New developments of the CHAOS ionospheric field model

The CHAOS-8 geomagnetic field model series describes the time-dependent near-Earth geomagnetic field under quiet conditions since 1999. It is derived from magnetic field observations from low-Earth orbit satellites, such as Swarm, CHAMP, MSS-1, and CSES, as well as annual differences of revised monthly means of ground observatory measurements. Starting with the 8th generation, the series co-estimates a climatological model of the ionospheric E-layer currents with a focus on accounting for their magnetic signals in the polar regions, which can be significant even under quiet and dark conditions. This model follows the AMPS approach (Laundal et al., 2018), utilizing magnetic apex coordinates and magnetic local time to describe large-scale patterns efficiently. Additionally, it uses multiple external parameters, including the Interplanetary Magnetic Field, dipole tilt angle, and magnetosphere-ionosphere coupling functions, to represent variability on seasonal, daily, and shorter time scales.Although the CHAOS ionospheric field model can successfully represent the average patterns in the polar ionospheric E-layer field, limitations remain. Most notably, it is less suitable at non-polar latitudes, where the Sq current system dominates, because it lacks longitude dependence. Moreover, the reliance on simple dependencies on external parameters to capture temporal variability may be overly restrictive, particularly for seasonal and long-term changes. Finally, since the CHAOS ionospheric field is estimated only from satellite data, both the internal and ionospheric contributions are treated as internal sources.This work presents ongoing efforts to address limitations in the CHAOS ionospheric field. Test models are estimated from satellite data using monthly and shorter time windows to capture seasonal variability better. Longitudinal dependence is introduced to provide a more accurate representation of the field at low latitudes, following the approach of the comprehensive model (Sabaka et al. 2003), while continuing to rely on apex coordinate systems. By comparing model predictions to ground observatory data, the potential of incorporating observatory measurements into the model estimation is explored to enhance the separation of internal and ionospheric contributions.