Refractivity-from-clutter using global environmental parameters

This paper examines the sensitivity of radar clutter returns to variations in parameters used to describe the refractive environment that is associated with surface-based ducts. This supports determining efficient parameters so as to minimize the search space required in the inverse problem of inferring the refractivity environment from observations of radar sea clutter. First, the sensitivity of replica fields to variations in range-independent parameters are considered. Next, variations in the parameters with range are modeled as a Markov processes. It is seen that either source of variation could explain variations in radar clutter observations obtained during a surface-based ducting event with the 3.0 GHz Space Range Radar (SPANDAR) at Wallops Island, VA. We then use the Simulated Annealing/Genetic Algorithm (SAGA) general purpose inversion code to infer refractivity parameters from observed clutter. SAGA is configured to use an embedded parabolic equation electromagnetic propagation model, a four-parameter model for atmospheric refractivity, and a linear least-squares objective function. The mismatch between (a) the optimal replica field and the observed clutter and (b) the inferred refractivity profile and the range-dependent refractivity structure obtained by in situ measurements, is discussed.