Near-Earth Asteroid Orbit Determination
- Problem
- How can a sparse set of asteroid observations support a defensible orbit?
- Role
- Member of a three-student observing and analysis team; report coauthored with P. Li and A. Zhuang.
- Method
- Calibrated CCD images, reduced reference-star astrometry, and applied an iterative Gauss method.
- Result
- Produced a heliocentric orbit estimate and compared it with JPL Horizons.
- Status
- Completed 2018 student study; orbit and uncertainty estimates are conditional on the observing arc and model assumptions.
Project Narrative
During the Summer Science Program in Astrophysics, I worked with P. Li and A. Zhuang on a five-week observing and orbit-determination study of the near-Earth asteroid 1991 PM5. The project began with a practical question: how do three students turn faint CCD images and imperfect observing conditions into a defensible heliocentric orbit?
Our team carried the work from telescope setup through image calibration, reference-star plate reduction, asteroid centroiding, and an iterative Gauss-method solution. The result was a coauthored technical report that records the full observation-to-orbit chain, along with comparisons to JPL Horizons and an uncertainty analysis. It was my first sustained encounter with scientific data that could not be separated from its measurement conditions, modeling assumptions, and numerical sensitivity.
Adapting the Observation Plan
Better inference started with changing the data acquisition, not tuning the final model.
We observed with Etscorn Observatory's 355.5 mm C-14 Schmidt-Cassegrain telescope, a 3,911 mm focal length, and a 1024 × 1024 SBIG CCD. Early 90-second visual-filter exposures left the reduced asteroid signal too faint, so the team moved to 120-second clear-filter exposures. That change is a useful reminder that a numerical pipeline cannot recover information the detector never captured.
Observing was a team effort; the report is credited to J. Casas, P. Li, and A. Zhuang.
We observed with Etscorn Observatory's 355.5 mm C-14 Schmidt-Cassegrain telescope, a 3,911 mm focal length, and a 1024 × 1024 SBIG CCD. Early 90-second visual-filter exposures left the reduced asteroid signal too faint, so the team moved to 120-second clear-filter exposures. That change is a useful reminder that a numerical pipeline cannot recover information the detector never captured.
From Pixels to Sky Coordinates
Centroids only become astrometry after calibration, plate reduction, and careful time handling.
The reduction chain used bias, dark, and flat-field calibration before matching reference stars and locating the asteroid centroid. Plate constants mapped detector coordinates into right ascension and declination, while observation times and Earth-Sun vectors from JPL Horizons supplied the geometry needed by the orbital solver. Each stage introduced its own error source, so image noise, catalogue fit, timing, and model residuals had to remain conceptually distinct.
Solving for the Orbit
A Gauss solution is conditional on the selected observations and the conventions used to refine it.
Using three observations at a time, the iterative Method of Gauss produced a heliocentric state estimate that could be converted into classical orbital elements. Light-time correction and subsequent refinement connected the idealized geometry to the measured observing sequence. Comparing different observation triplets and the resulting residuals was as important as reporting one final element set, because a sparse arc does not determine an orbit without sensitivity to measurement and initialization choices.
Using three observations at a time, the iterative Method of Gauss produced a heliocentric state estimate that could be converted into classical orbital elements. Light-time correction and subsequent refinement connected the idealized geometry to the measured observing sequence. Comparing different observation triplets and the resulting residuals was as important as reporting one final element set, because a sparse arc does not determine an orbit without sensitivity to measurement and initialization choices.
Uncertainty and Evidence Boundary
A large Monte Carlo sample describes the assumed error model; it does not erase systematic uncertainty.
The project used Monte Carlo perturbations to examine how measurement uncertainty propagated into the recovered orbital elements, then compared the solution with JPL Horizons. The linked coauthored report documents the methods and team authorship. The orbit is an estimate from a sparse observing arc and stated uncertainty model, not an independently confirmed determination.
Related Writings
Orbital Determination Research Paper