Statistical Estimation of Refractivity from Radar Sea Clutter

This paper summarizes current developments in the refractivity from clutter (RFC) techniques and describes the global parametrization approach in estimation of the lower atmospheric electromagnetic sea ducts. RFC uses radar clutter to gather information about the environment the radar is operating in. Range and height dependent atmospheric index of refraction (M-proflle) is statistically estimated from the sea-surface reflected radar clutter. These environmental statistics can then be used to predict the radar performance by taking multidimensional integrals of the posterior probability density. All of the following methods use a Bayesian framework and use split-step fast Fourier transform based parabolic equation approximation to the wave equation as the propagation model. Environmental parameters are inverted using genetic algorithms, Markov chain Monte Carlo samplers, and a hybrid genetic algorithm -Markov chain Monte Carlo technique. The methods are compared with respect to their estimated maximum a posteriori accuracy, speed and ability to sample correctly from posterior density. The inversion algorithms are implemented on S-band radar sea-clutter data from 1998 Wallops Island, Virginia experiment. Reference data are measured as range-dependent refractivity profiles obtained with a helicopter. The inversions are assessed by comparing the propagation predicted from the radar-inferred refractivity profiles and from the helicopter profiles.