Impact of model resolution and transport uncertainties on the representation of CO2 emission plumes by global and regional atmospheric transport models

Emissions of carbon dioxide (CO₂) will have to be drastically reduced in the coming decades to reach the goal of the Paris Agreement to limit the global temperature increase to no more than 2°C. To support this process, Europe is planning to establish a CO₂ anthropogenic emission monitoring system, which will assist countries, cities and facility operators in monitoring their emissions and evaluating the progress towards their reduction targets. The system will combine measurements from ground-based networks with observations from a new constellation of CO₂ satellites, which will provide high-resolution images of total column CO₂ allowing tracking the plumes of large emission sources. A suite of atmospheric transport modelling systems will assimilate these observations and inversely estimate emissions from the continental to the country scale and down to the scale of individual cities and power plants.

In the European project "CO₂ Human Emissions" (CHE), the components of such a modelling framework are explored, which includes the generation of a library of realistic atmospheric CO₂ simulations. These "nature runs" are obtained by running global and regional atmospheric transport models at the highest possible resolution affordable today and using state-of-the-art inputs of anthropogenic emissions and natural CO₂ fluxes. The library includes global simulations at 9 km x 9 km resolution with the CAMS-IFS model, European simulations at 5 km x 5 km resolution with WRF-GHG, COSMO-GHG and LOTOS-EUROS, and high-resolution simulations at 1 km x 1 km over the city of Berlin and several power plants with COSMO-GHG and LOTOS-EUROS.

Here we analyse and compare the model simulations to address the following questions: How realistically are atmospheric gradients in CO₂ caused by spatial and temporal variations in biospheric and anthropogenic fluxes and by atmospheric dynamics represented at the different model resolutions? What resolution is required to resolve the plumes of individual cities and power plants? How large are the differences in near surface and total column CO₂ due to uncertainties in atmospheric transport including uncertainties in vertical mixing? Information on transport uncertainties is derived from an ensemble of CAMS-IFS simulations and from the spread between the individual models.

Answering these questions is critical for the design of a future operational capacity to monitor anthropogenic CO₂ emissions, which should optimally support decision makers at facility, city, and country scale as well as the global stocktake process of the Paris Agreement.