Abstract
Abstract
In the present work we show a fundamental possibility of a short-term prediction of strong shallow-focus earthquakes in the Southern Sakhalin (on the Krillon earthquake example, 23.04.2017, Mw=5.0). Retrospective modeling of earthquakes sequences from 2013 to 2017 based on the method of self-developing process (SDP) and the catalogue data of the local network has required. An auxiliary tool was used to improve the accuracy of the SDP estimates. It is a medium-term earthquake prediction method called LURR (load-unload response ratio). The previous results were improved using different sample data in the zones of active fault structures. This avoids ambiguity in the choice of the computational domain for SDP calculations. The zone of maximal anomalous levels of LURR coincides with that one we obtained a solution by the method of SDP. It happened 77 days before the earthquake. When compared with average estimates by the LURR method (two years), the prediction time was concluded to go down by 11 times.