Katiyayni Balachandran, University of Texas at Arlington; Kamesh Subbarao, University of Texas at Arlington
Keywords: Light curves, Rotation Period, Resident Space Objects, Power spectrum, Simulation
Abstract:
A robust method for the estimation of Sidereal rotation period of Resident Space Objects (RSOs) in orbit using Periodogram analysis is presented. This can be applied to non-uniformly sampled light curves (a measurement of the objects brightness over time), which is often the norm with astronomical data collection. Most ground-based observations are difficult to achieve uniform spacing due to delegated windows, position of the target and time-delay of information. Other reasons such as weather and diurnal, lunar or seasonal cycles play into effect where sampling is concerned.
Spectral analysis of photometric observations using the Fast Fourier Transform (FFT) is complex and inadequate with an unevenly spaced dataset. Identifying the true periodicity in a signal then becomes a challenge. Increasing the sampling frequency does mitigate this problem, but by the slightest extent. The large gaps in data impose a problem and require re-sampling and interpolation for uniform spacing. Subsequently, the Power Spectral Density (PSD) estimation is inaccurate, and the additive interpolation noise compromises the computation of the rotation period. Each sampled observation has an associated error variance which multiplies when framing a time series with missing data points.
The Lomb-Scargle Periodogram addresses these issues and is optimal for non-uniformly spaced data with the added benefit of faster computational times. This paper will illustrate the technique in the context of rigid body estimation of asteroids, rocket bodies, and fragments of debris in an ellipsoidal shape. A simulation generating uniformly-sampled synthetic light curves of one of the test cases, at varying sampling frequencies and angular step sizes, validates the method within a bounded 10% error.
Date of Conference: September 11-14, 2018
Track: Poster