EarthCube-Supported “ICEBERG” Team Paves Way for Novel Polar Geosciences Research Methods

June 2, 2021
By: Kimberly Mann Bruch

heather_1.jpg

Iceberg Principal Investigator Heather Lynch is especially noted for using satellite imagery to discover a previously unknown population of Adelie penguins on the remote Danger Islands off the coast of the Antarctic peninsula. Credit: Heather Lynch

Human-led surveys of polar animals like penguins and seals are not only expensive, but often dangerous. Artificial intelligence (AI) techniques, using satellite imagery, has recently been studied as an alternate survey option by a group of EarthCube-supported geoscientists at Stony Brook University in New York. The researchers found that computers can learn to survey polar animals in a way similar to a human and their accuracy was promising for many types of similar surveys. 

Their most recent study, published in Remote Sensing of Environment, focused on training computers to search for Antarctic seals in satellite imagery. Their findings indicated that their AI-based survey method was more than ten times faster than an experienced human observer.

The seal study (SealNet) used “deep learning”, which makes use of computational models that can automatically learn patterns from satellite imagery data instead of being explicitly programmed. The “deep” in “deep learning” refers to the many layers of interconnected processing units in the model that allows it to learn representations of the data at multiple and increasingly complex levels of abstraction.

“One unique aspect of our research is the interdisciplinary nature of our work, as it requires close collaboration between computer scientists, ecologists, and remote sensing specialists,” said Heather Lynch, ICEBERG (Imagery Cyberinfrastructure and Extensible Building-Blocks to Enhance Research in the Geosciences) principal investigator and professor of ecology and evolution at Stony Brook University. “Ecologists like myself have to reach across the aisle and engage in a very meaningful and sustained way with software engineers, website developers, data scientists, visualization experts, and the list goes on and on.”

ICEBERG, which was partially funded by the EarthCube program, aims to build the cyberinfrastructure required to make the most of satellite imagery for geosciences, starting with researcher working in polar areas, where much of this science is already underway, and then branching out to the entire EarthCube community.

This research was funded by NSF EarthCube’s award number 1740595.

About EarthCube

EarthCube is a community-driven activity sponsored by the National Science Foundation to transform research in the academic geosciences community. EarthCube aims to create a well-connected environment to share data and knowledge in an open, transparent, and inclusive manner, thus accelerating our ability to better understand and predict the Earth’s systems. EarthCube membership is free and open to anyone in the Geosciences, as well as those building platforms to serve the Earth Sciences. The EarthCube Office is led by the San Diego Supercomputer Center (SDSC) on the UC San Diego campus. 

 

Media Contact: 

Kimberly Mann Bruch, San Diego Supercomputer Center Communications, kbruch@sdsc.edu

 

Membership Contact:

Lynne Schreiber, San Diego Supercomputer Center EarthCube Office, lschreiber@sdsc.edu

 

Related Links:

EarthCube: https://earthcube.org

Iceberg: https://iceberg-project.github.io/

San Diego Supercomputer Center: https://www.sdsc.edu/

UC San Diego: https://ucsd.edu/

National Science Foundation: https://www.nsf.gov/