Spatially and temporally explicit bottom-up information of CO2 anthropogenic emissions is needed to initialize transport models, separate biogenic and anthropogenic emissions, and attribute the exact sources (sectors) of anthropogenic emissions. With this aim, a global Prior Emission Dataset (PED) was developed by CoCO2 Task 2.1 providing spatially consistent anthropogenic emissions for the global assimilation system of WP3. However, the global PED cannot include all the latest information available regionally to keep a methodological consistency at global scale (e.g., with consistent use of spatial proxy data). The goal of this task is to complement the global PED by building a global mosaic of regional CO2 emission inventories that includes all the latest information available at regional level without the limitation of using consistent methodology and spatially harmonised dataset. The global mosaic PED, hereafter referred as CoCO2-MOSAIC, includes primarily official or widely used information in each region that can be considered as a regionally accepted reference. As so, the CoCO2-MOSAIC will be used to (i) assess the global PED and (ii) analyse the sensitivity of the WP3 global modelling system to changes in the bottom-up emissions used as input (and in particular their spatial disaggregation).