Probing the Outer Solar System: Searching Below the Noise

This research team from the University of Washington will develop an image processing tool for to detect Trans Neptunian Objects (TNOs) and Main Belt Asteroids that can be used with Rubin Observatory’s Legacy Survey of Space and Time (LSST) data. New generations of astronomical surveys such as LSST will soon open a new era in Solar System science by mapping the skies in unprecedented depth and detail.

LSST alone will increase the number of known TNOs from the 2729 known today to over 40,000 and the number of Main Belt Asteroids with measured orbits will also increase by an order of magnitude. Current analysis techniques make use of only a fraction of the information present within these data (requiring that asteroids and comets be detected in individual images).

By combining or coadding images from these surveys to increase the signal-to-noise ratio, the number of small bodies detected by the LSST could increase by over an order of magnitude. This project will develop codes building on an existing prototype to process these images. The tool will be made available to the community upon completion.

The broader impacts of this project are to develop tutorials and educational material for astronomical data science targeted towards undergraduate students. As part of this curriculum, a series of educational modules (tutorials, Jupyter notebooks, videos, and astronomical applications) will be developed, prototyped, and evaluated using the Pre-MAP program at the University of Washington (a program that targets traditionally underrepresented groups in STEM by introducing them to research in their first year).

This project is to develop open-source software to aid in the detection of moving objects in large data sets from upcoming all-sky surveys, such as LSST. A prototype of KBMOD (Kernel-Based Moving Object Detection) can search over 10 billion moving object trajectories in a stack of 10-15 4Kx4K images in under a minute using consumer-grade GPUs. This project will build upon the success of KBMOD to develop a framework capable of finding TNOs in the multi-year surveys from the LSST; develop a tool set for the robust classification of asteroids and comets from their light curves based on advanced statistical techniques including neural networks; implement optimized search algorithms and strategies for searching for moving sources in surveys with long temporal baselines and for faster moving sources (e.g.

Main Belt Asteroids); and extend the probabilistic framework to the question of image classification - initially focusing on the identification of binarity in barely resolved images of asteroids.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.