Optimization of satellite biotelemetry data in seaturtles through outlier removal techniques

This paper focuses on optimizing satellite telemetry data for green sea turtles (Chelonia mydas) by applying rigorous preprocessing and geospatial filtering techniques to eliminate outliers. The telemetry data, gathered from turtles equipped with SPOT-375B transmitters, included geographic positions and other relevant parameters. Initial preprocessing filtered out low-quality data, reducing the dataset by an average of 75.85%. A subsequent geospatial filter further removed data points recorded on land, accounting for an additional 5.25% reduction. These steps ensured that only accurate marine location data were used for analysis. The results demonstrate the critical importance of data cleaning and filtering in enhancing the accuracy and reliability of telemetry studies. This approach not only improves the precision of movement pattern analysis but also provides a robust framework for future studies in marine conservation. While effective, the study acknowledges the limitations of relying on coastline data resolution and suggests incorporating environmental variables in future research to further refine data interpretation.