Related Experiment Video
Updated: Oct 5, 2025
11:34
Scattering And Absorption of Light in Planetary Regoliths
Published on: July 1, 2019
10.5K
Unexplored Antarctic meteorite collection sites revealed through machine learning
Veronica Tollenaar1, Harry Zekollari1,2, Stef Lhermitte2
1Laboratoire de Glaciologie, Université libre de Bruxelles, Brussels, Belgium.
Science Advances
|January 26, 2022
Summary
Scientists used machine learning to map meteorite-rich areas in Antarctica, identifying over 600 potential meteorite stranding zones. This data-driven approach reveals that most Antarctic meteorites remain undiscovered, paving the way for efficient collection.
Area of Science:
- Planetary Science
- Geology
- Astronomy
Background:
- Meteorites offer crucial insights into Solar System origins and evolution.
- Antarctica is a prime location for meteorite recovery due to concentration in stranding zones.
- Current methods for identifying meteorite-rich areas rely on chance and expensive expeditions.
Purpose of the Study:
- To develop a data-driven method for identifying meteorite-rich areas across Antarctica.
- To estimate the probability of finding meteorites continent-wide.
- To guide future meteorite collection efforts.
Main Methods:
- Integration of advanced datasets with a machine learning algorithm.
- Generation of continent-wide meteorite probability maps.
- Identification and validation of meteorite stranding zones.
Main Results:
- Approximately 600 meteorite stranding zones were identified with over 80% accuracy.
- Discovery of previously unknown meteorite-rich areas, some near research stations.
- Estimation that less than 15% of Antarctic surface meteorites have been recovered.
Conclusions:
- A machine learning approach significantly enhances meteorite discovery in Antarctica.
- Vast numbers of meteorites remain uncollected on the Antarctic ice sheet.
- This method enables coordinated and cost-effective future meteorite recovery missions.
