MSGID: 1:317/3 64475775   
   PID: hpt/lnx 1.9.0-cur 2019-01-08   
   TID: hpt/lnx 1.9.0-cur 2019-01-08   
    A more precise model of the Earth's ionosphere    
    With the help of neural networks, the complexity of the layer around the   
   Earth can be reconstructed much better than before. This is important for   
   satellite navigation, among other things.    
      
     Date:   
         April 24, 2023   
     Source:   
         GFZ GeoForschungsZentrum Potsdam, Helmholtz Centre   
     Summary:   
         The ionosphere -- the region of geospace spanning from 60 to 1000   
         kilometers above the Earth -- impairs the propagation of radio   
         signals from global navigation satellite systems (GNSS) with its   
         electrically charged particles. This is a problem for the ever   
         higher precision required by these systems -- both in research   
         and for applications such as autonomous driving or precise orbit   
         determination of satellites.   
      
         Models of the ionosphere and its uneven, dynamic charge distribution   
         can help correct the signals for ionospheric delays, which are   
         one of the main error sources in GNSS applications. Researchers   
         have presented a new model of the ionosphere, developed on the   
         basis of neural networks and satellite measurement data from 19   
         years. In particular, it can reconstruct the topside ionosphere,   
         the upper, electron-rich part of the ionosphere much more   
         precisely than before. It is thus also an important basis for   
         progress in ionospheric research, with applications in studies on   
         the propagation of electromagnetic waves or for the analysis of   
         certain space weather events, for example.   
      
      
         Facebook Twitter Pinterest LinkedIN Email   
      
   ==========================================================================   
   FULL STORY   
   ==========================================================================   
   The ionosphere -- the region of geospace spanning from 60 to 1000   
   kilometres above the Earth -- impairs the propagation of radio signals   
   from global navigation satellite systems (GNSS) with its electrically   
   charged particles.   
      
   This is a problem for the ever higher precision required by these systems   
   - - both in research and for applications such as autonomous driving or   
   precise orbit determination of satellites. Models of the ionosphere and   
   its uneven, dynamic charge distribution can help correct the signals   
   for ionospheric delays, which are one of the main error sources in   
   GNSS applications.   
      
   Researchers led by Artem Smirnov and Yuri Shprits of the GFZ German   
   Research Centre for Geosciences have presented a new model of the   
   ionosphere in the journal Nature Scientific Reports, developed on the   
   basis of neural networks and satellite measurement data from 19 years. In   
   particular, it can reconstruct the topside ionosphere, the upper,   
   electron-rich part of the ionosphere much more precisely than before. It   
   is thus also an important basis for progress in ionospheric research,   
   with applications in studies on the propagation of electromagnetic waves   
   or for the analysis of certain space weather events, for example.   
      
   Background: Importance and complexity of the ionosphere The Earth's   
   ionosphere is the region of the upper atmosphere that extends from   
   about 60 to 1000 kilometres in altitude. Here, charged particles such as   
   electrons and positive ions dominate, caused by the radiation activity of   
   the Sun -- hence the name. The ionosphere is important for many scientific   
   and industrial applications because the charged particles influence   
   the propagation of electromagnetic waves such as radio signals. The   
   so-called ionospheric propagation delay of radio signals is one of the   
   most important sources of interference for satellite navigation. This is   
   proportional to the electron density in the space traversed. Therefore,   
   a good knowledge of the electron density can help in correcting the   
   signals. In particular, the upper region of the ionosphere, above 600   
   kilometres, is of interest, since 80 per cent of the electrons are   
   gathered in this so-called topside ionosphere.   
      
   The problem is that the electron density varies greatly -- depending on   
   the longitude and latitude above the Earth, the time of day and year,   
   and solar activity. This makes it difficult to reconstruct and predict   
   them, the basis for correcting radio signals, for example.   
      
   Previous models There are various approaches to modelling electron density   
   in the ionosphere, among others, the International Reference Ionosphere   
   Model IRI, which has been recognised 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 the important area of the topside ionosphere because of   
   the limited coverage of previously collected observations in that region.   
      
