Modelling representation errors of atmospheric CO2 concentrations at a regional scale

Inverse modelling of carbon sources and sinks requires an accurate estimate of the quality of the observations to obtain a realistic estimate of the inferred fluxes and their uncertainties. Representation errors, defined here as the mismatch between point observations and grid cell averages, may add substantial uncertainty to the interpretation of atmospheric CO2 concentration data. We used a high resolution (2 km) mesoscale model (RAMS) to simulate the variations in the CO2 concentration to estimate the representation errors for grid sizes of 10-100 km. Meteorology is the main driver of representation errors in our study causing spatial and temporal variations in the error estimate. Within the nocturnal boundary layer the representation errors are relatively large and mainly determined by unresolved topography at lower model resolutions. During the day, surface CO2 flux variability and mesoscale circulations were found to be the main sources of representation errors. Careful up-scaling of point observations can reduce the importance of the representation error substantially. The remaining representation error is in the order of 0.5-1.5 ppm at 20-100 km resolution.