Keywords

5.1 Introduction

The earthquake, by its name, can simply be defined as the shaking of the ground, which is caused by sudden release of the energy stored in the rock mass that had accumulated as strain over time along the geological faults. As per data base of the United States Geological Survey (USGS) every year in average 22,600 earthquakes occur around the globe. Among them nearly 157 events of magnitude 6–6.9, 15 of magnitude 7–7.9 and one of magnitude 8–9.9 jolt the different part of the earthquake prone countries causing different scale of damages due to seismic shaking as well as earthquake induced secondary disasters.

The historical as well as instrumental seismicity have evidently shown that the Himalayan region is very much prone to earthquake and associated disasters. The geological, seismological, geodetic and geophysical data have also indicated that great earthquake may hit the major seismic gaps putting lives of millions of people at risk. The damage intensity of any earthquakes depend on several factors; such as seismo-tectonics, local geology, topography, soil condition, magnitude of the event, distance from the epicenter, nature of earthquakes etc. For example, the Ms 8.1 Mexico earthquake of September 19, 1985 is one of the few earthquakes that caused extensive loss of life and property (Anderson et al. 1986). At least 9,500 people were killed, about 30,000 were injured, more than 100,000 people were left homeless, and severe damage was caused in parts of Mexico City and in several states of central Mexico. It is estimated that the earthquake badly affected an area of approximately 825,000 km2, caused between 3 and 4 billion U.S. dollars of damage, and was felt by almost 20 million people. The earthquake collapsed 412 buildings and another 3,124 were seriously damaged in Mexico City. About 60 % of the buildings were destroyed at Ciudad Guzman, Jalisco. The event had two unique features, first the massive damage was observed some 400 km long distance from the epicenter and second was subsurface condition; substantial amplification of seismic wave of low frequency in the lake beds (Celebi et al. 1987). The detail investigation of the local site effect has identified the amplification of the seismic wave as the main culprit. Having similar subsurface geological condition and frequency of destructive earthquake in the Himalaya, the Kathmandu valley may suffer massive damage during the great earthquake of similar magnitude. Therefore, first part of the this chapter mainly reviews the active tectonics, characteristics of active seismogenic faults, and seismicity in terms of earthquake risk in Nepal Himalaya and the second part is mainly dedicated to assess the seismic hazard in terms of site effects in Kathmandu valley, where capital city of Nepal, Kathmandu is situated aiming that such study may be key to understand level of seismic hazard in other intermontane valleys in the Himalayan region.

5.2 Seismo-Tectonics of Nepal Himalaya

The collision between the Indian and Tibetan plates some 50 Ma ago has initiated the Himalayan mountain building processes. The continuous plate convergence has resulted a typical seismotectonic features in the Himalaya, which governs the entire seismicity in the region. The major tectonic features of the Himalaya to the south of South Tibetan Detachment System (STDS) is controlled mainly by three major thrusts, e.g. from north to south the Main Central Thrust (MCT) Main Boundary Thrust (MBT) and Main Frontal Thrust (MFT) (Fig. 5.1). These thrust faults with a North to South propagation direction are running almost the entire length of Himalaya and are generally inferred to be splay thrusts of the Main Himalayan Thrust (MHT), which marks the underthrusting of the Indian Plate (Fig. 5.2) The MFT system consists of two or three thrust sheets composed entirely of Siwalik rocks, from bottom to top mudstone, multi-storied sandstone and conglomerate. These sedimentary foreland basin deposits form an archive of the final stage of the Himalayan upheaval and record the most recent tectonic events in the entire history of Himalayan evolution since ~14 Ma. The northernmost thrust sheet of the MFT is truncated by the Lesser Himalayan sequence and overlain by unmetamorphosed to weakly metamorphosed rocks of the Lesser Himalaya, where the Lesser Himalayan rock package is thrust over the Siwalik Group along MBT. In western Himalaya crystalline thrust sheets are frequently observed within the Lesser Himalaya. The Lesser Himalayan zone generally forms a duplex above the mid crustal ramp (Schelling and Arita 1991; Srivastava and Mitra 1994; Decelles et al. 2001). The MCT system overlies the Lesser Himalayan MBT system and was formed in ca. 24 Ma. This MCT system consists of high-grade rocks, e.g. kyanite-sillimanite gneiss, schist and quartzite and is mostly characterized by ductile deformation.

