IceBoost v2.0 AI model maps global glacier ice volumes

Researchers from Ca' Foscari University of Venice, in collaboration with the National Research Council's Institute of Polar Sciences (CNR-Isp) and international partners including NASA's Jet Propulsion Laboratory, have developed IceBoost v2.0. This new machine learning model aims to redefine the global map of terrestrial glacier ice volumes.

By analyzing over seven million direct ice thickness measurements and integrating 26 physical and geometric variables—such as terrain slope, curvature, ice flow velocity, and local temperature—the model reconstructs the thickness and volume of glaciers listed in the Randolph Glacier Inventory. The system provides a spatial and topographic distribution of ice that is up to 40% more accurate compared to previous models.

The study, published in Scientific Data (Nature), estimates that the world's glaciers contain approximately 150,000 cubic kilometers of ice. This volume, if completely melted, would result in a 32.3-centimeter rise in global mean sea levels (excluding the Antarctic and Greenland ice sheets). An interactive web application has been released to allow experts and the public to explore these detailed mapping data.