The deliverable provides a synthesis of the lessons learned from the activities in CoCO2 WP4.
It provides a review of the conclusions from the previous deliverables D4.1-7 in this work
package as well as new conclusions from the analysis in tasks T4.1-4. The main conclusions
lead to recommendations and guidance regarding the spatial resolution of the atmospheric
transport modelling and inversion and the observations to tackle the monitoring of the CO2
anthropogenic emissions, the development of specific branch of local scale inversion based
on computationally light data driven methods for the operational process of the XCO2
spaceborne images of plumes downwind of emission hotspots, the coupling of such a branch
with the CO2MVS multi-scale inversion prototype, the development of machine learning
techniques for the local scale inversions, and regarding the benchmarking and intercomparison
of inverse modelling configuration at local and national scale, including the use of
modular open source community codes. Further analysis and inter-comparisons with more
complex synthetic tests cases or experiments using real data will be needed to refine the
configuration assessment of the local scale inversion techniques based on lightweight
techniques or machine learning, but the experiments in WP4 already provide strong insights
on their strengths and accuracy, and on their level of readiness for operational applications.
Large (national) scale inversions of anthropogenic CO2 emissions keep on being more
exploratory, and may finally have to connect to local scale inversions via the coupling of
systems or scales within the multi-scale inversion prototype, or via the gradual increase in the
spatial resolution of the national scale systems.
Abstract
Files