The magnetic core field and its magnetosphere show relevant changes on time scales from hours (magnetic storms) to millions of years (polarity reversals). Geomagnetic observations contain contributions from: the main field produced by geodynamo processes in Earth's outer core; the lithospheric field from magnetized rocks and geological structures; and highly variable fields caused by current systems in ionosphere and magnetosphere with secondary fields from induction in conductive structures in the Earth's crust and mantle. Separation of the different contributions is essential to understand the underlying processes with implications for future main field evolution, space weather conditions or interpretation of lithospheric anomalies. This separation remains a challenge. Modern vector magnetic field satellite data lead to increasingly accurate geomagnetic main-field models that co-estimate large-scale magnetospheric fields. When using geomagnetic observatory time series for internal field secular variation studies external field variations and their induced counterparts are traditionally filtered out by using monthly or annual mean values. Comparisons of the modern models to such data confirm that multi-annual to decadal magnetospheric variations are present in these time series, which are essential to investigate internal secular variation and characteristics of main field evolution on a wider frequency range than available from satellite data. Geomagnetic activity indices like the Dst (disturbed storm time) index were developed to describe magnetospheric field variations. However, the Dst index by definition does neither describe the absolute value nor the long-term magnetic variations correctly.A number of improvements to Dst have been suggested, but none of them has specifically addressed the decadal stability and absolute level. This project aims to derive a method for a better separation of the multi-annual to decadal magnetospheric field variations and internal secular variation from geomagnetic observatory and repeat station time series. We will take advantage of the global data compilations of ground data available from the World Data Center Edinburgh and the results from separating internal and magnetospheric fields in the modern models based on data from the satellites CHAMP, Ørsted and Swarm (in particular including new findings obtained within this SPP) to develop a new magnetospheric and induced counterpart index using ground data, that can be extended back to the early 20th century. The benefits will be twofold: Firstly, an improved understanding of long-term magnetospheric variations is important, e.g., regarding space weather implications. Secondly, cleaner internal field ground data will enable more detailed studies of decadal secular variation and core field characteristics to better understand core dynamics and geodynamo processes and will also facilitate the use of repeat station data to augment total field lithospheric anomaly mappings with vector field information.