Bayesian Estimation of Altimeter Echo Parameters
Jerome Severini, Corinne Mailhes, Pierre Thibaut, Jean‐Yves Tourneret · 2008
This paper studies a Bayesian algorithm for estimating the parameters associated to Brown's model. The joint posterior distribution of the unknown parameter vector (amplitude, epoch and significant wave height) associated with this model is derived. This posterior is too complex to obtain closed form expressions of the minimum mean square error and the maximumaposterioriestimators. We propose to sample according to this distribution using an hybrid Metropolis within Gibbs algorithm. The simulated samples are then used to estimate the unknown parameters of Brown's model. The proposed strategy provides better estimations than the standard maximum likelihood estimator at the price of an increased computational cost.