Neural network based tomographic approach to detect earthquake-related ionospheric anomalies

Abstract

Abstract.A tomographic approach is used to investigate the fine structure of electron density in the ionosphere. In the present paper, the Residual Minimization Training Neural Network (RMTNN) method is selected as the ionospheric tomography with which to investigate the detailed structure that may be associated with earthquakes. The 2007 Southern Sumatra earthquake (M = 8.5) was selected because significant decreases in the Total Electron Content (TEC) have been confirmed by GPS and global ionosphere map (GIM) analyses. The results of the RMTNN approach are consistent with those of TEC approaches. With respect to the analyzed earthquake, we observed significant decreases at heights of 250-400 km, especially at 330 km. However, the height that yields the maximum electron density does not change. In the obtained structures, the regions of decrease are located on the southwest and southeast sides of the Integrated Electron Content (IEC) (altitudes in the range of 400-550 km) and on the southern side of the IEC (altitudes in the range of 250-400 km). The global tendency is that the decreased region expands to the east with increasing altitude and concentrates in the Southern hemisphere over the epicenter. These results indicate that the RMTNN method is applicable to the estimation of ionospheric electron density.

Atmospheric and ionospheric coupling phenomena related to large earthquakes

“…CC-BY 3.0 License. As before the GIM TEC has been investigated in this study and it shows in Figure that the TEC over the epicenter significantly decreases on September 9, around the noontime period, i.e., 3 days before the earthquake, which well agrees with the result reported by Hirooka et al (2011). The GIM TEC along the 101°E longitude has been extracted during the period August 25 -September 22, 2007.…”

“…There were many aftershocks with magnitude larger than 6 on 13, 14 and 20 September. Hirooka et al (2011) have established that 3 days before the earthquake at 14:00 to 15:00 LT, a strong negative TEC anomaly was detected around the earthquake epicenter. They have also investigated the three-dimensional structure of electron density in the ionosphere, using a tomographic approach.…”

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

“…Estimation accuracy for all altitudes was checked by simulation: at altitudes from 100 to 200 km (especially lower than 200 km) errors were rather high. At other altitudes the accuracy was slightly low at the boundary of the estimation region but even the largest error did not exceed ±1 TECU [18]. Figure 9 shows distributions of electron density in the form of slices made in longitudinal and latitudinal directions at 5° intervals.…”

Application of remote sensing for major land geohazards

“…Z-value) which defines the envelope of normal ionospheric variations. Previous studies suggested Z-values of Karatay et al 2010, Contadakis et al 2015), 1.34 ( Astafyeva and Heki 2011, Hasbi et al 2011, Aggarwal 2015, Nenovski et al 2015), 1.5 (Liu et al 2010, Akhoondzadeh 2013b), 1.96 (Dogan et al 2011), 2 (Molchanov andHayakawa 1998, Hasbi et al 2009, Yiyan et al 2009Jhuang et al 2010, Fuying et al 2011, Hirooka et al 2011, Jing et al 2011, Kon et al 2011, Contadakis et al 2015, 2.5(Jhuang et al 2010), 2.9(Jhuang et al 2010) and 3(Contadakis et al 2015.Figure 5.7 shows the variations of Z-value with confidence level (according to probability theory) plotted over the ranges of Z-values used in the literature. The Z-value governs the detection of PETA and is more critical for distinguishing false positives(Chen et al 2004).…”

“…However, currently there is no clear justification for the selection of the N-day value used in previous PETA studies. The N-day values that have been used are 5(Nenovski et al 2015), 10 (Fuying and Yun 2011; Jing et al2011;Xu et al 2013b;Yiyan et al 2009), 15(Aggarwal 2015;Astafyeva and Heki 2011;Chauhan et al 2009;Chen et al 2004;Hasbi et al 2011;He et al 2014;Hirooka et al 2012;Hirooka et al 2011;Karatay et al 2010;Ke et al 2016;Kon et al 2011;Le et al 2011;Thomas et al 2017;Ulukavak and Yalcinkaya 2016), 27 (Guo et al 2015), 30 (Guo et al 2016Jhuang et al 2010;Oikonomou et al 2016) or even 60(Xu et al 2013b). Consequently, statistical envelope methods face the challenging task of accounting for ionospheric variability (owing to different phenomena such as space weather and possible seismic effects) within a single N-day value.…”