Bayesian Linking of Geosynchronous Orbital Debris Tracks as seen by the Large Synoptic Survey Telescope

TL;DR

This paper introduces a Bayesian sampling approach using MCMC to link and constrain geosynchronous debris orbits from optical telescope streak observations, addressing challenges of short exposures and crowded fields.

Contribution

It presents a novel Bayesian model and algorithm for linking orbital tracks from optical observations, specifically tailored for short-exposure, crowded GEO debris fields with LSST data.

Findings

The method can forecast orbit constraints achievable with LSST.

Short exposure times lead to large, degenerate orbit errors.

Linking in crowded fields remains challenging with current LSST cadence.

Abstract

We describe a Bayesian sampling model for linking and constraining orbit models from angular observations of "streaks" in optical telescope images. Our algorithm is particularly suited to situations where the observation times are small fractions of the orbital periods of the observed objects or when there is significant confusion of objects in the observation field. We use Markov Chain Monte Carlo to sample from the joint posterior distribution of the parameters of multiple orbit models (up to the number of observed tracks) and parameters describing which tracks are linked with which orbit models. Using this algorithm, we forecast the constraints on geosynchronous (GEO) debris orbits achievable with the planned Large Synoptic Survey Telescope (LSST). Because of the short 15 second exposure times, preliminary orbit determinations of GEO objects from LSST will have large and degenerate errors on the orbital elements. Combined with the expected crowded fields of GEO debris it will be challenging to reliably link orbital tracks in LSST observations given the currently planned observing cadence.