Satellite Detection Model
An open-source pipeline that predicts, ahead of time, which satellites are about to streak across an observatory’s exposure.
Developing a Python simulation pipeline that ingests real-time Two-Line Element data from Space-Track and propagates orbits with SGP4. The result is the Satellite Detection Model, an open-source computational pipeline for predicting satellite interference in astronomical observations.
Given a pointing and an exposure window, the model propagates the current catalogue forward and reports which objects cross the field and when — turning satellite contamination from something you discover in the reduced data into something you can schedule around. The work was mentored by Dr. Niall Deacon at the MPIA, with input from Dr. Meredith Rawls at the University of Washington.
The ongoing extension is an AI Space Debris Analyzer: a tool that uses publicly available Space-Track data on orbiting man-made bodies to predict and quantify interference in ground-based optical telescope observations, specifically targeting the Vera Rubin Telescope. It estimates the expected debris-related noise within specific time windows, helping astronomers plan observations that minimise contamination.
Context
- Mentorship — Dr. Niall Deacon, Max Planck Institute for Astronomy; discussion with Dr. Meredith Rawls, University of Washington.
- Collaboration — actively seeking collaborators.
- Methods — TLE ingestion from Space-Track, SGP4 orbit propagation, field-crossing prediction.