Several researchers from UMR TETIS are taking part in the organisation of the Workshop on MAChine LEArNing for Earth Observation (MACLEAN), to be held in Vilnius from 9 to 13 September. The vast amount of data currently being produced by modern Earth observation missions and surface measurements is raising new challenges for the scientific community in the fields of remote sensing and atmospheric modelling. Sensors are now offering (very) high spatial resolution images at revisit frequencies never before achieved, by providing different signals, for example multi-(hyper)spectral optical signals, radar data, LiDAR and digital surface models. On the other hand, atmospheric composition and processes are measured at the Earth's surface, using molecular-scale measurements with mass spectrometers, particle counters and more traditional meteorological instruments.

Modern machine learning techniques can be crucial for dealing with such heterogeneous, multi-scale and multimodal data. Among the methods receiving attention in this area are deep learning, domain adaptation, semi-supervised approach, time series analysis, active learning, explainable artificial intelligence, uncertainty quantification, and interactive model building and visualization. Although machine learning and

This workshop is intended as an international forum where researchers in the above-mentioned fields can meet to exchange ideas, debate and define short- and long-term research objectives around the exploitation and analysis of Earth observation and atmospheric data using machine learning techniques. The workshop aims to provide an overview of current research on machine learning for Earth observation data and other atmospheric measurements. On the other hand, the organisers hope to stimulate concrete discussions to pave the way for new machine learning frameworks specifically designed to handle the data concerned.