Towards improving regional and national budgets from regional inversion system CSR by using ABLH measurements

Greenhouse gases (GHGs) play a key role in the Earth's climate due to their connection with the planet's energy balance. A scientific understanding of contemporary GHG fluxes and their drivers is essential for any climate change mitigation policies being implemented or considered. Furthermore, regular, scientifically accurate information on recent GHG emissions supports the implementation of existing policies by allowing to identify - and act on – potential deviations.

Atmospheric inversion systems, such as CarboScope Regional and ITMS-Demonstrator, are used to support climate policies (e.g. Paris Agreement) within the German ITMS (Integrated Greenhouse Gas Monitoring System). They provide independent estimates of anthropogenic and natural greenhouse gas emissions (currently targeting CO₂ and CH₄) on regional scales, with a particular focus on national anthropogenic fluxes. One of the main sources of uncertainty in fluxes retrieved by inversion systems is the misrepresentation of the height of atmospheric boundary layer’s height (ABLH). At the regional scale, inaccuracies in the ABLH can lead to biases in the estimated GHG fluxes that are challenging to identify and quantify, as they can vary in both space and time.

In this study, we evaluate the performance of the STILT (the transport model in the CSR system) and ICON (playing the same role for the ITMS-Demonstrator) in representing the ABLH across Europe. We compare the model performance with high-resolution radiosonde data collected over Europe between 2021 and 2023. We present the spatial and seasonal patterns of bias, and furthermore, we evaluate the models' performance against ceilometer data collected in Germany by the German Weather Service (DWD) ceilometer network. Finally, we show preliminary results on the application of Kriging with External Drift (KED) applied to the European and German domains, exploring the potential of using spatially resolved, data-driven fields as input for ABLH corrections in the inverse model.