An artificial neural network developed by University of California, Irvine (UCI) researchers can autonomously recognize and quantify the edges of glaciers in satellite images with far greater reliability than humans.
The team trained the Calving Front Machine (CALFIN) network on tens of thousands of images; afterwards, CALFIN could measure calving fronts to within an average of 82 meters (269 feet) from their true locations, outperforming previous models.
Said William Colgan of the Geological Survey of Denmark and Greenland, “I think machine learning now offers a robust way of upscaling a handful of site-specific and process-level observations to tell a bigger regional story."
From Columbia Climate School, Columbia University
View Full Article
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA
No entries found