The accuracy of data from ionosondes for the estimation of hmF2 and the identification of global change in the ionosphere

A long temporal series of simulated ionograms was generated with a superimposed secular variation of −14 km/century on the h_mF2 parameter. These ionograms were interpreted by the automatic scaling program Autoscala. By applying four different empirical formulas, four artificial series of h_mF2 were generated and then processed with the same methods used by other authors for real data sets. Data analysis of the simulated ionograms revealed the artificially imposed long-term trend. These results lead to the conclusion, that regardless of the empirical formula used, the accuracy of h_mF2 from ionosonde measurements would be adequate to observe a long-term trend of −14 km/century.     

Improvements in bottomside electron density definition in the Autoscala program

Some improvements introduced in the Autoscala program are presented. They include improvements in E valley modeling of the electron density profile N_e(h), and in the link between the E valley and bottom-side F regions. An abrupt variation in N_e(h) generated by the previous version of Autoscala under night conditions has been eliminated.A series of ionograms recorded by the Millstone Hill digisonde (42.6°, 288.5°) were automatically interpreted by the previous version of Autoscala and by the new one. Data from Incoherent Scatter Radar (ISR) were used to comparatively assess the performance of the two versions. For this purpose, the root mean square errors (RMSEs) of the N_e(h) provided by Autoscala were calculated relative to the corresponding values provided by ISR.A more accurate overall modeling of N_e(h) was achieved by the new Autoscala version (RMSE = 0.51 MHz for the new version against RMSE = 0.67 MHz for the previous one).     

A comparative study of decision tree,random forest,and convolutional neural network for spread-F identification

Ionospheric spread-F (SF) is a commonly observed phenomenon of electron density perturbation in the F-layer. The ionospheric irregularities structure has an adverse effect on the propagation of electromagnetic waves in the ionosphere. The automatic identification of ionospheric spread-F and statistical study of the formation of spread-F are of great significance to the study of the physical mechanism of ionospheric inhomogeneity and for prediction of ionospheric irregularities. In this paper, we describe and implement three automatic identification methods of spread-F based on machine learning: decision tree, random forest, and convolutional neural network (CNN). The performance of these automatic identification methods was verified using a large set of test data. Results show that the accuracy of all three methods on identifying ionograms with spread-F exceeded 90%. After comparing the results of the three methods, we found that the decision tree method was the simplest and with the structure easiest to be understood, and it required the shortest interpretation time. In terms of the identification results, the random forest method provided better results than the decision tree method, and the CNN method was the best at accurately identifying ionograms with spread-F.     

A method for automatic detection of equatorial spread-F in ionograms

A method is presented for automatic detection of spread-F. The method is based on an image recognition technique and is applied to ionograms recorded at the ionospheric station of Tucumán (26.9°S, 294.6°E). The performance achieved is statistically evaluated and demonstrated with significant examples. The proposed method improves Autoscala's ability to reject ionograms with insufficient information, including those featuring Spread-F. Automatic identification of cases of spread-F is of additional interest in Space Weather applications, when it helps detect degraded radio propagation conditions.The present data analysis is a retrospective study but forms the basis for real-time application as an extension of Autoscala’s capabilities.     

A statistical study of the F2 layer stratification at the northern equatorial ionization anomaly

Ionograms recorded at Puer station (PUR, 22.7°N, 101.05°E, Dip Latitude 12.9°N) in the Southwest of China from January 2015 to December 2016 were used to study characteristics of the F2 layer stratification at the northern equatorial ionization anomaly. Ionosonde observations show that the development of the F2 layer stratification is different under different conditions. Both the upward and downward movement of the F2 layer stratification could be observed. The F2 layer stratification could occur both at daytime and nighttime. The new cusp could originate from different positions on ionograms. Moreover, statistical results indicate that the F2 layer stratification occurred later in the winter than in other seasons at daytime, it occurred frequently in the local spring, and most of ionograms with the F2 layer stratification at post-midnight occurred in March and April. Our results also show that the F2 layer stratification has a correlation with solar activity.