Researchers at Zhejiang University have developed AtomDiT, a machine learning framework that builds a digital twin of dislocations from a single high-resolution electron micrograph. The twin simultaneously reveals dislocation type and position, local stress, and chemical fields. Demonstrated across metals and alloys, including dual-phase systems and hydrogen-segregated grain boundaries, AtomDiT uncovers how stress and chemistry govern dislocation motion, offering a new route to understanding crystal deformation and damage.

A review published in SCIENCE CHINA Earth Sciences synthesizes current knowledge of nitrous oxide (N₂O) production by marine planktonic ammonia-oxidizing archaea (AOA), an important microbial group involved in ocean nitrification. The authors summarize the potential metabolic pathways of AOA-derived N₂O and examine how environmental factors including oxygen, temperature, pH, substrate availability, and organic carbon regulate this process. The review highlights the need to understand multifactorial effects of ocean warming, acidification, and deoxygenation to better predict marine N₂O emissions under future climate change.

In a paper published in SCIENCE CHINA Earth Sciences, a team of researchers led by Professor Jia Xin from the Institute of Environmental Archaeology at Nanjing Normal University reconstructed the environmental evolution and human settlement adaptation processes at the Xuecheng archaeological site in the Lower Reaches of the Yangtze River during the Mid to Late Holocene. By integrating multiple environmental proxies and high precision topographic data, the study revealed how hydrological and landscape changes shaped prehistoric settlement migration and spatial organization, providing new evidence for understanding human environment interactions during the formation of early Chinese civilization.

With the growing demand for in-situ exploration of complex extreme terrains such as polar craters in lunar exploration missions, quadruped robots have demonstrated unique advantages owing to their efficient jumping and locomotion capabilities in weak gravitational fields. However, lunar gravity is only one-sixth of Earth's, which substantially prolongs the flight phase of underactuated jumping. Incorrect takeoff actions on rugged terrain generate rotational torques that cause the robot to continuously rotate in mid-air, with the angular momentum accumulating over time and leading to catastrophic consequences, thus imposing significant uncertainty on landing control. Conventional optimal control methods rely on offline preplanned trajectories and are unable to cope with the variability of unknown rugged terrains; existing deep reinforcement learning models lack effective utilization of temporal causal relationships, relying solely on current states for decision-making and struggling to predict the long-term consequences of actions. Therefore, how to leverage historical sequence information to model temporal dependencies between actions, enabling the robot to possess forward-looking prediction and adaptive adjustment capabilities during the jumping process, has become a key technical challenge for jumping control on complex lunar terrains.