Cross-assessment of oceanographic observations from sea turtle-borne sensors, argo floats, satellite data, and reanalysis products in the Mozambique Channel

Abstract

In this study, we aim to evaluate the quality and complementarity of oceanographic data collected by instrumented sea turtles, in comparison with satellite-derived products and ocean reanalysis datasets. This work was conducted within the framework of the international STORM (Sea Turtle for Ocean Research and Monitoring) project, using equipped turtles with satellite telemetry and environmental sensors (TD and CTD tags), collecting temperature, depth, and salinity data along their trajectories, enabling direct comparison with remotely sensed products (OSTIA, ESA SST CCI, SMOS-OI) and global reanalysis models (GLORYS12V1, GLORYS2V4). Turtle-derived SST showed high correlation and low error with OSTIA and ESA SST CCI, while reanalysis models tended to overestimate SST, especially nearshore. Vertical temperature profiles aligned well with ARGO and reanalysis data in the upper layers. Salinity data were reliable during the first weeks of deployment, but accuracy decreased over time, likely due to biofouling and potential freshwater inputs in coastal areas, which may cause sensor drift. Spatial bias analysis revealed greater discrepancies close to the coast. These results confirm the value of sea turtles as mobile platforms for ocean observation in data-sparse tropical regions.