GEOSTATISTICAL MODELING OF DISSOLVED-OXYGEN DISTRIBUTION IN ESTUARINE SYSTEMS

The probabilistic random function model is used in geostatistics to infer the attribute under study at unsampled locations. In many earth sciences applications, sparse information and incomplete sampling are not uncommon. This problem sometimes can be approached with an appropriate definition of the stationary random function model being used. In spatial statistics this decision is usually influenced by practical factors, such as availability and distribution of the samples. This paper describes a procedure that trades the third dimension for a larger number of samples in the two-dimensional space. This allows for a better characterization of the distribution of dissolved oxygen within the Mar Piccolo of Taranto, Italy.