Microtremor based earthquake disaster risk evaluation of the Kathmandu Valley

Abstract

Nepal occupies the central one-third portion of the Himalayan Arc, which is formed by the collision of the Indian and the Eurasian plates that started about 55 million years ago. Many researchers report interseismic strain convergence between these plates in Nepal at a rate of 16-18 mm/year, and the release of this strain is considered the main reason for high earthquake risk in this region. Records show that Nepal and Kathmandu Valley have experienced recurring destructive earthquake in 1255, 1408, 1681, 1803, 1810, 1833, and 1866, while the latest strong quake was experienced in 1934, which had a magnitude of Mw = 8.1, destroyed about 19% and damaged about 38% of the buildings in the valley. A loss estimation study indicates that if an earthquake of the1934 class recurs, massive damage to building structures and heavy loss of life can be expected. The trends of structural damage in the past earthquakes indicate that the valley is characterized by strong site effects, which was formerly a lake with sediment thickness as high as 500 m in the central part. Moreover, the thickness and properties of the valley sediments vary from place to place, which may cause trapping and focusing of seismic waves during an earthquake leading to an evident change of resonant frequency over a short distance. A few studies attempt to address the problem of earthquake hazard in the Kathmandu Valley, but most of them focus more on a generalized approach to understatand the earthquake hazard of the valley. Much of the risk reduction efforts have focused only on management and legal issues leaving the scientific aspects of potential damage in shadow. Particularly, the dynamic behavior of the sediment deposit must be considered while assessing the earthquake disaster risk in the Kathmandu Valley. Moreover, an in-depth study particularly in relation to the influence of sparsely distributed depth of sediment deposits and local variation in ground response to the structural damage is more important, considering the adoption of a structural design code and mitigation of expected disaster in the valley. One way to addressing this issue is ground response and building vulnerability analysis. There are different approaches for ground response analysis, such as from conventional method of numerical analysis using borehole data to advanced techniques like horizontal-to-vertical spectral ratio method using strong ground motion and microtremors. Evaluation of ground characteristics through numerical analysis from borehole drilling is expensive and slow, and installation of seismic observation stations at required sites is often not feasible in densely urbanized area like Kathmandu Valley. Therefore, the horizontal-to-vertical spectral ratio technique using microtremor measurements is the only option in the the Kathmandu Valley. In this study, an attempt has been made to characterize the Kathmandu Valley Basin in order to address the issue of earthquake hazard risk mitigation in the valley through analysis of microtremor data recorded at a total of 172 points using horizontal-to-vertical spectral ratio technique. This study is divided into four parts. In the first part, variation of dominant period in the study area is obtained, and then an estimated dominant period of shaking map is prepared. The obtained results and prepared map indicate that the central part of the valley that covers about 30 percent of the urban area of the valley falls under dominant period range from 1.0 s to 2.0 s, while the outer parts fall under dominant period range from 0.1 s to 1.0 s. The longer dominant period range in the central part is correlated with the extent of depth of Pliocene and Quaternary deposits, while the shorter dominant period in the outskirts corresponds to shallow stiffer deposits near the rocks flanking the valley sides. In the second part, the characteristic of the Kathmandu Valley is evaluated by analyzing the obtained horizontal-to-vertical spectral ratio curves of 172 measurement points. The investigation results particularly reveal that the valley sediment is characterized by two amplified frequencies, which vary from 0.48 Hz to 8.89 Hz in first resonance and 3.1 Hz to 7.5 Hz in second resonance. These two amplified frequencies were detected at about 20% of the sites that are mainly distributed in the central and northern parts of the valley. These two resonance frequencies are characteristic to particular site condition, and describe the overall seismic site response in two scales, including the deep and surface soil layers. Depending upon the area, especially in the central and northern parts, it is found that the top 10-20 m depth of the sediment plays an important role in causing the second resonant effect in the valley. In the third part, an analysis of the floor variation of the lacustrine sediments in the Kathmandu Valley is carried out. Due to lack of adequate and precise scientific studies on the floor variation of sediments in the valley, however, it is always difficult to ascertain their characteristics during earthquakes, which ultimately leads to erroneous and assumed data for ground modeling as well as analysis and design of the infrastructures. This particular study attempts to fulfill this gap by proposing an approximate basement topography of the valley, which is mainly based on the interpretation of the microtremor data and analysis results. According to the obtained results, the soft sediment depth can be expressed in terms of frequency by a non-linear regression equation h=146.01fr which helps estimate the basement topography of the Kathmandu Valley. In the final part, a rapid visual seismic vulnerability assessment of the buildings is done using a simple tool prepared specifically for the Kathmandu Valley’s buildings, since other tool may not be suitable because of the difference in building types in the valley. The survey results in four typical locations in three cities of the valley indicate that more than twothird buildings in the urban center of the valley are vulnerable to earthquakes. As an important conclusion, the sediment in the Kathmandu Valley is expected to shake the valley buildings during an earthquake in two ways: one due to shaking of the topmost layer and other due to the shaking of layer underneath. As higher fundamental frequencies are characteristic to the topmost layer, this mostly affects low- to medium-rise buildings, while the lower frequencies are characteristic to the layer underneath, so this may affect medium- to high-rise buildings during an earthquake. Additionally, because of the increasing population and development as a greater commercial hub, the central part has recently seen a sharp rise in the number of medium to tall buildings, which have been constructed without adequate geotechnical investigation. In order to reduce the disaster risk, therefore, the multilayer ground effect must be considered in seismic design of long period structures in the valley center. At the same time, it may also be necessary to consider enhanced seismic design criteria for medium- and short-period structures in the valley.