Association Analysis of Ocean Salinity and Temperature Variations

Ocean salinity and temperature variations have received increasing attention due to their relation to global climate changes such as warm winters, the universal drought, heat waves, floods and snowstorms. The factors that affect climate changes vary across the world. For example, Taiwan is subtropical where the climate is influenced by airstreams and ocean currents. In this study, association rule mining is used to discover interesting patterns from Argo ocean salinity and temperature data sampled in the waters surrounding Taiwan. The discovered rules may have the form: if the first month salinity rose from 0.15 psu to 0.25 psu in the A1 area and the second month temperature rose from 0degC to 1.2degC in B1 area, then the third month salinity will rise from 0.1 psu to 0.2 psu in the A2 area. A visualization interface is provided to display the association rules. Simulation results from Argo data are presented to verify the applicability of the proposed system.