Intensity attenuation relation at Chamba–Garhwal area in north-west Himalaya with epicentral distance and magnitude

Seismic hazard assessment of any region depends on the attenuation relation which relates the seismological data with parameters of engineering interest. In the absence of sufficient strong motion data for northwest Himalayan region, isoseismal maps of previous earthquakes were the only source of information for computing the attenuation characteristics of that region. The northwest Himalayan region had experienced major and moderate earthquakes in the past and isoseismal maps had been developed by various authors for each of these earthquakes representing intensity pattern and decay of intensity in different directions. With this information, attenuation relations of intensity with distance can be obtained. The present study includes 10 moderate and major earthquakes ( Ms ≥ 4.9) that had occurred during the last 100 years. These are primarily Kangra (1905), Chamba-1 (1945), Chamba-2 (1995), Kinnaur (1975), Dharamsala (1986), Uttar Kashi (1991), Garhwal-1 (1996), Garhwal-2 (1996), Sundarnagar (1997) and Chamoli (1999) earthquakes. Miezoseismal zones in majority of the isoseismal maps show two major directions, i.e., either the longest axis of these isoseismal maps which are aligned in NW–SE direction following the major longitudinal features or north-northwest to south-southeast direction, following the tectonic features which are transverse to the Himalayan trend. Isoseismal maps of Kangra (1905), Chamba (1945) and Uttar Kashi (1991) earthquakes are more symmetrical and elongated in south-east direction, but compressed in north-east and north-west directions, thus indicating fast attenuation characteristics in north-east and north-west directions reflecting lithological variation. On the other hand, Sundernagar (1997) and Kinnuar (1975) earthquakes show circular pattern of intensity distribution. Combining decay patterns from different magnitude earthquakes in different directions, two attenuation relationships have been developed using regression relationship (least square method).