Revealing Significant Differences in Horizontal Correlation Distances of E-Layer and F-Layer Based on COSMIC-2 GIS Data

Abstract

Building an accurate background covariance matrix is crucial for ionospheric data assimilation. This matrix quantifies the correlations between grid points and governs the transition between data-driven and model-driven regions. Due to the ionosphere's layered structure and the diverse physical mechanisms operating at different altitudes, the correlation distances vary significantly across layers. While previous studies have primarily focused on Total Electron Content data or F-layer peak electron densities, the E-layer has received little attention. To address this gap, we analyzed the Constellation Observing System for Meteorology, Ionosphere, and Climate-2 Global Ionospheric Specification (GIS) (COSMIC-2 GIS) data from 2020 to 2022, conducting the first study on horizontal spatial correlations in the E-layer (120 km). This study also compared the similarities and differences between the horizontal correlations of the E-layer (120 km) and F-layer (300 km). Our analysis examined how correlation distances vary with local time, geomagnetic latitude, and seasons. The results revealed significant differences in spatial coherence between the two layers: the F-layer exhibits a single-peak structure, with zonal correlation distances greater than meridional correlation distances, while the E-layer shows a more complex double-peak structure, with pronounced peaks at sunrise and sunset. Finally, we developed a horizontal correlation coefficient model based on dual-Gaussian functions at different altitudes. These findings enhance our understanding of ionospheric dynamics and support the construction of more accurate background error covariance matrices, ultimately improving ionospheric modeling and space weather forecasting accuracy.