Presentation: The Palaeo Data Cube: Introducing Next Generation Sharing of Geological Data

Presentation given during the American Geophysical Union (AGU) 2025 Annual Meeting in New Orleans (LA, USA), within Session "IN52A - Advancements in Earth Science Data Access and Visualization Through Emerging Technologies". Sharing of Earth observation (EO) assets like satellite imagery has greatly increased in the last two decades, with ever evolving technical advancements recuding the barrier for end users to access this data. These advancements have for instance been possible due to the devleopment of national to continental EO Data Cubes in Australia, Switzerland, Brazil, Africa, Armenia, of over Pacific Island nations. These advancement are often based on standard protocols and community developed tools that have made it easier to describe data (GeoNetwork), query it (GeoServer) and catalog it (STAC). Moreover, the growing use of cloud computing has increased the use formats such as cloud optimized GeoTIFFs (COGs). In other fields of Earth science however, these capabilities are applied in a very limited number of cases. We present the Palaeo Data Cube, a new framework to share global, high resolution thematic maps on the Earth evolution over the Phanerozoic (last 545 million years), based on the PANALESIS model, and using standard and open source solutions. The Palaeo Data Cube grants access to 225 maps from 5 image mosaics (palaeogeography, lithospheric thickness, crustal thickness, hydrothermal penetration depth and seafloor ages). Maps can be accessed for 45 time steps via WMS and are all indexed in a STAC catalog, with 5 collections (one per mosaic), for a total of over 1 billion pixels. All collections are also documented using GeoNetwork and archived in Zenodo. Through the Palaeo Data Cube, we aim to accelerate the dissemination of global maps derived from plate tectonic models, and other sources, and easy the access and use of these products to geoscientsts. We also put reproducubility at the core of our framework, by sharing and tracking versions of the code used to process the data, alongside the output themselves.