CATRINE researchers; Achraf Qor-el-aine and Alohotsy Rafalimanana, presented their latest project findings at European Geophysical Union ( EGU26) conference 3-8 May 2026 in Vienna.
Achraf Qor-el-aine gave an oral presentation "Diagnosing CO₂ Transport in the North Atlantic Upper Troposphere, evaluating ICON-ART and IFS against in-situ IAGOS observations" in the session on Dynamics and chemistry of the upper troposphere and lower stratosphere (UTLS).
The presentation focused on three transatlantic IAGOS flights from winter 2022 (08 January, 17 February and 20 February) and combined two complementary transport diagnostics: ELIAS 2.0, a machine-learning detector of warm-conveyor-belt (WCB) inflow, ascent and outflow stages, and HYSPLIT 72-hour Lagrangian backward trajectories. Both ICON-ART and IFS were run following the CATRINE modelling protocol so that any inter-model spread is, by construction, a transport signal.
Achraf's presentation showed that all three observed CO₂ spikes are robustly co-located with WCB outflow in the tropopause region, where steep vertical gradients and stratosphere–troposphere exchange amplify transport errors. Both models reproduced the synoptic-scale CO₂ structures, but their performance diverged with the length and complexity of the transport pathway: on the locally-driven 08 January WCB outflow the two models agreed closely, while on the multi-day, trans-Pacific ascent pathways sampled on 17 February (mid-Atlantic peak) and 20 February, IFS systematically underestimated the CO₂ magnitude by ~4–5 ppm whereas ICON-ART reproduced the observed peaks — a signature consistent with cumulative numerical diffusion of filamentary CO₂ structures and with the 24-hour cyclic re-initialization of the IFS meteorology.
The broader message was that, under the CATRINE identical-flux protocol, the inter-model UTLS spread provides an empirical lower bound on the transport-induced uncertainty that propagates into inversion-derived surface fluxes — with direct implications for the operational CO2MVS capacity and for the use of upper-tropospheric aircraft and satellite observations in CAMS-style CO₂ monitoring.
Alohotsy Rafalimanana’s oral presentation was on the CATRINE work on high-resolution simulation of urban CO2 emissions over the Paris area using WRF-Chem: Monitoring urban atmospheric CO2plumes from space: sensitivity to urban physics and scale effects over Paris
The presentation covered the two key sensitivities addressed in a two-part paper currently under review at Atmospheric Chemistry and Physics (ACP): urban physics schemes (Part 1: Atmospheric modeling of the urban boundary layer) and model resolution ranging from mesoscale down to Large Eddy Simulation at 100 meters (Part 2: Representation of fine-scale structures).
Alohotsy’s presentation showed that more sophisticated urban canopy schemes, in particular the Building Energy Model (BEP-BEM), significantly improved the representation of surface heat fluxes, turbulent mixing, boundary layer dynamics, and near-surface CO2 concentrations, especially in winter.
Alohotsy also demonstrated that increasing resolution from 900 m to 100 m captures localized CO2 plume structures much more accurately, with differences between the two resolutions reaching up to 40% near strong emitters, and better reproducing the observed concentration peaks at monitoring stations.
The broader message was that both urban physics and model resolution are critical ingredients for reliable urban greenhouse gas monitoring, with direct implications for the interpretation of satellite observations such as OCO-3.
Images courtesy of the project team.