Developing responses to the challenges posed by Artificial Intelligence (AI) and Machine Learning (ML) is of rapidly growing strategic importance for the CGMS community. AI/ML has become a cross-cutting topic affecting virtually all aspects of satellite meteorology, spanning satellite operations, on-board and on-ground data processing, product generation, numerical weather prediction, nowcasting, climate applications and future space system architectures.
CGMS will work to identify common priorities and establish a coordinated and collaborative approach within the CGMS community, recognising the need to cooperate with existing international initiatives, in particular WMO efforts on AI matters.
Many AI-related challenges, particularly those concerning satellite data, metadata, interoperability and technological infrastructure, are common across organisations. A joint WMO/CGMS AI workshop will therefore be developed with the objective to bring together experts from operational agencies, research institutions and international organisations to identify common challenges, exchange experiences and develop shared guidance.
The development of Best Practices has emerged as one of the principal themes of the broader discussion on AI/ML. CGMS considers that such guidance should initially focus on well-established application areas, including satellite-based nowcasting and numerical weather prediction, while remaining sufficiently flexible to accommodate the rapid expansion of more general AI/ML applications. Any guidance developed should promote openness and innovation, enabling the community to adopt emerging technologies to significantly enhance the discoverability, accessibility and exploitation of satellite data, while avoiding fragmented methodologies, duplicated efforts or incompatible practices.
A key element of the collaborative framework will be the development of a shared CGMS catalogue to support the discovery and reuse of AI/ML applications, expertise and AI-ready data formats.