Estimating emissions from urban areas on the basis of measurements is a challenging subject and typically relies on three key components: measurements of greenhouse gas (GHG) concentrations (or turbulent fluxes), modelling of the atmospheric transport and prior information on the spatio-temporal structure of the emissions. The high spatial and temporal heterogeneity of urban areas and their emission patterns is especially challenging for atmospheric transport models and gridded inventories, which are currently pushed to resolutions of kilometers and below in an effort to accurately represent these effects. In-situ systems remain the backbone of GHG concentration measurements in this context. However, it remains unclear how representative they really are for kilometer-scales. This open question of spatial representativeness becomes increasingly important with the ever-improving quality of measurements, models, and inventories themselves.

For more than a year, I performed continuous open-path measurements of CO₂ and CH₄ along a 1.55 km long absorption path over Heidelberg to compare them to in-situ measurements. While both measurements are mostly in good agreement, analysis reveals differences of up to 20 ppm in CO₂ for specific wind directions, most likely a result of a local atmospheric transport phenomenon. Further, the two measurements show differences in CH₄, which is potentially a result of different sensitivities to local emissions. These, in some cases different, sensitivities of either measurement approach to local source patterns showcases the potential of the continuously growing dataset to assess the representativeness of the different measurement approaches, but also the performance of atmospheric transport models and emission inventories in urban environments.