A qualitative assessment of seismic risk along the Peninsular coast of India, south of 19 degrees N

Many earthquakes have been recorded from the coastal margin of the Indian p eninsular shield during the last 200 years. Largely made up of Precambrian assemblages with variable cover of Jurassic to Quaternary sedimentary rocks and Cretaceous-Eocene volcanics, the peninsular shield was long held to be aseismic. Recent measurements, however, show that this continental fragmen t is being pushed northeastward by the Carlsberg and Central Indian ridges, and the Indo-Myanmar subduction zone is exerting vigorous slab pull toward s the east. Repeated cycles of sea level change during the Quaternary have also induced continuing hydro-isostatic adjustment due to variable melt wat er loading in the Bay of Bengal and the Arabian sea. All these forces produ ce space-time fluctuations of strain around many small to large faults, whi ch occur in the upper crust of the shield. Some of the faults have been int ermittently active (during the past 100 kyr); others were active earlier. A lthough the Shillong plateau and the associated hill ranges of northeastern India and Myanmar are subject to the maximum seismic hazard. the peninsula r coast is also vulnerable to intermittent seismicity. We present illustrat ive evidence of some active faults. which are recognisable (a) on coastal l and by displaced Pleistocene weathered cover, hot springs, leakages of nati ve mercury and allochthonous geochemical anomalies of base metals and (b) o ffshore below the inner shelf by horst-shaped uplifted segments and intra-f ormational slump folds on and below the top shallow seismic (3.5 kHz) refle ctor. On the other hand. there are long stretches of the east coast at Vish akhapatnam and Manappad Point, which do not show active faults. Step-like m arine terraces, which occur up to + 6 in above the low tide level (LTL) pre serve records of relative sea level fluctuations during the Holocene and th e Last Interglacial. In such sectors, absence of tectonic disturbance durin g the last 100 ka is also corroborated by lateral continuity of shallow sei smic reflectors below the inner shelf over many kilometers. Since authentic historical (200-1000 years B.P.) records of seismicity along the Peninsula r coast are virtually unavailable., the likely recurrence interval between earthquakes in each sector cannot be gauged. We, therefore. propose a scale of seismic risk. based on geometry of the mappable faults and available se ismic records of the last two centuries. These could be used in combination to rank the densely populated coastal tracts sector-wise. (C) 2001 Publish ed by Elsevier Science Ltd. All rights reserved.