Application of on-line system for monitoring and forecasting ocean surface changes around Korean Peninsula

This study applies an on-line system, which employes an adaptive reconstruction technique to monitor and forecast ocean surface changes. The system adaptively generates an appropriate synthetic time series by recovering missing measurements for sequential images. The reconstruction method incorporates temporal variation according to physical properties of targets and anisotropic spatial optical properties into image processing techniques. This adaptive approach allows successive refinement of the structure of objects that are barely detectable in the observed series. The system sequentially collects the estimated results from the adaptive reconstruction and then statistically analyzes them to monitor and forecast the change in surface characteristics.