Fig. 5.1
figure 1

Geological map of Nepal (modified after Upreti and Le Fort 1999). LH lesser Himalaya, HH higher Himalaya, TTS Tibetan-Tethys sequence, MBT main boundary thrust, MCT main central thrust, MFT main frontal thrust, STDS South Tibetan detachment system

Full size image
Fig. 5.2
figure 2

North–South geological cross-section of Nepal Himalaya showing major thrust system and microseismicity (Pandey et al. 1999). The southern flat along the main Himalayan thrust (MHT) is locked during the inter-seismic period causing an accumulation of elastic stress, which is released during microseismic events in Nepal Himalaya. The major (great) earthquake is generally originated along the northern flat just front of the Higher Himalaya and deformation is propagated along the MFT, a southern expression of MHT

Full size image

5.3 Genesis of Earthquakes in Nepal Himalaya

The occurrence of the earthquakes is well explained by the theory of Plate tectonics. According to this theory entire earth is divided into 14 major lithospheric (i.e. consisting of crust and upper mantle of thickness 100–150 km) plates. These plates are in continuous motion relative to each other. Three kinds of plate motion are well established in this theory; first two plates converge each other, second two plates move apart and the third one is moving plate slide past each other. Because of these motions 90 % of the total earthquakes, termed as tectonic earthquake, occur in the plate boundary zones globally. The genesis of earthquake in this region is well explained by the ‘mid-crustal ramp model’ of Pandey et al. (1995). According to this model crustal ramp along the MHT that marks the boundary between colliding Indian and Eurasian plates is the main geological structure responsible for the accumulation of elastic strain. The frontal part of the ramp structure, i.e. southern flat, is locked during the inter-seismic period, which acts as a geometrical asperity and accumulates elastic stress (Fig. 5.2). The accumulated stress is intermittently released in the form of micro-seismicity. This phenomena has been well approved by the intense microseismicity and moderate earthquake recorded throughout the Himalaya that cluster along the foothills of the Higher Himalaya (Fig. 5.3). However, the great earthquakes generally occur along the locked portion of the décollement beneath the Siwalik and Lesser Himalaya and the induced deformation is propagated towards the southern part of the Himalaya along the HFT, a southern expression of the MHT.

Fig. 5.3
figure 3

Seismicity in the Himalayas of Nepal (after Jouanne et al. 2004). The intense microseismicity (monitored between 1985 and 1998) drawn with small grey circles, tend to cluster south of the Higher Himalayas (Pandey et al. 1999) at a mid-crustal level. Star represents medium size earthquake

Full size image

5.4 Seismicity in Nepal Himalaya

The Himalaya was formed by the subduction followed by collision between the Indian and Eurasian plates at the rate of 35–38 mm/year towards N-NE (Chen et al. 2000; Holt et al. 2000; Paul et al. 2001; Wang et al. 2001; Sella et al. 2002; Jouanne et al. 2004). The convergence is highlighted through shortening across the Himalaya, Tibetan Plateau, and Tien Shan, and by Tibetan deformation through eastward movement of crustal material and southern rotation about the eastern syntaxis (Molnar and Lyon-Caen 1989; Wang et al. 2001; Zhang et al. 2004). Almost half portion of the convergence is absorbed across the Himalaya in straining the crust due to stress built over it as mentioned. The aggregated stress is, thus, responsible in generating large earthquakes in the region. Owing to the short instrumental record of the seismic events, the distribution of seismicity throughout the Himalaya appears non uniform although a major trend has been recognized. The general trend consists of a narrow belt of predominantly moderate sized earthquakes beneath the Lesser Himalaya just south of the Higher Himalayan front (Ni and Barazangi 1984) where all available fault-plane solutions indicate thrusting (Fig. 5.3). The great Himalayan earthquakes, however, occur along the basal décollement beneath the Siwalik and Lesser Himalaya. The focal depth for the Himalayan earthquakes varies from 10 to 20 km.

