Planar trend model for the midlatitude ionosphere

Abstract

Midlatitude Ionosphere, which is a passage region between polar and equatorial ionospheres, poses its own peculiarities. Total Electron Content (TEC) is one of the main parameters of the Ionosphere, corresponding to the total number of electrons on a ray path. TEC can be considered as the projection of electron density incorporating the inherent spatiotemporal variabilities of the Ionosphere occurring at various scales. In this study, we have demonstrated that midlatitude TEC can be represented as a planar model whose coefficients can be estimated in closed form in the Least Squares sense using Global Ionospheric Maps (GIM). For 24 regions in the Northern Hemisphere, and their 24 geographic conjugates with respect to the geomagnetic equator in the Southern Hemisphere, the three model coefficients are estimated from January 1, 2000, 00:00 UT to December 31, 2022, 22:00 UT covering about two solar cycles with 2 h temporal resolution. The regions of 10 degrees of latitude and 20 degrees of longitude are arranged around a center coordinate chosen to correspond to a GIM grid point and a time frame in Universal Time (UT). The validity and accuracy of the model coefficients are tested using Square of the Normalized Metric Distance (SNMD) and Symmetric Kullback-Leibler Distance (SKLD). We have also computed the coefficient of determination between the SNMD and SKLD revealing a fit of 99.995 % to the planar model, where deviations from the planar model occur only on severe geomagnetic disturbance days and hours. 77 % of the regions indicate a very strong fit with coefficients of determination higher than 0.95. The remaining 23 % of the regions which are either on or neighboring the geomagnetic field anomalies, the coefficient of determination values are higher than 0.88, indicating a good fit to the planar model. Unlike previous region-specific models proposed in the literature with high computational complexity and problems in duplicability, the whole midlatitude region is covered in our model directly through region, date and time specific coefficients. For all 48 regions and 23 years, the coefficients are made available to the public in a user-friendly fashion and they can be downloaded from https://www.ee.hacettepe.edu.tr/∼solen/_PlanarTrendModel. These coefficients can be used in further interpolation, prediction and tracking routines to reveal the underlying random field structure of midlatitude ionosphere.