Testing the MEE prediction algorithm in various seismically active regions in the 1985-2000 period: Results and analysis

The paper presents results of the 15-year testing of the MEE (Map of Expected Earthquakes) algorithm using a set of various precursors and intended for the medium-term prediction of strong, K greater than or equal to 12.5 (M greater than or equal to 5.5), earthquakes in various seismically active regions of the world. The average prediction effectiveness of the algorithm is shown to be hi-her than that of the random guessing by a factor of 2.5. On average, 65% of predicted earthquakes occurred in zones where the conditional probability of strong earthquake occurrence was P(D-1\K) greater than or equal to 70% and the average area of alarm zones amounts to 30% of the observation area. It is established that the most effective prognostic indicator used in the MEE algorithm is the density K-sf of seismogenic faults, whose average prognostic effectiveness was J = 4.96 +/- 2.39 in the regions tested. Results of this study suggest that the MEE algorithm is advantageous for more intense observations of other precursors in selected zones with a high (more than 70%) level of the conditional probability and for taking necessary preventive measures to reduce the economic and social damage that may be caused by a future strong earthquake.