   Recently, however, large amounts of data have become available for   
   this area.   
      
   Therefore, Machine learning (ML) approaches lend themselves to deriving   
   regularities from this, especially for complex non-linear relationships.   
      
   New approach using machine learning and neural networks A team from   
   the GFZ German Research Centre for Geosciences around Artem Smirnov,   
   PhD student and first author of the study, and Yuri Shprits, head of the   
   "Space Physics and Space Weather" section and Professor at University   
   Potsdam, took a new ML-based empirical approach. For this, they used   
   data from satellite missions from 19 years, in particular CHAMP, GRACE   
   and GRACE-FO, which were and are significantly co-operated by the GFZ,   
   and COSMIC. The satellites measured -- among other things -- the electron   
   density in different height ranges of the ionosphere and cover different   
   annual and local times as well as solar cycles.   
      
   With the help of Neural Networks, the researchers then developed a model   
   for the electron density of the topside ionosphere, which they call the   
   NET model.   
      
   They used the so-called MLP method (Multi-Layer Perceptrons), which   
   iteratively learns the network weights to reproduce the data distributions   
   with very high accuracy.   
      
   The researchers tested the model with independent measurements from   
   three other satellite missions.   
      
   Evaluation of the new model "Our model is in remarkable agreement with   
   the measurements: It can reconstruct the electron density very well in   
   all height ranges of the topside ionosphere, all around the Globe, at all   
   times of the year and day, and at different levels of solar activity,   
   and it significantly exceeds the International Reference Ionosphere   
   Model IRI in accuracy. Moreover, it covers space continuously," first   
   author Artem Smirnov sums up.   
      
   Yuri Shprits adds: "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   
   numerous physical processes that govern the dynamics of the topside   
   ionosphere and can have broad applications in ionospheric research."   
   Possible applications in ionosphere research The researchers see possible   
   applications, for instance, in wave propagation studies, for calibrating   
   new electron density data sets with often unknown baseline offsets,   
   for tomographic reconstructions in the form of a background model,   
   as well as to analyse specific space weather events and perform long-   
   term ionospheric reconstructions. Furthermore, the developed model can   
   be connected to plasmaspheric altitudes and thus can become a novel   
   topside option for the IRI.   
      
   The developed framework allows the seamless incorporation of new data   
   and new data sources. The retraining of the model can be done on a   
   standard PC 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 communication and   
   navigation systems that rely on GNSS.   
      
       * RELATED_TOPICS   
             o Earth_&_Climate   
                   # Atmosphere # Weather # Geomagnetic_Storms #   
                   Severe_Weather # Earth_Science # Climate #   
                   Environmental_Issues # Geography   
       * RELATED_TERMS   
             o Ionosphere o Global_Positioning_System o Solar_cell o   
             Earth_science o Global_climate_model o Weather_forecasting o   
             Meteorology o Solar_power   
      
   ==========================================================================   
   Story Source: Materials provided by   
   GFZ_GeoForschungsZentrum_Potsdam,_Helmholtz_Centre. Note: Content may   
   be edited for style and length.   
      
      
   ==========================================================================   
   Journal Reference:   
      1. Artem Smirnov, Yuri Shprits, Fabricio Prol, Hermann Lu"hr, Max   
         Berrendorf, Irina Zhelavskaya, Chao Xiong. A novel neural network   
         model of Earth's topside ionosphere. Scientific Reports, 2023; 13   
         (1) DOI: 10.1038/s41598-023-28034-z   
   ==========================================================================   
      
   Link to news story:   
   https://www.sciencedaily.com/releases/2023/04/230424103340.htm   
      
   --- up 1 year, 8 weeks, 10 hours, 50 minutes   
    * Origin: -=> Castle Rock BBS <=- Now Husky HPT Powered! (1:317/3)   
   SEEN-BY: 15/0 106/201 114/705 123/120 153/7715 218/700 226/30 227/114   
   SEEN-BY: 229/110 112 113 307 317 400 426 428 470 664 700 292/854 298/25   
   SEEN-BY: 305/3 317/3 320/219 396/45   
   PATH: 317/3 229/426