Researchers from Sun Yat-sen University, the University of Chinese Academy of Sciences, and North Carolina State University have created a dataset tracking monthly surface air temperatures across Antarctica from 1979 to 2023. The dataset covers the entire continent and helps fill major gaps in climate records, especially in remote areas with little or no direct measurement. To build it, the team used deep learning to estimate missing temperatures, training their model on over 250 sources, including weather stations, ocean observations, and past climate data from global models. The result offers a much clearer picture of how Antarctica’s climate is changing, especially the strong warming seen in West Antarctica and the Antarctic Peninsula.
Image Credits: henrique setim
