Deep learning + Sentinel-1 radar yields first high-res Antarctic map of nearly 40,000 grounded icebergs
anselm · x · 2026-09-14
Scientists combined deep learning with Sentinel-1 synthetic aperture radar to produce the first high-resolution, continent-wide dataset of grounded icebergs around Antarctica, identifying nearly 40,000 stationary icebergs.
- Published in Earth System Science Data (ESSD, 18, 6017–6041) as a highlight paper
- Surprising finding: most icebergs are small, but together they may help anchor coastal sea ice and shape Antarctic ecosystems
- Demonstrates how AI-driven Earth observation reveals changes at scales never before visible
More from Research
- There is no single "AI alignment problem": researcher offers a structured taxonomy of its many faces — Dr_Atoosa · 2026-09-14
- Nature: Scientists test whether LLMs trained on historical data can reproduce creative breakthroughs — TZahavy · 2026-09-14
- alphaXiv open-sources OpenResearch, turning coding agents into reproducible research agents — aigclink · 2026-09-14
- Toby Ord: 10x more RLVR compute cuts tokens-to-target ~3x; gains may be math-specific — tobyordoxford · 2026-09-14
- New scaling curve has half the slope: 10,000x compute for 20%-to-80%, but bigger generational jumps — tobyordoxford · 2026-09-14
- Fudan NLP paper explains why max reasoning settings can backfire on SWE benchmarks — karminski3 · 2026-09-14