Development of an algorithm for the Bayesian fusion of multi-angle, multi-polarisation and multi-frequency remotely sensed data

This work addresses the possibility of retrieving soil moisture information from remotely sensed data in the microwave domain and develops an algorithm to efficiently merge point measurements. The inversion process is based on the Bayes's theorem and applied to data of a radiometer and scatterometer observing the same area. The flexibility of such algorithm allows incorporating as many as possible sources of information, as multi-angle, multi-polarisation and multi-frequency data. An error analysis indicates that multi-polarisation information provides the best results. Advantages, disadvantages and future improvements of this procedure are also discussed in the analysis.