Filtering/Smoothing Coupled Technique Implementation on the Data Assimilation Algorithm

Due to the limitations of models and measurements especially during geomagnetic storm time, the data assimilation technique is widely used for reducing the information gap and estimating the optimized ionospheric drivers [1]. Estimating Model Parameters Ionospheric Reverse Engineering (EMPIRE), one of the data assimilation algorithms is a physics-based algorithm led by the ion continuity equation [2]. It’s a Variational method that utilizes the Kalman filtering technique and explicitly calculates the ion drifts and neutral winds effects by primarily ingesting the modeled global maps of electron density. The recent developments on EMPIRE, which take direct measurements of neutral winds [3] and ion velocities [4], has been validated that the algorithm can provide a more comprehensive understanding of ionospheric drivers. In a preliminary study of augmenting superDARN ion drift measurements and validating with the MillstoneHill dataset on March 17th, 2015, the mean error of ion drifts estimations in meridional direction is reduced by approximately 50% compared to the EMPIRE run of only ingesting electron density.

Beyond the numerical validation with Millstone Hill measurements, we explored the scientific principles of a storm-related phenomenon, nighttime ionospheric localized enhancement (NILE) by assessing the ion drift calculations at different storm phases. NILE is one of the phenomena being observed through the study of the October 2003 extreme storm is that of the enhanced ionospheric plasma, which corotates persistently with earth during the late-night period (18 LT - 24 LT), initially named the “Florida effect.” It was observed through the data collected from over 400 ground-based multi-frequencies GNSS receivers. Studies on August 2018 and November 2003 storms [5], based on the implementation of data assimilation tool IDA4D [6] coupled with ionosphere model SAMI3 [7], showed the evolution of NILE in terms of occurrence time and region along with GPS and ionosonde validations [8]. The proposed driving mechanism is that plasma is uplifted due to the pre-reversal enhancement effect, then drifts downwards after the local dusk terminator. Based on the polarization terminator definition, the plasma drifts westward and poleward to form the density enhancement in the Florida region during local nighttime [8]. To validate the driving mechanism with the ion drift augmentation method, the evaluation period was extended from one day on March 17th to three days from March 16th to 18th. The ion estimation results partially supported the proposed theory that the ions are uplifted before the dusk terminator, then drift downward and poleward, but the zonal drift direction is opposite to the theory. As indicated by the augmentation results, the ion drift estimation in the zonal direction was worse than the calculations from both of climate model and primary EMPIRE setup.

In this study, we are updating the Kalman filtering technique to the filtering/smoothing coupled technique to test whether it can improve the analysis of ionospheric driver estimations. The filtering/smoothing technique analyzes the ionospheric drivers recursively in a time forward and backward sequence. Since all measurement and model information is available to access for a past storm event, the EMPIRE can update the state vectors (i.e. ion drifts and neutral winds in EMPIRE), using the past and “future” data to yield the optimized estimations with the loop closure process. To be consistent with the preliminary study of the augmentation method, we select the analysis time to be March 16 to 18th 2015, ingest the superDARN ion drifts measurements, and validate the coupled technique with the Millstone Hille dataset by comparing with the Kalman filtering outputs. If it is proved that the coupled technique can improve the ion drift calculations, we will investigate the ion motion in the Florida region to testify to the proposed NILE driving mechanism.

References
[1] Schunk, R. W., Scherliess, L., & Thompson, D. C. (2011). Ionosphere data assimilation: Problems associated with missing physics. Aeronomy of the Earth's Atmosphere and Ionosphere, 437-442.
[2] Datta‐Barua, S., Bust, G. S., Crowley, G., & Curtis, N. (2009). Neutral wind estimation from 4‐D ionospheric electron density images. Journal of Geophysical Research: Space Physics, 114(A6).
[3] Miladinovich, D. S., Datta-Barua, S., Bust, G. S., & Makela, J. J. (2016). Assimilation of thermospheric measurements for ionosphere-thermosphere state estimation. Radio Science, 51(12), 1818-1837.
[4] Hu, J., Rubio, A. L., Chartier, A., Bust, G. S., & Datta-Barua, S. (2021, September). Data Assimilation of Ion Drift Measurements for Estimation of Ionospheric Plasma Drivers. In Proceedings of the 34th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2021) (pp. 3833-3847).
[5] Datta-Barua, S., Mannucci, A. J., Walter, T., & Enge, P. (2008). Altitudinal variation of midlatitude localized TEC enhancement from ground-and space-based measurements. Space Weather, 6(10), S10D06.
[6] Bust, G. S., Garner, T. W., & Gaussiran, T. L. (2004). Ionospheric Data Assimilation Three-Dimensional (IDA3D): A global, multisensor, electron density specification algorithm. Journal of Geophysical Research: Space Physics, 109(A11).
[7] Huba, J. D., Joyce, G., & Fedder, J. A. (2000). Sami2 is Another Model of the Ionosphere (SAMI2): A new low-latitude ionosphere model. Journal of Geophysical Research: Space Physics, 105(A10), 23035-23053.
[8] Chartier, A. T., Datta‐Barua, S., McDonald, S. E., Bust, G. S., Tate, J., Goncharenko, L. P., ... & Schaefer, R. K. (2021). Night‐Time Ionospheric Localized Enhancements (NILE) Observed in North America Following Geomagnetic Disturbances. Journal of Geophysical Research: Space Physics, 126(9), e2021JA029324.