The historical records indicate that the Kathmandu valley, where the capital city is located, has experienced effects of large earthquakes in the past centuries. Major damage of probable seismic origin is reported to have occurred in 1255, 1408, 1681, 1803, 1810, 1833 and 1866 (Chitrakar and Pandey 1986). Some of these events might be related to the repetition in the past of the 1934 Bihar-Nepal earthquake. That might be the case in particular for the 1833 earthquake (Bilham 1995). Most of the others are probably related to smaller magnitude earthquakes that would have occurred close to the Kathmandu valley. The major historical earthquakes are listed in Table 5.1. National Seismological Centre (NSC) under the Department of Mines and Geology has been continuously monitoring the earthquake events since 1978. NSC documentation shows that between 1994 and 1999, the average frequency of earthquakes having magnitude between 2 and less than 5 was approximately 10 per day, magnitude 6 and less than 7 was 1 in per 6 years. Similarly, total number of earthquakes of magnitude 2 to less than 7 between 1994 and 1999 were approximately 700, 900, 1,500, 1,700, 2,200, and 1,600 respectively (Upreti 2001). Although the general trend shows that number of earthquakes is increasing in Nepal Himalaya, no very large earthquake seems to have struck Far-Western and Western Nepal over the past few centuries.

Table 5.1 Past earthquakes in Nepal Himalaya and associated damages (adopted from Dixit et al. 2013)
Full size table

Intense microseismicity and moderate earthquake events throughout the Nepal Himalaya cluster along the foothills of the Higher Himalaya (Pandey et al. 1995, 1999). It makes an E-W trending zone as shown in Fig. 5.3. In western Nepal it lies between 80.5 and 82.5°E whereas in central Nepal it is bounded between longitudes 82.5 and 86.5°E. The eastern Nepal cluster is characterized by higher level of events between 86.5 and 88.5°E (Fig. 5.3). The projection of microseismic events along the structural cross-section shows distinct clusters in Nepal Himalaya (Fig. 5.2). In central Nepal the cluster has a rounded form and is located in the vicinity of the flat-ramp transition of the MHT. The cluster in western Nepal shows an elongated form and is nearly horizontal.

5.5 Active Faults and Paleoseismological Evidences of the Great Earthquakes

5.5.1 Active Faults

Active faults, in this paper, are considered those faults, which have repeatedly moved during the recent geological time and have high probabilities of recurrent slip associated with earthquakes in future. The N-S relative motion of the plates has developed many mega-thrusts in the regional scale as well as numerous active faults at the local scale in the Himalaya (Nakata et al. 1990). These active faults are the direct evidences of recent crustal and sub-crustal movements due to ongoing continental collision. The identification, nature, location and recurrence slip history of these faults can have great contribution in the assessment of earthquake hazard. In this regard, previous studies based on the interpretation satellite images provided important information for the Himalaya-Tibet region (Molnar and Tapponnier 1975). However, the distribution of active faults in relation to ongoing active tectonics is still the major topic of debate (Ni and Barazangi 1984; Nakata and Kumahara 2002).

