The ionosphere of the Earth is no doubt an important part of the global electric circuit and is a subject to dramatic perturbations related mainly with geomagnetic and solar activity. It also reacts on going from “below” impacts from different processes like thunderstorms, dust storms, radioactive pollutions, volcanic eruptions, earthquakes, etc. Recently the total electron content (TEC) of the ionosphere became one of the most often used parameter to describe features of the ionosphere behavior before, during and after the considered seismic events. TEC is the total number of electrons along the vertical path between the satellite and the ground in 1 m² cross section column; TEC is measured in TEC units (1 TECu = 10¹⁶ el/m²). It is derived from the regular ionosphere observations using GPS and GLONASS satellites (as sources of stable signals) and ground-based network of receivers of these signals.

Basing on numerous ground-based and satellite observations many papers reported significant ionosphere disturbances (including the TEC increases or depressions) existing during quiet periods of the geomagnetic and solar activity. As these disturbances happen just before and spatially near the earthquakes’ locations many investigators consider these ionosphere anomalies as precursors. Nevertheless the current knowledge state and field of the investigation are far from making reliable EQ’ forecasts.

This paper provides a summary of our investigations’ results in the field of the pre-earthquake TEC variations. We present the pre-earthquake GPS-observed TEC disturbances analysis’ results for 50 cases of M ≥ 6 seismic events happened during 2005-2006, and propose a physical interpretation for considered seismo-ionosphere anomalies supported by the Upper Atmosphere Model (UAM) numerical calculations.

Numerous case studies reported the following features of the TEC disturbances observed before strong earthquakes [1-8] (see Figure 1).

• Anomalous strong (≥30% - 90% relatively to the quiet values) positive or negative modifications (increases or reductions) before M ≥ 5 earthquakes linked to the

near-epicenter area. According to [7,8] the positive anomalies are more often observed.

• The size of the anomalies depends on the earthquake magnitude and extends larger than 1500 km in latitude and 3500 - 4000 km in longitude. The structure and dimensions of the disturbed areas are kept rather stable during 4 - 8 hours. These anomalies do not propagate along the magnetic meridians in contrast of the ionospheric disturbances caused by solar and geomagnetic activities.

• Local long-living negative and positive disturbances are reported to appear from several hours to couple of weeks before the earthquake’s release moment. The sign of the anomaly may change to the opposite one. Sometimes anomalies depress up to full decay before the earthquake release moment or a few hours earlier.

• The vertical projection of the epicenter does not mandatory coincides with the maximum phenomenon’s manifestation location.

• Similar effects are often observed at the magnetically conjugated region.

• In case of the strong low-latitudinal earthquakes increase or decrease of the equatorial anomaly with the trough deepening and filling takes place [9-11].

This set of the features was extended with terminator and sub-solar point effects in [11-14].

• Relative TEC disturbances reduce sometimes to almost full disappearance when the ionosphere conductivity increases, i.e., with the sunrise terminator and sub-solar point income.

• The disturbed areas shift from the terminator to the night-sector. TEC disturbances’ renewal takes place after sunlit ionosphere (sunset terminator) leave.

Table 1 presents the results of application of this feature-list for 2005-2006, –50˚ < φ < 50˚, M ≥ 6, D < 60 km earthquakes, where φ is the considered seismic event epicenter’s magnetic latitude [6]. The criteria were even hardened—we strongly required the geomagnetic conjugation of the pre-EQ TEC anomalies. According to the