Merged Sea Ice Motion from Pattern-Matching and Feature-Tracking Vectors

<p>Under the background of global warming, sea ice changes rapidly. Sea ice drift is an important indicator for sea ice flux, atmospheric and ocean circulation, and ship navigation. Considering that pattern-matching vectors and feature-tracking vectors retrieved from remotely sensed data provide complementary information, a novel sea ice motion estimation method was proposed by merging these two types of vectors. In this study, the DTU sea ice product was used as pattern-matching displacements, whereas feature-tracking vectors were derived from Sentinel-1 SAR images. The proposed method of sea ice motion retrieval is summarized as follows. Initial feature-tracking vector fields were first derived from Sentinel-1 SAR image pairs. Then, mismatched feature-tracking vectors were discarded based on a bad matches filtering step. Finally, the filtered feature-tracking vectors were merged with pattern-matching vectors based on Co-Kriging algorithm. The proposed method was evaluated by comparing merged vectors with buoy displacements, DTU sea ice drift product and filtered feature-tracking vectors. The results showed that the proposed method can retrieve accurate, high resolution and evenly distributed sea ice motions.</p>