Here we test the precursory enhancement in ionospheric total electron content (TEC) which has been reported by Heki (2011) and numerous Global Navigation Satellite System (GNSS) TEC observational studies before the 2011 Mw9.0 Tohoku-Oki and many great earthquakes. We verify the frequency of this TEC enhancement via analysis of a two-month vertical TEC (VTEC) time series that includes the Tohoku-Oki Earthquake using the procedure, based on Akaike’s information criterion, and threshold of Heki and Enomoto (2015). The averaged occurrence rate of the TEC enhancement is much larger than that reported by Heki and Enomoto (2015) when all of the visible GPS satellites at a given station are taken into account. We cannot rule out the possibility that the pre-seismic VTEC changes before the great earthquakes that were reported by Heki and Enomoto (2015) are not precursors but just a product of chance. We also analyze the spatial distribution of the pre-seismic TEC enhancement and co-seismic TEC depletion for the Tohoku-Oki Earthquake with the data after reducing inter-trace biases. We observe significant post-seismic depletion that lasted at least 2 h after the earthquake and extended at least 500 km around the center of the large-slip area. This means that evaluation of the enhancements using reference curves which was adopted by Heki 2011 and even by the recent papers (e.g. He and Heki 2016, 2017, 2018) is in danger of mistaking a large and long-lasting post-seismic TEC depletion for a pre-seismic enhancement.

We investigate the veracity of the reports by Iwata & Umeno (2016, doi:10.1002/2016JA023036) and Iwata & Umeno (2017, doi:10.1002/2017JA023921), both of which claimed that the observed perturbations in GNSS-based ionospheric total electron content (TEC) could serve as a “precursor” of large earthquakes based on correlation analysis. Iwata & Umeno (2016) defined a spatial correlation of residuals between the observed and predicted TEC time series. They reported that the correlation value is significantly larger before large earthquakes than those observed during non-earthquake periods. Iwata & Umeno (2017), who applied the same method to other large earthquake, claimed that the preseismic ionospheric disturbances can be distinguished from other non-earthquake phenomena based on the small percentage of area where the correlation value exceeds the criterion. They also claimed that the low propagation velocity of the correlation peaks is also a pre-seismic characteristic. Here we tested their claims using a larger dataset. As a result, these three characteristics they claimed to have captured as evidence of earthquake precursors are not significant being frequently observed during normal (non-earthquake) days. In addition to that, the criteria of Iwata & Umeno (2017) cannot be applied to the large earthquake discussed by Iwata & Umeno (2016), and vice versa. Therefore, we can find no basis for claiming that they detected precursors to the earthquakes. The calculation procedure of the correlation function shows that the value is more of an indicator that amplifies small variations synchronized between nearby stations, like medium-scale traveling ionospheric disturbances rather than earthquake precursors.

We conduct numerical experiments to examine two studies that reported preseismic anomalies in the ionospheric total electron content (TEC) and argued for the significance of their respective analyses based on statistical evaluations. The first study is Liu et al. (2018), who statistically studied the relationship between 62 M≥6 earthquakes in the Chinese interior over an 18-year period and the TEC, which was deduced from the Global Ionospheric Map. The TEC showed anomalies with specific polarities at set times during certain days that preceded the earthquakes. They defined alarms based on this and drew receiver operating characteristic curves, which yielded a significantly better performance (higher area under the curve (AUC) and lower p-value) than random alarms. We conduct this analysis using random synthetic earthquakes. The resulting AUC and p-values are very similar to those for real earthquakes, indicating that the high performance of the Liu et al. (2018) alarm is an artifact. The second study is Le et al. (2011), who classified the TEC time series into anomalous and non-anomalous days based on the TEC perturbation. They found that the anomalous day rate increased as the nucleation time of the earthquakes was approached, especially for larger and shallower earthquakes. We conduct the same analysis using random synthetic earthquakes. The anomalous day rate that is comparable to their result occurs in ~40 % of the 1,000 random trials, thereby suggesting that their result may also be an artifact.

Deep tectonic tremors occur in the Nankai subduction zone, defining a belt-like zone with a width of a few ten km located at depths between 30 and 55 km along upper surface of the subducting Philippine sea slab. We interpret the geometry of the tremor belt based on temperature calculations. Time evolution of temperature is calculated using a 3-D heat conduction assuming constant geometry of the slab and the present-day plate convergence velocity. The results show that the tremors occur along the 450°C isotherm along the slab surface where the mantle wedge can be non-convective and already well-serpentinized. Two gaps of the tremor belts are explained by different mechanisms. The Ise gap is where the hanging wall is not mantle wedge but crust. The Kii Gap is not a gap orthogonal to the plate convergence direction, but only appears to be a gap because the isotherm of the plate surface steps in the plate convergence direction. The tremors can be caused by high pore-fluid pressure conditions due to aqueous fluids released by dehydration reactions in blueschist-facies oceanic crust to form eclogite facies. The depth of the tremor belt corresponds to that of decoupled, non-convective mantle wedge. Since the temperature and pressure are within the serpentinite stable condition, the released fluid would normally be absorbed by serpentinization of the mantle wedge peridotite. However, since the non-convective mantle wedge is already well serpentinized from exposure to previously released fluids, the newly released fluid is not absorbed and increases the pore pressure.