In the regional scale, a number of studies have been carried out in the Himalaya and adjacent areas (Nakata 1982; Nakata et al. 1984; Nakata 1989; Nakata et al. 1990; Nakata and Kumahara 2002). Nakata (1982) grouped active faults of the Himalayas as the several distinct fault systems namely Main Central Active Fault System, Main Boundary Active Fault System, active faults of the Lower Himalaya and Main Frontal Active Fault System. Among these, active faults along the MBT and MFT are most active and have potential to produce large earthquakes in the future (Lave and Avouac 2000; Chamlagain et al. 2000). The Himalayan front in the western Nepal is characterized by several discontinuous segments of the MFT and its subsidiary faults (Nakata et al. 1984) (Fig. 5.4). In the west between the Karnali and Babai rivers, sineous trace of the fault is recognized for about 20 km by a series fault scarplets dislocating the alluvial fan surfaces. These features are characteristics of the reverse faults in the convergent tectonic environments. Similarly, several discontinuous traces of active faults are also found along the Central Churia Thrust and the MFT that were reactivated several times in the past. A common tectonic feature is a pressure ridge and the apparent vertical slip along the fault is down to the north. Active faults along MBT, though considered to be basically reverse faults, do not always have a displacement of the same nature in this area (Nakata et al. 1984) that is active faults close to the MBT are of normal type (Mugnier et al. 1994).

Fig. 5.4
figure 4

Distribution of the active faults in and around Nepal Himalaya. Thick lines without tick marks show newly found active faults. Arrow indicates the direction of strike slip. Down-thrown side is shown by tick marks (after Nakata and Kumahara 2002)

Full size image

In central Nepal, traces of the active faults are continuous, especially in Hetaunda and south of the Kathmandu valley around the Bagmati River (Chamlagain et al. 2000; Kumahara et al. 2001). Active faults along the HFT also show northward downthrow. Active faulting along the MBT is represented by the Arung Khola Fault, the Hetauda Fault and Udaipur Fault in the east. The traces of these faults are continuous and north-facing scarplets suggest northward dipping fault planes (Fig. 5.4). The Himalayan front in the eastern Nepal is characterized by tectonic landforms produced by active faulting along both the MBT and MFT. In eastern Nepal, active faults trending NW-SE along the MBT merge with the active faults striking E-W along the MFT and extending farther south into the Gangetic plain (Fig. 5.4). These active faults exhibit downthrows to the north. The characteristics of active faults along Himalayan front vary from place to place in the Nepal Himalaya and are attributed to the difference in volume of the Siwaliks resulting from the subsurface conditions of the collision. The distribution of the active faults in the Himalayas, their recent movements and rapidly growing unplanned urbanization have clearly indicated that major cities located near these fault lines are much prone to earthquake disaster.

5.5.2 Paleoseimological Evidences of Great Earthquakes

Paleoseimology is defined as the study of the prehistoric earthquakes, particularly their location, timing and size. It provides geological evidences created during the paleoearthquakes and permits study of the distribution of the individual paleoearthquakes in space and over time periods of thousands or tens of thousands years. Such a long paleoseismic histories, in turn help us to understand many aspects of neotectonic deformation and seismogenic behavior of specific faults or regions. And it is equally important to assess the probability and severity of future earthquake, generally understand from the recurrence interval of the prehistoric earthquakes. Paleoseismic data are important because they supplement historic and instrumental records of the seismicity by characterizing and dating of large prehistoric events, which are very much useful for precise seismic hazard assessment in both regional and site specific seismic hazard.

Although numbers of pre-historical and historical earthquakes have been found in the Himalaya (Fig. 5.5), there are limited works to understand the prehistoric deformation and recurrence interval of these great earthquakes. Along the Himalayan range, the seismic behavior, characterization and paleoseismological reconstruction of the MFT is key to understand seismic risk. Some remarkable works, in this regard, have been carried out in the Himalaya identifying location, size and ruptured areas of the paleoearthquakes (Fig. 5.5). In Nepal Himalaya, there are limited studies on paleoseismicity along the active faults associated with the MFT. Lave et al. (2005) have discovered evidences of great medieval earthquake (1100 ad) in Nepal Himalaya. The trench survey and other paleoseismological evidences have revealed an earthquake of Mw 8.8 had jolted the region with surface displacement of about 17 (+7 − 3) m. The total length of surface rupture is estimated to be 240 km.

