In the light of the important role played by the return period in seismic hazard analysis, two types of statistical analyses (Gutenburg-Richter's law and Markov chain) were applied to study the earthquake periodicity in the Chiayi-Tainan area. There are two data sets prepared for this study; one contains earthquakes with ML ≥ 4 (283 events) during the 1900-1995 period, the other contains earthquakes with ML ≥ 2 (3816 events) during the 1973-1995 period. The mean return period and h values estimated with Gutenburg-Richter relation for earthquakes in the study area are in agreement with previous studies. The mean return period for ML ≥ 6 earthquakes is 9.7 years from the data in 1900-1995 period, and 17 years from the data in 1900-1995 period. The Markov chain measures the transition characteristics of earthquakes, and estimates that ML=6-6.9 and ML ≥ 7.0 earthquakes repeat at intervals of 28 and 135 of ML ≥ 4 earthquake events, respectively, in the Chiayi-Tainan area. This study also demonstrated that investigation of "substitutability" phenomenon is applicable to earthquake data, which suggests high similarity for low-magnitude earthquakes and dissimilarity for high-magnitude earthquakes. The result from these statistical approaches for earthquake occurrences leads to a conceptual model inferring fault behavior. This model is supported by the recent geologic findings, which may be potentially useful in earthquake hazard assessment.
All Science Journal Classification (ASJC) codes
- Geotechnical Engineering and Engineering Geology