Toggle light / dark theme

More pocked with craters than any other object in our solar system, Jupiter’s outermost and second-biggest Galilean moon, Callisto, appears geologically unremarkable. In the 1990s, however, NASA’s Galileo spacecraft captured magnetic measurements near Callisto that suggested that its ice shell surface—much like that of Europa, another moon of Jupiter—may encase a salty, liquid water ocean.

But evidence for Callisto’s subsurface ocean has remained inconclusive, as the moon has an intense . Scientists thought this electrically conductive upper part of the moon’s atmosphere might imitate the magnetic fingerprint of a salty, conductive ocean.

Now, researchers have revisited the Galileo data in more detail. Unlike in prior studies, this team incorporated all available magnetic measurements from Galileo’s eight close flybys of Callisto. Their expanded analysis much more strongly suggests that Callisto hosts a subsurface ocean.

The bibliometric infrastructure of citations has become an inescapable organising feature of academic life. Drawing on a range of evidence of the use and misuse of citations data, Stuart Macdonald argues its ubiquity has rendered authorship a questionable concept in modern scholarship.

The same material from which you drink your morning coffee could transform the way scientists detect disease, purify water, and insulate space shuttles thanks to an entirely new approach to ceramic manufacturing.

Published in Advanced Science, 3D-AJP is an aerosol jet 3D nanoprinting technique that allows for the fabrication of highly complex ceramic structures that—at just 10 micrometers (a fraction of the width of human hair)—are barely visible to the naked eye. These 3D structures are made up of microscale features including pillars, spirals, and lattices that allow for controlled porosity, ultimately enabling advances in ceramic applications.

“It would be impossible to machine ceramic structures as small and as precise as these using traditional manufacturing methods,” explained Rahul Panat, professor of mechanical engineering at Carnegie Mellon University and the lead author of the study. “They would shatter.”

Summary: Researchers have developed a geometric deep learning approach to uncover shared brain activity patterns across individuals. The method, called MARBLE, learns dynamic motifs from neural recordings and identifies common strategies used by different brains to solve the same task.

Tested on macaques and rats, MARBLE accurately decoded neural activity linked to movement and navigation, outperforming other machine learning methods. The system works by mapping neural data into high-dimensional geometric spaces, enabling pattern recognition across individuals and conditions.