An Observational Error Model, and Application to Asteroid Mass Determination
Conventionally, asteroid orbit determination algorithms assume that optical observations are unbiased and uncorrelated, with uniform uncertainties for all observers of a given era. This "naive error model" (NEM) is an oversimplification. We present a draft observational error model (DEM) that seeks to more accurately represent the correlations, biases, and uncertainties in asteroid observations. An extensive validation effort is described, in which the orbit of each numbered (and provisionally designated) asteroid is calculated, minus the observations from a single opposition; predictions for these missing observations are then used to compare the performance of each error model. Initial results demonstrate that the DEM exhibits a consistent and significant reduction in optical RMS error, indicating that we are more accurately representing the asteroid's true orbit. As a benefit of this improved orbital accuracy, we present new masses for 8 Flora (4.26 ± 0.45 x 10-12 SM), 13 Egeria (8.2 ± 1.6 x 10-12 SM), 14 Irene (4.13 ± 0.73 x 10-12 SM), 18 Melpomene (1.51 ± 0.51 x 10-12 SM), 21 Lutetia (1.29 ± 0.12 x 10-12 SM), 31 Euphrosyne (3.13 ± 0.59 x 10-12 SM), 47 Aglaja (1.09 ± 0.43 x 10-12 SM), and 49 Pales (1.35 ± 0.25 x 10-12 SM).
AAS/Division for Planetary Sciences Meeting Abstracts #40
- Pub Date:
- September 2008