New inversion method with expanded potential

August 10, 2026

As part of the ITMS project, an inversion method was developed that enables both observation-based emission estimates and the optimization of concentration fields. The scientific findings have now been published. The study confirms the ITMS demonstrator results and strengthens confidence in the ITMS inversion system and its modular approach.

The task of the Integrated Greenhouse Gas Monitoring System (ITMS) is to determine Germany’s emissions of various greenhouse gases. For methane (CH4), the ITMS demonstrator and an accompanying scientific publication were released in 2025. In this work, methane emissions were derived from ICOS observations, applied to the example year 2021, and examined in depth with respect to methodological limitations. Researchers at the Deutscher Wetterdienst (DWD) simultaneously developed an additional method to solve the so-called inverse problem—determining emissions from observations. This new method avoids the limitation in the spatial distribution. Consequently, different methodological approaches can be compared, and their respective strengths and weaknesses evaluated more thoroughly. The new study presents the method and compares both approaches. It has been published for scientific discussion: https://doi.org/10.5194/egusphere-2026-3221 [1].

The alternative method is based on the data assimilation, more specifically the Local Ensemble Transform Kalman Filter (LETKF). In this process, many slightly different model simulations are compared with the observed data and gradually adjusted. In this way, the algorithm improves both the calculated methane concentrations and the estimated emissions. The measurement data comes from the Integrated Carbon Observatory System (ICOS), which continuously measures the concentrations of greenhouse gases in the atmosphere.

Figure 1: Top: Estimates of methane emissions in Europe for 2021. Bottom: The difference between a priori and a posteriori. Left: The new study and Right: the ITMS-Demonstrator. The geographic distribution of the corrections is similar, particularly within Germany.

Figure 1: Top: Estimates of methane emissions in Europe for 2021. Bottom: The difference between a priori and a posteriori. Left: The new study and Right: the ITMS-Demonstrator. The geographic distribution of the corrections is similar, particularly within Germany.

The results of this study agree with those of the ITMS demonstrator. The inversions suggest that methane emissions in Germany are being underestimated, particularly in the northwestern part of the country. Overall, the study estimates methane emissions to be 33 ± 20% higher than those reported to the UNFCCC for the year 2021. Despite the agreement between the two methods, further uncertainties remain, such as potential systematic errors in the modeled transport by the DWD weather model ICON. The ITMS project is in close contact with its partners at the Federal Environment Agency and the Thünen Institute to explain this discrepancy in the estimates.

With this second, independent method, ITMS is expanding its toolkit for determining greenhouse gas emissions. The consistency between the two approaches strengthens confidence in the results and provides an important foundation for further expanding monitoring to include other greenhouse gases.

[1] Becker, N., Bruch, V., R¨ osch, T., Keil, N. H., Steil, M. J., Jimenez de la Cuesta Otero, D., Ellerhoff, B., Potthast, R., and Kaiser-Weiss, A. K.: Concur- rent assimilation of methane fluxes and concentrations with a 4D LETKF for Germany in 2021 based on ICON–ART, EGUsphere [preprint], https://doi. org/10.5194/egusphere-2026-3221, 2026.