Project Details
Projekt
Modeling and forecasting of the electron density distribution within the ionosphere
Subject Area
Atmospheric Science
Term
since 2024
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 462853228
From the geodetic perspective, the electron density plays a key role in applications such as satellite navigation and point positioning, since the quality of ionospheric correction models depends strongly on the knowledge of the electron density. Besides the electron density, the Slant Total Electron Content (STEC), i.e., the integration of the electron density along the signal propagation path between transmitter and receiver, and the Vertical Total Electron Content (VTEC), i.e., the mapping of STEC into the vertical, are referred to as the three key quantities of the ionosphere. Since STEC and VTEC describe only the integrative effect, local or regional irregularities of the electron density, such as those caused by space weather events, can be effectively detected only by investigating the electron density itself. The monitoring of electron density is challenging in two respects: on the one hand there is a relatively small number of direct observations because only a few measurement techniques exist that provide values for electron density itself, and on the other hand the electron density is highly variable in space and time, strongly effected by solar and geomagnetic activity as well as coupling processes between the atmosphere layers. New approaches for modeling and forecasting the electron density values will be developed and tested within this project. It addresses two main questions, namely (1) what conclusions can be drawn about the influence of solar and geomagnetic activity on the electron density distribution in the ionosphere, and (2) how can we use Machine Learning (ML) techniques to quantify relationships between the ionospheric key parameters and the solar and geomagnetic parameters for an improved modelling and forecasting of the electron density? To answer these questions, we first use a four-dimensional (4-D) ionosphere-plasmasphere model that describes the electron density vertically through a multi-layer Chapman model with 3 key parameters (peak value, peak height, scale height) for each of the 4 ionosphere layers. The dependency of the key parameters on latitude and longitude is realized by 2-D series expansions. Thus, the developed model transforms all measurements including ground and space-based GNSS, DORIS, radio occultations, etc., into the series coefficients of selected key parameters using an optimization approach. This procedure allows not only the generation of 2-D maps of selected key parameters and 3-D electron density grids, but also the computation of correlations with solar and geomagnetic data, as well as feature importance diagrams using ML techniques. Based on these feature priority lists describing, e.g., the influence of solar and geomagnetic parameters as well as coupling processes, the electron density grids can be forecasted up to 2 days using ML methods such as ensemble learning which allows to quantify the accuracy of the results by the computation of confidence intervals.
DFG Programme
Research Units
