房総半島の完新世旧汀線からみた「大正型」関東地震の平均再来間隔
スポンサーリンク
概要
- 論文の詳細を見る
Several Holocene former shorelines are observable in the southern part of the Boso Peninsula, southern Kanto, central Japan. They are considered to have emerged by co-seismic uplifts, and their height distribution must indicate the accumulated amount of seismic vertical crustal movements of 1703 (Genroku) and 1923 (Taisho) types. In this study, the mean recurrence time of the 1923 type earthquake is estimated on the basis of the most fitted combination of the two type movements to the height distribution of the N-surface shoreline, <I>i.e.</I> the raised shoreline formed at the culmination stage of the sea level in the middle Holocene (<I>ca</I>. 6, 200 years BP).<BR>The data of 36 localities (Fig. 2) on <I>H<SUB>N</SUB>, HG</I> and <I>h<SUB>T</SUB></I> (Table 1) has been used to find the most fitted combination of the 1703 and 1923 type movements, where <I>H<SUB>N</SUB></I> is the height of the N-surface shoreline, <I>H<SUB>G</SUB></I> is the height of the Genroku shoreline <I>i.e</I>. the raised shoreline which emerged in 1703 coseismically, and <I>h<SUB>T</SUB></I> is the geodetically measured co-seismic uplift amount at the 1923 earthquake. The relationship between the three variables are as follows : <BR><I>H<SUB>N</SUB></I>=2.8<I>H<SUB>G</SUB></I>+2.6<I>h<SUB>T</SUB></I>+5.8<BR>The high value of the multiple regression coefficient of 0.911 of the equation indicates that the main Holocene crustal movements in the southern part of the Boso Peninsula have been the 1703 and 1923 type seismic crustal movements undoubtedly. The equation also shows that the sum of the sea level drop amount and the place-independent uplift amount afer the emergence of the N-surface shoreline is approximately six meters. Apart from the two types of the seismic crustal movements, fault displacement is likely to have been playing a certain role for the Holocene crustal movement.<BR>Two possible models, models A and B, of the seismic vertical crustal movement has been adopted in this study. Model A is the case that the 1703 and 1923 type seismic crustal movements occur independently (Fig. la). Model B is the case that only an event of several events is of 1703 type and the other events are of 1923 type (Fig. 1b).<BR>It is known that the mean recurrence time of the 1703 type earthquake has been around 2, 000 years during the past 6, 200 years. Observed value of mean annual uplift amount for the recent 32 years at the Mera Tide Observatory (Loc. 19) has been -0.0019 meters. Based on these data, the mean recurrence time of the 1923 type earthquake has been calculated to meet the above-mentioned regression equation. The resultant values are 890 years in the case of model A and 610 years in the case of model B. The interseismic recovery ratios have been also calculated. The recovery ratio of the 1703 type movement (<I>d<SUB>G</SUB></I>) is 0.44 in both cases of model A and B, and that of the 1923 type movement (<I>d<SUB>T</SUB></I>) is 0.37 in the case of model A and 0.39 in the case of model B. The result of model B is presumably more harmonious with the geomorphologically deduced Holocene shoreline shift history.<BR>It is likely that the data have included certain errors. The calculation has been based on very simple models. Moreover, some unconfirmed assumptions, such as the 1703 and 1923 events are the typical representatives of the 1703 and 1923 type earthquakes, respectively, have been used for the calculation. Therefore discussion in detail is not so meaningful. In this study, it can be concluded that the mean recurrence time of the 1923 type earthquake has been estimated as several hundred years to one thousand and several hundred years.