The interaction between electromagnetic waves and a dynamic sea surface remains an open problem. We propose to use a model for the sea clutter based on stochastic differential equations, in which the clutter is modelled by stationary Markov stochastic processes. Transition probabilities can be used in the context of a moving sensor to synchronize data recorded at different times from different viewpoints. The three parameters of the model, A and alpha for the radar cross section, and B for the speckle, can be estimated by maximum likelihood. We derive here the stochastic differential equations for the model including a target. We derive formula for maximum likelihood estimation of the target parameters, and apply them to numerical experiments to illustrate their performance.