Site Response Characterization Investigation of Kathmandu Valley by HVSR Ambient Noise Analysis

Abstract Microtremor measurements for the seismic microzonation of the Kathmandu Valley were conducted at 75 stations spread across a 500 m × 500 m grid in the current study. The amplification factor (Ap) and predominant frequency (Fp) of the surface soils are estimated using the Horizontal-to-Vertical Spectral Ratio (HVSR) approach. The findings indicate that Fp ranges from 0.7 to 5.45 Hz and Ap spans from 1.0 to 6.2 Hz throughout the valley. Higher frequencies are found on the basin borders and in the southern regions where shallower stiff deposits predominate, whereas lower frequencies are found in the northern regions with thick lacustrine sediments. The soil classification revealed 1.3% very soft, 24% soft, and 74.67% stiff soils based on the fp. Bhaktapur (B-21) and Lalitpur (L-24), two highly populated locations with a fragile building stock, were found to have significant amplification sites, suggesting a high potential for soil-structure resonance. The stable zones below 1.5 Hz showed low amplification, while Bhaktapur (4.39 Hz) and Lalitpur (5.45 Hz) showed the highest amplification, suggesting potential liquefaction or slope instability risks that need to be verified by geotechnical data and shear-wave velocity profiling. By facilitating neighborhood-level hazard assessments and offering helpful data to architects, engineers, and urban planners, the 500 m grid spacing greatly increased the microzonation resolution. Despite its limitations, such as the inability to completely quantify soil dynamics, the absence of two or three-dimensional basin modeling, and ambient noise interference, the HVSR approach is a reasonable starting point for valley-wide microzonation. To validate risk zones and improve engineering applications for seismic hazard assessment and catastrophe mitigation, future research should include these findings with MASW surveys, borehole Vs30 data, and numerical basin response models. The findings show that macro-scale danger maps often oversimplify the highly varied soil in the Kathmandu Valley.