GNGTS 2015 - Atti del 34° Convegno Nazionale

52 GNGTS 2015 S essione 1.1 al. (2001; Fig. 1) and Amatya and Jnawali (2006). The South Tibetan Detachment normal fault zone is not depicted in Fig. 1, for simplicity. In the following, we conform the fault-plane solutions of various authors to the convention of the fault plane dipping to the right of the positive direction of the strike (ranging 0°-360°) with the rake angle seen from the hanging wall and measured counterclockwise from 0 to 360 between the positive direction of the strike and the direction of the slip vector. We mention this kind of data in the following order: strike-, dip-, and rake-angle. We come now to the M S 8.2 Bihar-Nepal, 1934 earthquake. Dunn et al. (1939) erroneously imagined a strike-slip sense of motion along a west-northwest striking fault from the intensity observations. Chen and Molnar (1977) did not follow this interpretation. Rather, they: i) started from low-angle thrust faults found from fault-plane solutions of more recent earthquakes in the India-Nepal border region by Fitch (1970) and Molnar et al. (1973); ii) assumed a mechanism of this type, and iii) calculated the following source parameters: 27.55 N epicentral latitude, 87.09 E epicentral longitude, M 8.3. This epicenter was adopted by all subsequent studies up to that of Sapkota et. al. (2013) [its position in Fig. 4 by Hough and Bilham (2008) is wrong]. Singh and Gupta (1980) claim they used nine first motions from long-period records to attempt the calculation of the fault-plane solution, but only eight are seen in their fig. 2. They reached the questionable result of 100°, 30°, 40°, that implies a thrust plane dipping 30° almost south with 100 -ward directivity. Then, Hough e Bilham (2008) assumed the epicenter by Chen and Molnar (1977) and took into account «the region of maximum shaking intensity and subsidence as proxy measures of the centroid of the 1934 earthquake» (Hough e Bilham, 2008; page 776) in a semi-quantitative reasoning to draw the projection of the inferred rupture for the 1934 Bihar-Nepal earthquake. They also proposed up-dip rupture, with propagation from east to west (opposite to the direction calculated by Singh and Gupta (1980)); on an expert judgment basis, they chose a 90 rake-angle and admit that «Our preferred rupture is not well constrained». Unfortunately, in Hough and Bilham (2008) there are some typos and for this reason we cannot include their source in Fig. 1. Recently, Dixit et al. (2015, in press) moved the rectangular projection of Hough e Bilham (2008) to match the surface rupture of MFT found by Sapkota et al. (2013) and obtained the polygon (A) shown in Fig. 1 for the 1934 earthquake. In conclusion, the present automatic KF source inversion is the first attempt of constraining quantitatively the approximate area of main moment release and fault slip of the 1934 earthquake using only the MSK intensities reported by Ambraseys and Douglas (2004). Finally, the meizoseismal area of the M S 8.1, 1897 Shillong earthquake is out of the Himalayan belt and south of the so called Assam syntaxis, mostly in the Assam territory, in small part in Bangladesh. This zone is in rapid upheaval and bordered, south-ward, by a north- dipping thrust (see fig. 1A in Sapkota et al. , 2013.) Source inversion of the M S 8.1 Shillong, 1897 earthquake. Fig. 2A shows the MSK intensities of the M S 8.1 Shillong, 1897 earthquake reported by Ambraseys and Douglas (2004) within 120 km from the epicenter in 31 towns and villages (dots). Fig. 2B shows the synthetic intensities and the beach ball diagram that was produced by the minimum variance model (column 3 of Tab. 1) in the KF inversion (interpolation as in Fig. 2A). The verification of this result with measurements-based inversions is not possible. In the case of the 1934 earthquake, we just saw however that measurements-derived geometry and kinematics can give questionable results for earthquakes of the early instrumental era. The match of Fig. 2B is striking. The geometry and kinematics of column 3 of Tab. 1 is almost in perfect agreement with the seismotectonic situation shown by Sapkota et al. (2013) in their fig. 1A; we refer to the presence of an active thrust dipping north in the Assam territory close to the border with Bangladesh, with the epicenter of 1897 on the side of the hanging wall. ���� �������� ��������� This tectonic structure coincides with the ��������� ������� ��� �������� ����� ������ ������ ������� ��� �� ����� northerly dipping E-W trending Dauki thrust (Sarma (2014); 250 km long, Evans (1964)) which borders south the Garo, Khasi and Jainta Hills in Assam, close to the Assam-Bangladesh border.