Atypical Site Effects and Insufficiency of the Conventional Seismic Site Classification Methods: Experimental observations in the Geological Deposits of the Kashmir Valley (NW Himalayas), Jammu and Kashmir

Kashmir Valley is one of the two administrative divisions of the Jammu and Kashmir region, lying within the Himalayan belt at the confluence of the Indian and Eurasian tectonic plates. It is an elongated basin filled with deep sedimentary deposits in the central portion, bounded by the Himalayan Mountain ranges at the periphery. The Pleistocene Karewa deposits overlain by the Recent Alluvium form the major types of sediments in the valley, showing a thickness of over 1300m (Burbank and Johnson, 1982). To evaluate the seismic response of these geological deposits, we recorded microtremors and performed multichannel analysis of surface waves (MASW) testing at ~190 sites using TROMINO^® equipment. Horizontal-to-vertical-spectral ratio (HVSR) curves and dispersion curves were respectively developed from the raw data, which were further used to estimate the site fundamental frequency (f₀) and H/V amplitude (A) as well as average shear wave velocity over 30m depth (V_S,30). A comparison of f₀ and A with V_S at each site was conducted to assess the suitability of codal provisions (NEHRP, EC8, etc.) and the specified amplification factors.

We found certain issues with the use of V_S,30 as the sole parameter for estimating the expected amplification at a site for seismic design. A direct correlation between amplification and stiffness (V_s30) at the sites was not attained, as opposed to the most common assumption in the prevalent code-based site classifications. Low amplification can be expected even at soft soil sites (NEHRP E) in the absence of an impedance contrast, and vice versa in stiff deposits. Observations in the deep sedimentary deposits in the valley show the importance of the contribution of deeper stratigraphy beyond the assumed 30m depth in the codes. Moreover, the notion of f₀ being directly proportional to V_s30 irrespective of the geological conditions is not true. Besides, rock sites (NEHRP A and B) may not always present the conventional flat HVSR response, instead, significant high frequency amplifications can occur due to weathered or fractured material. Test results also revealed unusually high amplitudes within the fractured portion of a fault zone, signifying that the assumption of all rock sites being safe in terms of negligible amplifications may not be entirely correct. Topographic amplifications near basin edges, hill slopes, and within small valleys resulted in broad peaks over wide frequency range in the HVSR curves, which cannot be explained through the limited information contained in the V_s30 parameter.

These incongruities observed at various sites in the Kashmir region point out the inadequacy of the site classifications based on the single-parameter (V_s30)approach, which do not always explain the actual amplifications observed during earthquakes in different geological conditions. The results obtained in the Kashmir Valley, backed by the similar arguments provided in literature (e.g., Lombardo and Rigano, 2006; Castellaro et al., 2008; Rovelli et al., 2009; Panzera et al., 2014; etc.), clearly indicate the need for adopting alternate site classification schemes (Di Alessandro et al., 2012; Pitilakis et al., 2019; Paolucci et al., 2021).