Spatiotemporal Tracking of Dynamic Structures

Academic Research · Algorithm Design & Data Processing · TU Delft

Contributed to building a full-stack Python pipeline to detect, filter, and track dynamic spatiotemporal structures (oceanic eddies) from large-scale, noisy satellite data.

Python Satellite Data Geophysics Tracking Completed

technical stack

detection pipeline

Sea level anomaly map of an eddy
(a) Sea level anomaly map of an eddy
Geostrophic velocities over SLA map
(b) Geostrophic velocities overlaid on SLA map
Okubo-Weiss parameter W
(c) Okubo–Weiss parameter W
Eddy point locations
(d) Detected eddy points (after correction & filtering)
Figure: Main steps of eddy detection. Sea-level anomaly map (a). Geostrophic velocities used to compute W (b). The Okubo–Weiss parameter (c). Detected eddy points (d).

key contributions