Estimation of carbon dioxide fluxes in the city of Paris using the ICON-ART-CTDAS model

Cities around the globe are aiming to reduce their carbon dioxide emissions, but monitoring and validating urban CO₂ emissions is a major challenge. This motivates the ICOS cities PAUL project to use a combination of different measurement and modelling techniques to provide observation-based emission estimates in three pilot cities: Zurich, Paris, and Munich. The challenge comes due to large variations in emissions and concentration gradients, high uncertainties in prior estimates, and inherent modeling errors. Here we present the results of an inverse modeling study for the city of Paris, which builds on the insights gained from similar simulations conducted for Zurich. Our approach employs the state-of-the-art atmospheric mesoscale model ICON-ART, which we ran in conjunction with an ensemble Kalman smoother to optimize CO₂ fluxes based on simulated and measured concentration differences.

Paris offers advantages for mesoscale model simulations due to its flat terrain and large size, unlike Zurich, where simulations were challenged by the city's complex topography. Furthermore, the CO₂ measurements in Paris, which were collected from a network of 2 tall towers inside and 7 towers outside the city, were easier to represent by the model due to their larger spatial representativeness compared to the more locally influenced rooftops measurements in Zurich.

The ICON-ART model simulations were performed for two offline nested model domains. The outer domain covers Central Europe with a spatial resolution of 6.5 km and was chosen large enough to serve as initial and boundary conditions for the simulations over both Zurich and Paris. The inner, high-resolution domain is centered on the Île-de-France region with a spatial resolution of 1 km. According to our previous experience with Zurich simulations, the atmospheric transport is well simulated by ICON-ART in most weather situations with the exception of low wind conditions, where relative errors in wind speeds and the corresponding dilution of CO₂ emitted from the city are the largest. The prior anthropogenic CO₂ fluxes were based on the anthropogenic inventory data prepared by AIRPARIF for the Île-de-France area at 0.5 km spatial resolution and on TNOGHGco 2018 data (1 km) for the rest of Europe. Biogenic fluxes were computed online using the Vegetation Photosynthesis and Respiration Model (VPRM), integrated online into ICON-ART.

In this presentation, we analyze the performance of ICON-ART model against meteorological and CO₂ observations in and around Paris, and demonstrate initial results from emission inversion experiments. Furthermore, we contrast the results with those obtained for Zurich to emphasize the different challenges and modelling capabilities in the two cities.