Abstract: In this paper, the goodness-of-fit test (GFT) method for estimating the minimum magnitude of completeness (M_C) is introduced first and then applied to determine the temporal-spatial distribution of M_C of the Longmenshan fault zone, Sichuan, and its surrounding area. Finally, advantage and problems of the method in application are summarized and discussed. Our study suggests that based on the assumption that the G-R relationship has the power-low distribution, the GFT method is used to estimate a minimum magnitude of completeness M_C through fitting a G-R relationship for a set of observed data by adjusting various minimum magnitudes gradually from small to big ones and eventually taking that minimum magnitude when the fitted G-R relationship can match best the set of observed data to be the result. Also, in the method a fit evaluation function is defined and used to measure the goodness of fitting. Our application results show that in the studied region the minimum magnitudes of completeness M_C show a tendency of decreasing gradually with time from 1970, reflecting a gradual improvement in the seismic monitoring ability of the region. The spatial distribution of M_C are basically comparable with that of the theoretical monitoring ability of the regional seismic network, and have more details about the distribution. In addition, we also found that, as a seismicity-based arithmetic, the GFT method may underestimate or overestimate M_C when seismic data for limited areas and for short time-period are used. Based on the estimated M_C, b-value scanning is carried out for the studied region. The result shows that b-values of background small earthquakes along the Longmenshan fault zone and its nearby area seem to have slightly risen 3 to 8 years before the 2008 Wenchuan M8.0 earthquake.

Key words: Minimum magnitude of completeness, Goodness-of-fit test method, Longmenshan fault zone, b value