Roman Galactic Variability Project

PI: De, Kishalay, Columbia University
Wide-Field Science – Regular

The Roman space telescope will conduct a 700-hour survey of about 1000 sq. deg. of the Galactic plane, with at least 100-250 hours dedicated to time domain surveys with cadences ranging from hours to years. Even as its exact scope is being defined by a community-led process, it is clear that there broad interest for time domain surveys that span cadences of minutes to years, encompassing a variety of science cases ranging from star formation to the evolution of black holes. Of particular interest is the concurrent operation of this survey with some of the largest ground-based surveys in the optical bands, opening up entirely new opportunities for joint scientific and technical utilization. However, despite its prospects to push completely new frontiers in Galactic transient science, the lack of prior systematic efforts at similar scale requires both scientific and technical preparation. We propose the development of the Roman Galactic Variability Project – a set of curated tools to support preparation and complete utilization of this data stream for transient science. We highlight and address two of the biggest opportunities in Galactic transient science that is likely to stem from this survey.

The first involves the routine identification of the faint quiescent counterparts of Galactic transients discovered in both the Roman and ground-based transient surveys like the Rubin observatory – with Roman providing the exquisite infrared depth and spatial resolution in crowded fields, and Rubin providing denser coverage and longer baseline sampling. We propose to develop tools that will use the high spatial resolution Roman maps to identify progenitors of Galactic transients by jointly modeling images from ground-based (i.e. coarser resolution) surveys like Rubin, using the standard “alert” packets that are scheduled to be delivered from the surveys. Unlike the joint modeling of static imaging from synoptic surveys, cross-identification of transient sources will require the use of time-resolved images contained in the alert packets together with Roman imaging, as well as the likely use of multi-band information to securely identify counterparts based on their colors in extremely crowded fields.

The second opportunity capitalizes on Roman’s unique potential to serve as a transient discovery engine in itself via the multiple epochs that will (likely) be obtained during the survey. While the infrared sensitivity promises new discoveries in the lowest luminosity Galactic transients together with their resolved structures, optimal image subtraction in crowded fields remains a formidable challenge that requires revision of the standard assumptions made in transient discovery algorithms adopted for sparse fields. We propose to design and validate optimal image subtraction and transient detection algorithms in the expected confusion-limited images, that is expected to be broadly applicable in Roman Galactic Plane observations as well as other crowded fields like globular clusters.

As the Galactic Plane Survey is scheduled to occur in the first two years of the mission, these tools have to be developed now to enable Galactic transient discovery and follow-up during the course of the survey. We outline plans to i) test the developed codes with archival datasets prior to the launch of Roman and ii) release the resulting tools to both the Roman Project Infrastructure Teams for implementation in existing efforts, as well as to the broader community.