Tracking a single pigeon using a shadowing filter algorithm

Area of Science:

• Animal movement ecology
• Biophysics
• Data science

Background:

• Miniature GPS devices provide high-resolution animal tracking data.
• Existing methods struggle with measurement errors and extracting full dynamical states (velocity, acceleration) from position data alone.
• Direct differentiation of position data is often inadequate for accurate state estimation.

Purpose of the Study:

• To implement and validate a shadowing filter algorithm for reconstructing the full dynamical state space of animal movement.
• To minimize measurement errors inherent in GPS tracking.
• To provide a practical prototype for applying Newtonian dynamics-based filtering to real-world animal tracking data.

Main Methods:

• Development and application of a shadowing filter algorithm based on Newtonian dynamics.
• Minimization of measurement errors from GPS position recordings.
• Reconstruction of velocity and acceleration from position data for a flying pigeon.

Main Results:

• Successful implementation of the shadowing filter algorithm.
• Demonstrated minimization of measurement errors.
• Accurate reconstruction of the full phase-space (position, velocity, acceleration) from GPS data.
• Exploration of filter parameter selection and minimum data requirements.

Conclusions:

• The shadowing filter offers a robust method for enhancing animal tracking data quality and extracting detailed movement dynamics.
• This approach provides a practical solution for overcoming limitations of standard GPS data analysis.
• The study presents a prototype for wider application of shadowing filters in animal movement studies.