Conor Benson, University of Colorado Boulder; Charles Naudet, Jet Propulsion Laboratory; Daniel Scheeres, University of Colorado Boulder; William Ryan, New Mexico Tech/MRO; Eileen Ryan, New Mexico Tech/MRO; Joseph Jao, Jet Propulsion Laboratory; Lawrence Snedeker, Goldstone Deep Space Communications Complex; Jeffrey Lagrange, Goldstone Deep Space Communications Complex
Keywords: GEO, debris, spin states, radar, light curves, YORP effect
Abstract:
The GEO debris population keeps growing with continued launches and no natural deorbit mechanisms. Understanding the long-term dynamical evolution of these debris is necessary to protect active assets and preserve GEO for future use. The long-term orbital evolution of GEO debris has been studied extensively, while research on rotational dynamics has been much more limited. Nevertheless, the spin states of retired and otherwise defunct GEO satellites are diverse and evolve significantly over time. A better understanding of long-term defunct satellite spin state evolution would aid GEO space situational awareness, material shedding predictions, modeling of attitude-dependent solar radiation pressure, active debris removal (ADR), and satellite servicing. ADR and servicing promise to figure prominently in future efforts to manage the GEO debris population and lower satellite costs. As a result, a number of organizations are developing ADR/servicing missions. These missions will require accurate spin state estimates to grapple and de-spin large target satellites. With evolving spin states and many potential targets, early spin state predictions will be extremely valuable for mission planning and execution. Spin state estimates are also valuable for satellite anomaly resolution and recovery.
Recent observations and dynamical modeling have shown that spin states of some defunct GEO satellites are driven by the Yarkovsky-OKeefe-Radzievskii-Paddack (YORP) effect, spin state evolution due to solar radiation and thermal re-emission torques. With much still unknown, accurate observations are needed to gain insight and validate general long-term dynamical models. A satellites rotational angular momentum direction (pole) seems to greatly impact its YORP-driven evolution. So both spin period and pole estimates are needed to advance our understanding. Even with complete knowledge of the mechanisms driving long-term evolution, accurate predictions require accurate initial estimates. So prediction hinges on obtaining unambiguous spin state estimates from observations.
Unfortunately, extracting GEO satellite spin periods and poles from ubiquitous non-resolved photometry generally require accurate satellite models, which are usually unavailable. Light curves are very sensitive to optical properties and detailed satellite geometry. So even using high fidelity (e.g. ray-traced) photometric models for satellites with well-constrained geometry and optical properties, inversion can result in numerous local minima. Some defunct GEO satellites are also in non-principal axis rotation (i.e. tumbling), and recent work indicates that the YORP effect can cause defunct satellites to transition from uniform rotation to tumbling where their dynamical evolution continues. Tumbling light curve analysis is more challenging than for uniform rotation because there are two fundamental spin periods: one corresponding to precession of the satellites long axis around the pole and the second to rotation of the long axis about itself. Dominant tumbling light curve frequencies consist of several or more low-order harmonics of the two fundamental frequencies which are a priori unknown. The common analysis approach requires testing numerous candidate tumbling period pairs over the sphere of possible attitude phasing and pole directions using simulated observations. This often results in many similarly well-fitting period pairs and poles within model uncertainty.
The aim of this work is to greatly reduce spin state ambiguity by incorporating ground-based Doppler radar measurements. The spatial resolution afforded by Doppler radar allows for clear identification of satellite features. The resolved Doppler echoes also provide unambiguous long axis precession period estimates for both uniform rotators and tumblers. Leveraging the time-varying antenna to satellite position vector, we can also obtain unambiguous pole solutions or greatly narrow the possibilities. This is achieved with just a pair of co-located transmitting/receiving antennas. Most importantly, this pole estimation approach does not require a detailed satellite model, just knowledge of the satellites radial extent and/or prominent satellite features that can be identified over successive rotations (e.g. solar panel outer edge or upper stage rocket nozzle). And unlike optical telescopes, radar can be easily employed day or night.
For this work, we observed both uniformly rotating and tumbling defunct GEO satellites and rocket bodies with Deep Space Network (DSN) antennas at NASAs Goldstone Deep Space Communications Complex in California. Two 34 meter DSN antennas were used in a bi-static configuration with one antenna transmitting and the second receiving. The transmitted signal consisted of continuous wave (cw) carrier at X-band. Reflection of the signal off of the rotating target and back to the receiving antenna yielded time-varying Doppler spectra. Each target was observed periodically over the course of several hours to sample different viewing geometries. The targets were observed in this manner on a number of days in early 2021. For several targets, near-simultaneous light curve data was also collected. Building on earlier studies, we revisit the previously observed GOES 8-12 satellites to compare spin rates/pole directions and gain insight about their ongoing evolution. This will help us better understand how YORP drives defunct satellites spin states. We also investigate high altitude rocket bodies.
Date of Conference: September 14-17, 2021
Track: Non-Resolved Object Characterization