High-resolution CO2 flux modelling on the building-scale using GRAMM/GRAL and in-situ measurements for the Paris metropolitan area

The city of Paris aims to reach net zero emissions by 2050, an ambitious target whose achievement will need to be verified. Atmospheric measurements of CO₂ can provide independent information on the city emissions and therefore, play an important role in monitoring the effectiveness of emission reduction plans.   

To derive emissions from measured concentrations, an atmospheric transport model is needed. This model should cover long time periods to detect trends and emission patterns, and run at high-resolution to simulate the air flow around urban structures. We use GRAMM/GRAL to model CO₂ transport over Paris at 10m resolution with a catalogue approach. The hourly occurring meteorological situation and its respective concentration field is selected from a catalogue of around 1000 precomputed meteorological conditions, which are representative of wind situations over Paris. The selection of the appropriate catalogue entry is based on minimizing differences to wind measurements in the modelling domain. Thus, long time series of concentration enhancement maps can be calculated with low computational costs. Our setup for Paris includes anthropogenic fluxes, biogenic fluxes from Sentinel-2-based VPRM, and boundary conditions derived from in-situ measurements to allow a direct comparison to the observed concentrations in the city. 

We compare the simulated CO₂ concentrations to measurements for 2023 from the ICOS Cities project. The modelled signals generally capture the diurnal dynamics and agree with the measured CO₂. There are certain meteorological conditions where GRAMM/GRAL fails to capture the measured signal. GRAMM/GRAL does not accurately capture meteorological situations with lower boundary layer heights which most often occur during nighttime and in winter. However, we present a method of estimating a time-dependent uncertainty using concentration distribution from multiple catalogue entries. This uncertainty can be used in an inversion.  

We determine the underlying emission patterns and analyse the importance of the resolution of the emission inventory for emission quantification and emission sector disaggregation. Such detailed sector-specific information can help to inform policymakers about progress towards reduction goals and the effectiveness of specific reduction measures.