The Impact of Perturbing Eddy Diffusion and Upper Boundary on the Ionosphere EnKF Assimilation System

• Published in: Journal of geophysical research. Space physics 2021-08, Vol.126 (8), p.n/a
• Main Authors: He, Jianhui, Yue, Xinan
• Format: Article
• Language: English
• Subjects: Atmospheric circulation; Atmospheric research; Data assimilation; Data collection; Diffusion; Diffusion coefficient; Eddy diffusion; Electrodynamics; EnKF data assimilation; Ensemble forecasting; ensemble member generation; General circulation models; Geomagnetic activity; Ionosphere; ionosphere forecast; Kalman filters; Mathematical models; Night; Parameters; Thermosphere; Vortices
• ISSN: 2169-9380; 2169-9402
• DOI: 10.1029/2021JA029366

Accurate representation of ensemble members of coupled Ionosphere‐Thermosphere (I‐T) is crucial to the ionosphere forecast in the ensemble Kalman filter (EnKF) data assimilation system. In this paper, besides the model drivers used in previous studies, including the solar 10.7 cm radio flux (F10.7), auroral hemispheric power (HP) and cross‐tail potential drop (CP), the eddy diffusion coefficient (ED), and nighttime vertical O+ flux at the top boundary are also perturbed to generate ensemble members. Based on our developed EnKF data assimilation system, the impact of perturbing different model external forcing parameters on ensemble generation and forecast capability of ionosphere and thermosphere has been investigated in detail through a series of sensitivity tests and data assimilation experiments. This system uses the National Center for Atmospheric Research (NCAR) Thermosphere Ionosphere Electrodynamics General Circulation Model (TIEGCM) as a background model. The findings are summarized as follows: (a) Associated with perturbing two additional model forcing parameters, better ensemble members of ionosphere and thermosphere states of the background model can be generated during both daytime and nighttime. (b) The improvement of the forecast capability of the ionosphere and thermosphere variables can be further enhanced. This study can provide a reference for ensemble generating strategy for the coupled I‐T EnKF data assimilation in the future. Plain Language Summary The ensemble member of EnKF data assimilation considering as much as possible variabilities existed in the coupled I‐T system should represent model uncertainties better. In the previous researches on EnKF ionosphere forecast, ensemble members are mainly generated by perturbing solar and geomagnetic activity index common model drivers (e.g., F10.7, KP). However, solely perturbing those indices is not sufficient to distribute the ionosphere and thermosphere states at all local times. In this study, in addition to perturbing those indices, the key model forcing parameters of the lower and upper boundary of the TIEGCM are also perturbed to generate the ensemble member of the coupled I‐T system. We found that an improved ensemble spread can be obtained at all local times in this way, which can further contribute to the forecast capability of the ionosphere and thermosphere states. Key Points The eddy diffusion and upper boundary are additionally perturbed to generate ensembles in EnKF I‐T assimi