| 摘要 |
FIELD: geophysics.SUBSTANCE: invention relates to geophysics and can be used for prediction of seismic event. Disclosed is a method for prediction of seismic events, based on combined processing of control parameter measurement results obtained in real time from a plurality of measurement points that cover a seismically active region. Data processing comprises generating a regular network of N×M nodes for seismically active region under investigation, selecting a sliding time window, determining, for each node in regular network of N×M nodes, degree of correspondence S of changes in parameters being monitored and/or multifractal parameter Δα(τ) of singularity spectrum carrier width (hereafter SSCW) Δα as mean of Δα values at a certain number of measurement points closest to node (i, j) amongst total number n of measurement points covering seismically active region. Then, using values for degree of correspondence S at each node (i, j), for each time window at current moment in time τ, sub-area of region under investigation which is deemed to be an area with increased risk of seismic activity within current sliding time window is determined by comparing degree of correspondence S with a threshold value. Threshold value is determined based on statistical analysis of value S for preceding seismic events in seismically active region.EFFECT: technical result is high accuracy of prediction of expected seismic event.19 cl, 7 dwg |
