Spatio-temporal Knowledge Discovery in Very Large METOC Data Sets

A system allowing for the efficient processing and viewing of dense METOC data sets stored in Network Common Data Format (netted) files is developed using advanced bitmap indexing. A method for netted data extraction and bitmap index creation is presented. Efficient geospatial range and pseudo-KNN queries are implemented. A two step filtering algorithm is shown to greatly enhance the speed of these geospatial queries, allowing for extremely efficient processing of the netted data.