Application of convolution neural networks for critical frequency fₒF2 prediction
Ionosphere has an important impact on the quality of radio communication, radar, and global positioning. One of the essential characteristics describing the state of the ionosphere is its critical frequency fₒF2. Its prediction provides effective modes of operation of technical radio equipment as we...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
INFRA-M
2023-03-01
|
| Series: | Solar-Terrestrial Physics |
| Subjects: | |
| Online Access: | https://naukaru.ru/en/nauka/article/51424/view |
| _version_ | 1827975524921638912 |
|---|---|
| author | Salimov B. G. Berngardt O. I. Hmelnov A. E. Ratovsky K. G. Kusonsky O. A. |
| author_facet | Salimov B. G. Berngardt O. I. Hmelnov A. E. Ratovsky K. G. Kusonsky O. A. |
| author_sort | Salimov B. G. |
| collection | DOAJ |
| description | Ionosphere has an important impact on the quality of radio communication, radar, and global positioning. One of the essential characteristics describing the state of the ionosphere is its critical frequency fₒF2. Its prediction provides effective modes of operation of technical radio equipment as well as enables calculation of the corrections needed to improve the accuracy of its functioning. Different physical and empirical models are generally used for fₒF2 prediction. This paper proposes an empirical prediction technique based on machine learning methods and observational history. It relies on a regression approach to the prediction based on the known daily quasi-periodicity of ionospheric parameters related to solar illumination. Algorithmically, this approach is implemented in the form of convolutional neural networks with two-dimensional convolution. The input data for the analysis is presented as two-dimensional solar time — date matrices. The model was trained on data from the mid-latitude ionosonde in Irkutsk (RF) and tested using data from several mid-latitude ionosondes: Arti (RF), Warsaw (Poland), Mohe (China). It is shown that the main contribution to the prediction value of fₒF2 is made by the data on the nearest few days before the prediction; the contribution of the remaining days strongly decreases. This model has the following forecast quality metrics (Pearson correlation coefficient 0.928, root mean square error 0.598 MHz, mean absolute error in percent 10.45 %, coefficient of determination 0.861) and can be applied to fₒF2 forecast in middle latitudes. |
| first_indexed | 2024-04-09T20:13:05Z |
| format | Article |
| id | doaj.art-598e526c1f064463b5c7907c74845f31 |
| institution | Directory Open Access Journal |
| issn | 2500-0535 |
| language | English |
| last_indexed | 2024-04-09T20:13:05Z |
| publishDate | 2023-03-01 |
| publisher | INFRA-M |
| record_format | Article |
| series | Solar-Terrestrial Physics |
| spelling | doaj.art-598e526c1f064463b5c7907c74845f312023-03-31T14:44:05ZengINFRA-MSolar-Terrestrial Physics2500-05352023-03-0191566710.12737/stp-91202307 Application of convolution neural networks for critical frequency fₒF2 predictionSalimov B. G.0 Berngardt O. I.1Hmelnov A. E.2Ratovsky K. G.3Kusonsky O. A.4Institute of Solar-Terrestrial Physics SB RAS, Irkutsk, RussiaInstitute of Solar-Terrestrial Physics SB RAS, Irkutsk, RussiaMatrosov Institute for System Dynamics and Control Theory SB RAS, Irkutsk, RussiaInstitute of Solar-Terrestrial Physics SB RAS, Irkutsk, RussiaInstitute of Geophysics, Ural Branch RAS, Ekaterinburg, RussiaIonosphere has an important impact on the quality of radio communication, radar, and global positioning. One of the essential characteristics describing the state of the ionosphere is its critical frequency fₒF2. Its prediction provides effective modes of operation of technical radio equipment as well as enables calculation of the corrections needed to improve the accuracy of its functioning. Different physical and empirical models are generally used for fₒF2 prediction. This paper proposes an empirical prediction technique based on machine learning methods and observational history. It relies on a regression approach to the prediction based on the known daily quasi-periodicity of ionospheric parameters related to solar illumination. Algorithmically, this approach is implemented in the form of convolutional neural networks with two-dimensional convolution. The input data for the analysis is presented as two-dimensional solar time — date matrices. The model was trained on data from the mid-latitude ionosonde in Irkutsk (RF) and tested using data from several mid-latitude ionosondes: Arti (RF), Warsaw (Poland), Mohe (China). It is shown that the main contribution to the prediction value of fₒF2 is made by the data on the nearest few days before the prediction; the contribution of the remaining days strongly decreases. This model has the following forecast quality metrics (Pearson correlation coefficient 0.928, root mean square error 0.598 MHz, mean absolute error in percent 10.45 %, coefficient of determination 0.861) and can be applied to fₒF2 forecast in middle latitudes.https://naukaru.ru/en/nauka/article/51424/viewionospheremachine learningneural networksfₒf2 |
| spellingShingle | Salimov B. G. Berngardt O. I. Hmelnov A. E. Ratovsky K. G. Kusonsky O. A. Application of convolution neural networks for critical frequency fₒF2 prediction Solar-Terrestrial Physics ionosphere machine learning neural networks fₒf2 |
| title | Application of convolution neural networks for critical frequency fₒF2 prediction |
| title_full | Application of convolution neural networks for critical frequency fₒF2 prediction |
| title_fullStr | Application of convolution neural networks for critical frequency fₒF2 prediction |
| title_full_unstemmed | Application of convolution neural networks for critical frequency fₒF2 prediction |
| title_short | Application of convolution neural networks for critical frequency fₒF2 prediction |
| title_sort | application of convolution neural networks for critical frequency fₒf2 prediction |
| topic | ionosphere machine learning neural networks fₒf2 |
| url | https://naukaru.ru/en/nauka/article/51424/view |
| work_keys_str_mv | AT salimovbg applicationofconvolutionneuralnetworksforcriticalfrequencyfof2prediction AT berngardtoi applicationofconvolutionneuralnetworksforcriticalfrequencyfof2prediction AT hmelnovae applicationofconvolutionneuralnetworksforcriticalfrequencyfof2prediction AT ratovskykg applicationofconvolutionneuralnetworksforcriticalfrequencyfof2prediction AT kusonskyoa applicationofconvolutionneuralnetworksforcriticalfrequencyfof2prediction |