Medium-term earthquake forecasts by the LURR method on the example of the strongest earthquakes of the 21st century

Abstract

Aim. To carry out a retrospective forecast of the eight strongest earthquakes (with a magnitude greater than 8) of the 21st century using the Load/Upload Response Ratio (LURR) method.Methods. Calculations were performed according to the authors’ original data processing technique, which had been successfully tested while studying the seismicity of Sakhalin Island. A significant advantage of this methodology consists in using fixed (basic) parameters of mathematical processing and sampling of calculated earthquakes (geometry of the calculation area and sequestration by magnitude). This enables researchers to conduct not only retrospective studies, but also real-time forecasting. In the present paper, a seismic catalog of the United States Geological Survey (USGS) was used.Results. For four of the considered earthquakes, precursors in the form of anomalies of the LURR parameter were identified. These anomalies had manifested themselves in periods from six months to two years before the earthquake, which is consistent with our results for Sakhalin. At the same time, no other anomalies (false alarms) were recorded in three cases. For the other four earthquakes, no anomalies were recorded at all. It is noted that the level of seismic activity (in the operating range of magnitudes) in these cases allows the smoothing window to be reduced. In some cases, the calculation for a window twice as small (180 days) revealed pre cursors, at the same time as contributing to the appearance of additional, false anomalies.Conclusions. It was confirmed that the authors’ methodology for data preparation and calculation parameters selection, as well as the LURR method, apparently have limited possibilities for extrapolating to various seismically active regions of the planet. There may be different reasons, which have not been considered in the present paper due to the uncertainty associated with the capabilities of the USGS network.