AAS Meeting #193 - Austin, Texas, January 1999
Session 77. Astrophysical Processes and Computational Techniques
Display, Friday, January 8, 1999, 9:20am-6:30pm, Exhibit Hall 1

[Previous] | [Session 77] | [Next]


[77.06] Bayesian Approach to Lunar Data Analysis

J.G. Ries (UT McDonald Observatory), W.H. Jefferys (UT Department of Astronomy), P.J. Shelus (UT McDonald Observatory)

In Lunar Laser Ranging, most of the photons detected are due to noise (except when the reflector is in the dark and most photons are laser returns). Unfavorable observing conditions can further deteriorate the return signal, and some data cannot be extracted by the present analysis method. For sufficiently strong data, the width of the returning laser pulse can be recovered by the Bayesian approach. Taking this information into account could improve the error estimate of the normal point. Earth orientation determination could be done simultaneously with data filtering, providing almost instantaneous Earth orientation parameters values.

In our investigation, the likelihood function of a generic lunar experiment has been derived. Previously, only two unknown parameters of the Earth orientation were considered. This time the strength of the returning pulse and the background noise level have been added to the parameters to be determined. Instead of the marginalizing the distribution by numerically integrating for the possible range of parameters, the marginal distributions are obtained using a Markov chain Monte Carlo method. The method has been tested on poor quality simulated lunar data and demonstrated good recovery of the unknown parameters. Testing of the method on actual lunar data is underway.


[Previous] | [Session 77] | [Next]