HR: 1340h
AN: S53B-1101    [Abstracts]
TI: Long-term evaluation of seismic risks from sub-characteristic earthquakes with new insight from smaller surface faulting events on the North Anatolian fault and in Japan
AU: * Okumura, K
EM: kojiok@hiroshima-u.ac.jp
AF: Department of Geography, Hiroshima University, 1-2-3 Kagamiyama, Higashihiroshima, 7398522 Japan
AB: Severe ground shaking from Mw 6.5--7.0 blind faulting events as in Kobe 1995, in Bam, Iran 2004 or in Chuetsu, Japan 2004 often causes intensive damages. These unexpected but intensive shocks raised skepticism on long-term earthquake forecasts based on historic and geologic records. This paper aims to examine the validity and limitation of geologic data on the smaller but hazardous earthquakes, and proposes a strategy to better understand those risks. The characteristic earthquake model assumes that a possible largest earthquake from a geologically defined source (as a segment) is to be repeated, rather than partial faulting events in random manner. Here the smaller but hazardous earthquakes are defined as sub-characteristic because they usually break only a part of source fault plane. But they are usually too large to be regarded as a part of non-characteristic background seismicity. Thus the earthquakes are to be classified as characteristic (CC), sub-characteristic (SC), and non-characteristic (NC). Another common criterion to tell the nature of faulting events is blind faulting (BF) or surface faulting (SF). Combination of these two categories brings 6 types of earhtuqakes while a simplistic model refers only to CC-SF and NC-BF. Now we know there are possible CC-BF as 1811-12 New Madrid and SC-BF like Bam 2004 and Chuetsu 2004. In addition, there are possible SC-SF earthquakes along large transform faults like the North Anatolian fault in Turkey. Following discussions concentrate on two types of the sub-characteristic earthquakes. Most of the recent SC-BF earthquakes occurred either at around a termination or a discontinuity of major seismogenic structures. Those SC-BF events rupture a small deeper portion of the structure or an adjacent minor structure without generating significant offset on the major ones that have built up clear signature of CC-SF events at the surface. Geologic information on the major structures does not tell us the timing of such SC-BF events but clearly indicates the likely place. It is necessary to analyze the relationship between CC-SF and SC-BF events. This may reveal time-dependent behavior of the SC-BF events. On the other hand, if the background seismicity is to include SC-BF events, time-independent probabilities should be evaluated. The smaller surface faulting events from the North Anatolian fault like 1942, 1951, and 1957 events may be SC-SF events. If this is the case and we are able to separate the time series of SC-SF from CC-SFs like in 1939, 1943, and 1944, there is good chance to know the behavior of sub-characteristic events. Other historic and geogloci data on possible sub-characteristic and surface-faulting events from Turkey and Japan will be analyzed to explore this hypothesis.
UR: http://home.hiroshima-u.ac.jp/kojiok
DE: 7212 Earthquake ground motions and engineering seismology
DE: 7221 Paleoseismology (8036)
DE: 7223 Earthquake interaction, forecasting, and prediction (1217, 1242)
DE: 8107 Continental neotectonics (8002)
SC: Seismology [S]
MN: Fall Meeting 2005