Joint Interpolation of Multisensor Sea Surface Temperature Fields Using Nonlocal and Statistical Priors

This paper addresses the joint analysis of multisource and multiresolution remote sensing data for the interpolation of high-resolution sea surface geophysical fields. As case-study application, we consider the interpolation of sea surface temperature (SST) fields. We propose a novel statistical model that combines two key features: an exemplar-based prior and statistical priors. The exemplar-based prior, referred to as a nonlocal prior, exploits similarities between local patches (small field regions) to interpolate missing data areas from previously observed exemplars. This nonlocal prior also sets an explicit conditioning between the multisensor data. Two complementary statistical priors, namely a prior on the spatial covariance and a prior on the marginal distribution of the high-resolution details, are considered as sea surface geophysical fields that are expected to depict specific spectral and marginal features in relation to the underlying turbulent ocean dynamics. We report the experiments on both synthetic and real SST data. These experiments demonstrate the contributions of the proposed combination of nonlocal and statistical priors to interpolate visually consistent and geophysically sound SST fields from multisource satellite data. We further discuss the key features and parameterizations of this model as well as its relevance with respect to classical interpolation techniques.