Enhancing the Accuracy of the Earth’s Ionosphere Model

The ionosphere – the region of Earth’s space extending from 60 to 1,000 km above the Earth – impedes the propagation of radio signals from Global Navigation Satellite Systems (GNSS) with its electrically charged particles. This is an issue with the high precision that these systems require—both in research and for applications such as autonomous driving or precise satellite orbit determination.

Models of the ionosphere and the uneven dynamic charge distribution can help correct for delayed signals in the ionosphere, which is a major source of error in GNSS applications. Researchers led by Artem Smirnov and Yuri Shprits of the German GFZ Research Center for Geosciences presented a new model of the ionosphere in the journal. Scientific reportsdeveloped on the basis of neural networks and satellite measurement data from 19 years ago.

In particular, it can reconstruct the top of the ionosphere, the electron-rich upper part of the ionosphere with much greater accuracy than before. Hence it is also an important basis for progress in ionospheric research, with applications in studies of the propagation of electromagnetic waves or for the analysis of some space weather phenomena, for example.

Background: The importance and complexity of the ionosphere

The Earth’s ionosphere is the region of the upper atmosphere that extends from about 60 to 1,000 kilometers in altitude. Here, charged particles such as electrons and positive ions dominate, due to the sun’s radioactivity – hence the name. The ionosphere is important for many scientific and industrial applications because charged particles affect the propagation of electromagnetic waves such as radio signals.

The ion propagation delay of radio signals is one of the most important sources of interference in satellite navigation. This is proportional to the electron density in the space being traversed. Therefore, a good knowledge of the electron density can help correct the signals. In particular, the upper region of the ionosphere, above 600 km, is of interest, since 80 percent of the electrons gather in the so-called upper ionosphere.

The problem is that electron density varies greatly – depending on the latitude and longitude over Earth, the time of day and year, and solar activity. This makes them difficult to reconstruct and predict, as a basis for correcting radio signals, for example.

previous models

There are various methods for modeling the electron density in the ionosphere, among others, the International Reference Ionosphere Model IRI, which has been recognized since 2014. It is an empirical model that establishes a relationship between input and output variables based on the statistical analysis of observations. However, it still has weaknesses in an important region of the upper ionosphere due to the limited coverage of observations previously collected in that region.

Recently, however, large amounts of data have become available for this region. Therefore, machine learning (ML) approaches lend themselves to deriving regularity from this, especially for complex nonlinear relationships.

A new approach using machine learning and neural networks

A team from the GFZ German Research Center for Geosciences is about Artem Smirnov, Ph.D. Student and first author of the study, Juri Shpritz, Head of the Department of “Space Physics and Space Weather” and Professor at the University of Potsdam, has carried out a new experimental approach based on ML.

For this purpose, they used data from 19 satellite missions, in particular CHAMP, GRACE and GRACE-FO, which were and still are largely operated by the GFZ and COSMIC. The satellites measured – among other things – electron density at different altitude ranges of the ionosphere and covering different annual and local times as well as solar cycles.

With the help of Neural Networks, the researchers have developed a model of electron density in the upper ionosphere, which they call the NET model. They used a so-called MLP (Multi-Layer Perceptrons) method, which iteratively learns network weights to reproduce data distributions with very high accuracy.

The researchers tested the model with independent measurements from three other satellite missions.

Evaluate the new model

“Our model is remarkably consistent with the measurements: it can reconstruct electron density very well at all upper ionosphere altitude ranges, worldwide, at all times of the year and day, and at different levels of solar activity, and it significantly exceeds the International Reference Model for the IRI.” In terms of accuracy. Moreover, it covers space constantly, ”summarizes first author Artem Smirnov.

Adds Yuri Shpritz: “This study represents a paradigm shift in ionospheric research because it shows that ionospheric densities can be reconstructed with very high accuracy. The NET model reproduces the effects of several physical processes that govern the dynamics of the upper ionosphere and can have wide applications in Ionospheric Research”.

Possible applications in ionospheric research

The researchers see possible applications, for example, in studies of wave propagation, for calibrating new electron density datasets with often unknown baseline offsets, for tomographic reconstructions in the form of a background model, as well as for analyzing specific space weather events and performing long-distance reconstructions. ionosphere. Moreover, the developed model can be correlated with plasma spikes and thus can become a new upstream option for IRI.

The developed framework allows for seamless integration of new data and new data sources. The model can be retrained on a standard computer and can be performed on a regular basis. Overall, the NET model represents a significant improvement over traditional methods and highlights the potential of neural network-based models to provide a more accurate representation of the ionosphere for GNSS-based communication and navigation systems.