GEOTRACES Online Seminar Series 2026

April 28, 2026
First Seminar — April 28, 2026. Topic: Data synthesis, modelling & integration
Moderator: Dr. Anh Pham (University of California, Los Angeles)

GEOTRACES Online Seminar Series 2026

The GEOTRACES Early Career Scientist (ECS) Committee has launched a new monthly seminar series bringing together senior scientists and early career researchers around shared themes in trace element and isotope biogeochemistry. Each seminar features two connected talks — a senior scientist providing synthesis and big-picture framing, and an early career researcher presenting new data, methods, or emerging perspectives. Seminars run from April to November 2026 on the last Tuesday of each month, at either 09:00 or 15:00 CEST. No registration required. Recordings are posted on the GEOTRACES YouTube Channel.

For more information and upcoming seminars visit the seminar web page.


First Seminar — April 28, 2026 Topic: Data synthesis, modelling & integration Moderator: Dr. Anh Pham (University of California, Los Angeles)

Dr. Gregory De Souza (ETH Zurich, Switzerland) Bringing it all together: using models to build process understanding

Two decades of GEOTRACES observations provide a wealth of TEI data at a range of spatial scales. This talk gave an overview of the diversity of approaches that have used conceptual and numerical models to synthesize observational data into an understanding of TEI cycling in the sea and at its interfaces.

Dr. Arianna Olivelli (Flanders Marine Institute, Belgium) Mapping ocean trace metal distributions: A global data-driven view of lead and its isotopes

Despite the rapid growth of GEOTRACES observations, resolving large-scale distributions of trace metals remains challenging due to sparse and heterogeneous coverage. This talk presented the first global maps of lead (Pb) concentrations and isotope compositions from surface to depth using the tree-ensemble algorithm XGBoost and explainable machine learning tools, providing an observation-based view of Pb distributions that is difficult to achieve from measurements alone.