Development of ionospheric tomography using neural network and its application to the 2007 Southern Sumatra earthquake

Abstract

An ionospheric anomaly prior to the 2007 Southern Sumatra earthquake (M8.5) was observed by GPS receivers around the Sumatra islands. In this paper, to investigate the three-dimensional structure of electron density in the ionosphere, a tomographic approach (Residual Minimization Training Neural Network; RMTNN) has been used. Results of the tomographic approach are consistent with those of the total electron content (TEC) approach. We found that a significant decrease of electron density takes place at altitudes of 250 to 400 km, especially at an altitude of 330 km. But the altitude at which the maximum electron density has been observed remains unchanged. The obtained structure has a region of decreased density of Integrated Electron Content (IEC) in the southwest area (at altitudes of 400 to 550 km) and in the northern area (at altitudes of 250 to 400 km). The global tendency of the decreased electron density expands to the east with higher altitudes and it is concentrated in the southern hemisphere over the epicenter. These results show that the high capability of the RMTNN method can be used for estimation of the ionospheric electron density distribution possibly associated with earthquake.

Ionospheric F2 region perturbed by the 25 April 2015 Nepal earthquake

“…Various types of numerical approaches, such as data assimilation and tomography, have been used to reconstruct 3-D ionospheric Ne structures from ground-based GPS TEC and space-based RO observations [Hirooka et al, 2012;Lin et al, 2015]. Here the observations from F3/C (Figure 3) and ground-based GPS receivers (Figure 1) show that the earthquake-generated Ne perturbation in the vertical direction (~10%) is much more significant than the TEC perturbation in the horizontal direction (~1%).…”

Validation of Ionospheric Tomography Using Residual Minimization Training Neural Network

“…As regards the application of RMTNN to ionosphere studies, there are reports of its use in numerical experiments under specific conditions, and in the visualization of ionospheric anomalies preceding earthquakes. The effectiveness of the method has been confirmed by comparison with other data . However, on the practical side, the noise characteristics (tolerance to possible measurement errors) and detection ability (peak and spread of recognizable structures) of RMTNN‐based ionospheric tomography have not been sufficiently verified.…”

Online user‐friendly slant total electron content computation from IRI‐Plas: IRI‐Plas‐STEC

“…The widely used Global Positioning System (GPS) provides a cost‐effective means for computation of STEC (GPS‐STEC) using pseudorange and phase delay recordings of dual‐frequency Earth‐based receivers [ Komjathy , ; Schaer , ; Davies and Hartmann , ; Arikan et al , ]. GPS‐STEC is used both in correction of positioning errors due to ionospheric delays and also in ionospheric physics to capture the underlying structure of ionosphere using computerized ionospheric tomography such as in Erturk et al [], Sibanda [], Jin and Jin [], and Hirooka et al []. The GPS‐STEC measurement model includes instrumental biases that need to be determined to increase the positioning resolution and reduce nonionospheric components from STEC.…”