Two improvement measures pointed to the past error to calculate b‐values have been proposed: (1) adopt seismic moment magnitude MW which is the uniform absolute magnitude scale; (2) discard the magnitude point in which some earthquakes are missing, then compute the b‐value using least square method to fit the data. Therefore b‐value is more accurate. The b‐values all approach to 1.0 for three sets of various precision and sources. The earthquake occurrence periods of great earthquakes for coming 50~100 year (extend 1~2 magnitude) have been computed using 13 years recent earthquake observation data and fix b=0.85. The results demonstrate that the earthquake occurrence periods with magnitude greater and equal to 6 and 7 are well consistent with that which are calculated from the data of 100 years historical earthquakes. So, the extension is reliable.

No abstract

No abstract

In the paper, we analyze 117 moderate-strong earthquakes occurred in Chinese mainland (Ms>5.5 in the east and Ms>6.0 in the west) since 1970, among them, 11 earthquakes (about 9%) have direct foreshocks and 63 earthquakes (about 51%) have generalized foreshocks. The predominant time interval between foreshock and main earthquake is no more than 30 days with a spatial distance less than 50 km and a magnitude difference over 1. From the digital seismic data in Liaoning Province, we know that direct foreshock had an obvious shear-stress background before the Ms=5.6 and Ms=5.1 Xiuyan earthquakes occurred on Nov. 29, 1999 and Jan. 15, 2000.

No abstract