RELATIONSHIP BETWEEN PRECEDING SEISMICITY AND THE PROBABILITY OF STRONG AFTERSHOCK OCCURRENCE

The paper considers the relationship of seismicity, preceding the main shock, with the emergence probability of strong aftershocks in the future series. (Strong aftershock in the sense of Bath's law means that its magnitude not lower than the average difference of magnitudes of the strongest aftershocks and their mainshocks.) A hypothesis of the research – strong aftershocks are more likely to occur due to mainshocks occurred in places of high background seismicity. Testing the hypothesis at the global level was carried out using ANSS ComCat earthquake catalog US Geological Survey, at the regional level – using earthquake catalogs provided by Geophysical Service of the Russian Academy of Sciences for seismic regions of Russia (Kamchatka and the Kuril Islands, Baikal and Transbaikalia, the North Caucasus). We tested several functions that characterize previous background seismic activity relative to the mainshock. The values of the functions were considered as a possible precursor or anti-precursor of a strong aftershock. The effectiveness of the precursor (anti-precursor) was evaluated by a specially developed criterion, representing the ratio of the sum of all successful forecasts to the number of all unsuccessful forecasts. The value at which the maximum efficiency is achieved was taken as a threshold. The value of the previous activity greater or less than the threshold was considered as a precursor or anti-precursor of a strong aftershock, respectively. As a result, the hypothesis of the study was confirmed at the global and regional levels, regardless of the method of measuring previous seismic activity. The most informative characteristic of activity is the ratio of the accumulated seismic moment of background earthquakes preceding the main shock to the time of the main shock, normalized to the area of the circle bounding the background seismicity region. The probability of expected repeated shocks was estimated using the Reasenberg–Jones model depending on time and magnitude. We estimated the model parameters for the Earth and seismically active regions of Russia both with and without preceding seismicity. Comparison of theoretical and model values of the probability of occurrence of at least one strong aftershock at different time intervals showed that the model corresponds well with the initial data. Using the probabil ity gain, we shown that Reasenberg–Jones model, which takes into account preceding background seismicity, is more preferable than the model without it.