Residual Ionospheric Error Correction in GNSS Radio Occultation Bending Angles: Parametric Analysis using Electron Density Profiles Derived from COSMIC-II Data

Neutral atmospheric bending angles derived from GNSS Radio Occultation (GNSS-RO) data are essential for estimating atmospheric properties such as temperature, humidity, and pressure. The region of interest for atmospheric properties extends up to 80 km, where ionospheric effects remain and require ionospheric corrections for accurate RO bending angle retrievals. First-order ionospheric terms are typically removed using a linear combination of L1 and L2 bending angles. However, this approach leaves behind higher-order terms, known as residual ionospheric errors (RIEs), which introduce systematic biases into the RO data.

Healy and Culverwell (2015) demonstrated that RIEs are theoretically proportional to the square of the difference between L1 and L2 bending angles, scaled by a coefficient, kappa, which varies with ionospheric conditions. Kappa correction is a convenient method to estimate RIEs directly from bending angle data without relying on external ionospheric data such as electron density profiles. Angling et al. (2018) proposed a simple linear model to estimate kappa as a function of altitude, F10.7, and solar zenith angle. They used the NeQuick model to generate electron density profiles and derived the linear model for kappa estimation. However, since NeQuick is a monthly median ionospheric electron density model, it has limitations in representing real-world ionospheric variability, leading to discrepancies between the kappa values from the NeQuick-based model and those estimated from actual data. Therefore, a more realistic derivation of kappa using actual RO data is needed to develop an improved kappa model.

This study aims to enhance kappa correction by using real electron density profiles derived from GNSS-RO data. A double Chapman layer is fitted to electron density profiles from COSMIC-II data, incorporating the characteristics of the E and F layers to provide continuous representations of the real electron density profiles. Ray-tracing simulations are conducted to obtain L1 and L2 ionospheric bending angles, which are then used to derive kappa values. These kappa values are analyzed under various ionospheric conditions, characterized by user-end parameters such as F10.7, local time, geomagnetic latitude, and altitude.

To examine more accurately the numerical relationship between kappa and these parameters, kappa data is classified by F10.7 to represent different solar activity conditions (e.g., solar minimum and maximum), and is also divided by local time (e.g., noon, midnight, and transition periods). Kappa values for each class are then fitted to the remaining parameters. The findings suggest that kappa values from the model proposed by Angling et al. (2018) differ from those estimated using observational data in this study. By directly deriving kappa values from real data and applying separate fits for different classes of solar activity and local time periods, the modeling accuracy can be enhanced. This study shows the necessity of tailored kappa corrections for different ionospheric conditions, improving techniques for correction of RIEs in GNSS-RO data. 

References

Healy,S.B., & Culverwell,I.D. (2015). A modification to the standard ionospheric correction method used in GPS radio occultation. Atmospheric Measurement Techniques, 8(8), 3385–3393.https://doi.org/10.5194/amt-8-3385-2015

Angling,M.J., Elvidge S., & Healy,S.B. (2018). Improved model for correcting the ionospheric impact on bending angle in radio occultation measurements. Atmospheric Measurement Techniques, 11(4), 2213–2224.https://doi.org/10.5194/amt-11-2213-2018