ECCOViz: Interactive Multivariate Visualization of Ocean Dynamics in the ECCO LLC4320 Dataset

High-resolution ocean simulations produce large, multivariate datasets that characterize complex physical processes across space and time. Variables such as temperature, salinity, and vertical velocity provide complementary information about ocean structure and dynamics, but examining these fields separately can make their spatial relationships difficult to interpret. This work presents ECCOViz, which focuses on the joint visualization of these three variables within the ECCO LLC4320 dataset. Salinity isosurfaces are used to represent the internal structure of the water column; the local temperature anomaly is mapped onto these surfaces to reveal thermal variations. Vertical velocity extremes determined by percentile-thresholding are introduced to identify regions of significant upward and downward motion. A case study of the north Atlantic region is conducted to prove the effectiveness of the proposed approach. The resulting images emphasize the connections between U-shaped salinity surfaces and regions of up-welling and down-welling. By integrating these variables into a single interactive visualization, the proposed approach supports the exploration of thermohaline structure and localized vertical exchange in high-resolution ocean simulation data. 

You can access the project here https://github.com/shunliru/SciViz-ECCO-Webapp