Well-posed Bayesian inverse problem in total variation

A Bayesian inverse problem is well posed in total variation when a unique posterior exists for every datum and the map from data to posterior is continuous in total variation distance.