This paper addresses the problem of estimating abrupt changes in synthetic aperture radar (SAR) images. The problem is formulated as a Bayesian estimation problem. Appropriate priors allow to take into account the correlations between images recorded at different dates. Unfortunately, the posterior distribution of the unknown parameters cannot be expressed in closed-form. The proposed Bayesian implementation consists of generating samples distributed according this posterior by using MCMC methods. These samples are then used to estimate various interesting features including the posterior changepoint probabilities and the posterior changepoint number. Simulations on synthetic data illustrate the proposed methodology.