Revisiting the 1897 Shillong and 1905 Kangra earthquakes in northern India: Site Response, Moho reflections and a Triggered Earthquake

Re-valuated instrumental magnitudes and intensity distributions for the Mw8.0 Shillong Plateau earthquake of 1897 and the Mw7.8 Kangra earthquake of 1905, combined with newly available geodetic constraints on rupture geometries, allow us to compare observed distributions of intensity with those predicted from theoretical models for shaking produced by each earthquake. The difference between predicted and observed shaking is interpreted in terms of the site response of the Ganges and Brahmaputra basins. As expected, these comparisons identify regions of enhanced shaking near the main rivers. We infer amplifications of 1-2 intensity units, corresponding to an amplification factor of 2-4 in peak ground acceleration. We also find two unexpected results in our comparison of observed and predicted shaking from the Kangra earthquake: 1) The epicentral region is surrounded by a halo of enhanced intensity at 150-200 km radius and 2) The Dehra Dun region was the locus of a broad region of anomalously high intensities. We interpret the former result as the signal from post-critical Moho reflections and the latter observation as a probable second large earthquake (M>7) at 30-50 km depth triggered within minutes of the 1905 mainshock. These results have important consequences from future earthquakes in the Himalaya.