Data Visualization
Copernicus
- Date
- June 2026
- Client
- Copernicus Marine Service — Dataviz Challenge
- Deliverables
- Infographics, interactive explorative visualization
- Tools used
- D3.js, React, Adobe Illustrator
The Copernicus Marine Dataviz Challenge invites designers and data visualization experts to transform open marine data into compelling visual stories about the ocean.
I chose to work with sea water temperature data from the Global Ocean Ensemble Physics Reanalysis: a dataset that captures long-term ocean temperature trends across the world's major basins. What makes this data particularly compelling to me is the tension between its scale and its specificity: it spans decades and the entire globe, yet tells unmistakably local stories. Each ocean basin warms at its own pace, driven by a unique mix of circulation patterns, atmospheric exchange, and geographic context. I was especially drawn to the question of depth: not just how the surface warms, but how far that heat actually penetrates into the water column, and how that varies between basins and over time. Visualizing these differences side by side makes visible something that statistics alone cannot convey.
To serve two different audiences and purposes, I created both a static infographic and an interactive visualization. The static piece is designed for immediate impact: a single, carefully composed image that communicates the key story at a glance, suitable for publication and social sharing. The interactive version is for exploration: it lets viewers move through ocean basins and time themselves, shifting from passive reading to active discovery. Together, they make the same data work at two different depths (no pun intended) of engagement.
Data Processing
- 1Source — Monthly sea water temperature (thetao_mean) from the Copernicus Marine reanalysis GLOBAL_MULTIYEAR_PHY_ENS_001_031, covering January 1993 through December 2024.
- 2Depth clip — Clipped to 2000 m. Below that, the data becomes unreliable in the early years: Argo floats didn't achieve global coverage until around 2006, so deeper readings before then are model guesses rather than real observations.
- 3Anomaly — For each location and depth, subtracted the average temperature for that same month over 1993–2010. This turns absolute temperatures into deviations from a stable baseline, making the long-term warming visible.
- 4Deseasonalize — Removed the regular seasonal cycle so the chart shows the underlying trend, not just summer warming up and winter cooling down.
- 5Color mapping — Scaled anomalies to a 0–1 range and applied the Ember colormap: cool tones near baseline, deepening to bright yellow-white at the hottest end.
Interactive Visualization
Explore the data yourself