Temporal and Spatial Variability of Multiscale Neutral Density Perturbations During Storm-Time: Insights From Multi-Satellite Observations

Abstract

Abstract

The neutral density perturbations exhibit multiscale features during geomagnetic storms, playing a crucial role in ionosphere-thermosphere (I-T) dynamics. However, the variations across various temporal and spatial scales remain underexplored. This study compared Gravity Recovery and Climate Experiment (GRACE) satellite data with simulations of Global Ionosphere-Thermosphere Model (GITM) during the 2015 St. Patrick's Day storm. In general, GITM captures large- and mesoscale density structures well, with some underestimations of mesoscale variations due to inaccuracy in the geomagnetic forcing. GRACE-A/B distinguish between temporal and spatial variations, revealing enhanced mesoscale structures at high latitudes over 35 s (temporal) and 220 km (spatial) scales. Virtual satellite results show that three latitudinally spaced satellites efficiently capture the longitudinal propagation of the large-scale traveling atmospheric disturbances (LSTADs), while six longitudinally spaced satellites (30°apart) significantly improve polar map accuracy. The logarithmic string-of-pearl configuration in the latitudinal plane adequately extracts neutral density variations across different temporal and spatial scales. Temporal variations, \({\Delta }\rho\), increase with the time scale with a threshold of 3 min for clear detection, while the change rate, \({\Delta}\rho /{\Delta}t\), decreases with time. Spatial variation in magnitude increases with spatial scale with a threshold of 3.5° or 418 km to generate clear variation (>10%). However, the neutral density gradient, \({\Delta }\rho /{\Delta }\mathrm{km}\), remains roughly the same on different spatial scales. Coherence analysis was applied to assess the relationship between satellite distance and observed scales, highlighting the importance of multi-satellite observations in understanding of multiscale thermosphere dynamics.