Earth observation provides repeated and consistent measurements across most of the planet, including regions where field observations and monitoring infrastructure are limited. AI makes it possible to process these data systematically and to develop global workflows that study different regions through a consistent analytical framework and at comparable levels of detail. This combination can widen access to evidence on climate change and reduce some of the geographical, economic and technical barriers that still shape climate research.

Global coverage, however, does not automatically guarantee global knowledge. Models developed in data-rich regions may perform differently under other environmental or socioeconomic conditions. Their evaluation is often constrained by sparse reference data, uneven monitoring capacity and limited local validation. Geographical bias, uncertainty and transferability therefore need to be assessed explicitly rather than assumed.

This session focuses on the use of AI and Earth observation to make climate change research more accessible, systematic and geographically inclusive. It welcomes studies of climate impacts, risks, vulnerability and adaptation, including their links with natural hazards, public health, ecosystems, population, infrastructure, early warning and disaster response.

Relevant contributions may address scalable and reproducible workflows, validation in data-scarce settings, uncertainty assessment and generalisation across regions. Studies combining satellite observations with field measurements or local knowledge are particularly relevant, as are open data, accessible tools and methods that can be adapted to different contexts. The aim is to understand how global Earth observations can make robust and actionable climate knowledge available across different communities and contexts, including those where conventional monitoring and research capacity remain limited.