Electron density modelling at Swarm height using Neural Networks for space weather monitoring

Abstract

The Swarm data base is well-suited to address a number of topics of serious interest in space weather science and monitoring as for instance: spatial and temporal characteristics of ionospheric electron density, improving topside approaches in ionospheric models for monitoring and forecasting the dynamics of the geo-plasma environment. In this study, we developed a neural network-based electron density model using the electron density measured by Langmuir probes on the Swarm A and C satellites. Data from the years 2014 till 2021 has been used for this study, where the satellites have an approximate altitude range of 470-430 km. The model’s capability of showing large and small-scale features of the ionosphere was tested and the results show that the model is capable of showing the crest formations on both sides of the magnetic equator, as well as seasonal and diurnal variations. Furthermore, using the neural network-based model predictions, the nighttime winter anomaly (NWA) feature was investigated. The NWA is a small-scale feature that can be observed during low solar activity conditions at nighttime in the Northern Hemisphere at the American sector and in the Southern Hemisphere at the Asian sector. Such electron density models at specific height region can be used for three-dimensional ionosphere model validation as well as for the development of improved ionosphere models. Again, accurate modelling and monitoring of ionospheric electron density at certain height region can help prediction of space weather impact.