic time-domain synthetic catalogs. We will release validated benchmark datasets for GBTDS and RGPS fields, replicating their cadence patterns and filter sets, and demonstrate the module through three showcase projects: (1) forecasting variable star yields across Roman surveys; (2) predicting compact object binary populations for LISA and Gaia; and (3) developing a prototype detection tool for rare hybrid pulsator-binary systems. All software and data will be open-source, permanently archived, and deployed on the Research Nexus, providing essential infrastructure to turn theoretical stellar models into Roman-ready predictions and Roman data into robust astrophysical constraints. |