Fig. 5.5
figure 5

Historical earthquakes and estimated rupture area along the Himalaya (adopted from Avouac 2007)

Full size image

Despite the relatively recent occurrence of earthquake around magnitude 8, many researches claimed that these great earthquakes did not produce surface rupture along the MFT zone and assumed as a blind seismogenic thrust (Lave et al. 2005). Recently, Sapkota et al. (2012) using geomorphological mapping of fluvial deposits, palaeo-seismological logging of river-cut cliffs and trench walls, and modelling of calibrated 14C ages found at least 150 km long traces of the rupture of the MFT fault in Nepal due to 15 January 1934 Bihar-Nepal earthquake. Further, they also claimed to have found surface rupture of 7 June ad 1255 earthquake that severely damaged Kathmandu Valley killing King Abhaya Malla. This study concluded that in the past 1,000 years, two great earthquakes, 679 years apart, rather than one giant earthquake of eleventh-century jolted eastern Nepal and all the earthquakes had not occurred along the blind thrust. Although the data on historical earthquakes have shown that Nepal Himalaya has witnessed great earthquakes in the past, however, the recurrence interval for such great earthquakes has not been understood yet.

5.6 Earthquake Disasters in the Himalaya

Because of its location in the active plate boundary zone, the Himalayan region is considered as one of the hot spots for earthquake disaster. Over the last century, the Himalayan arc has been struck by four major earthquakes (Fig. 5.5) with magnitude ca. 8.5 in 1897 (Shilong earthquake) 1905 (Kangra earthquake) 1934 (Bihar-Nepal earthquake) 1950 (Assam earthquake) (e.g., Seeber and Armbruster 1981) and associated other destructive events killed thousands of people perishing hard-won economy of the Hindu-Kush Himalayan region (Table 5.2). The 1905 Kangra earthquake produced severe damage in the Kangra area and, about 100 km to the east, in the Dehra Dun area. The estimated rupture is about 280 km long segment, from Kangra to Dehra Dun, that must have extended eastward to about 78°E, near the border with Nepal (Chander 1988; Yeats and Lillie 1991; Gahalaut and Chander 1997). The 1934 Bihar-Nepal earthquake was believed to rupture a 200–300 km long segment to the east of Kathmandu (Pandey and Molnar 1988). In 2005, western Himalaya was hit by an earthquake with magnitude 7.4 killing more than 74,000 people from Pakistan and India. These major disasters in the region are briefly described below to provide insights on scientific facts, death tolls and damage.

Table 5.2 Earthquake fatalities in the Hindu-Kush Himalayan region till 2013
Full size table

5.6.1 1897 Shillong Earthquake

The remarkably documented June 12, 1897 earthquake (Ms = 8.0) occurred in the Shillong Plateau of north-eastern India, which is only high ground between the Himalaya and Bay of Bengal. The earthquake rupture was of 550 km long with east-west strike (Seeber and Armbruster 1981) and 300 km wide rupture leading a strong evidence for a shallow dipping, detachment like fault source. A 17,000 km2 rupture was predicted to be occurred in predominant thrust fault dipping north at about 5° having a depth of about 15 km and 23 km below the southern and northern margin of the rupture zone respectively (Gahalaut and Chander 1997). It was found that the northern edge of Shillong plateau rose violently more than 11 m during rupture of a buried, 110 km long, reverse fault dipping steeply away from the Himalaya thereby resulting in the destruction of structures in the vicinity. The earthquake induced liquefaction and the large magnitude devastated the surrounding areas. About 1,500 people were believed to loss their life during the earthquake.

5.6.2 1905 Kangra Earthquake

On 4 April 1905 Kangra earthquake was felt on the foothills of northwest Himalaya at 33°N and 76°E. The highest intensity of X was observed in Kangra and Dharmashala for Rosi-Forel scale. The magnitude of the earthquake was estimated to be M = 8.4. As many as seven foreshocks were felt before the mainshock which cost 19,000 lives and 10,000 buildings. The distribution of foreshocks of this event has suggested that large rupture is comparable to other great events of Himalaya except 1,897 earthquake. Several different possible rupture zones are found to be correlated with observations; the first is rupture occurred beneath the area delimited by intensity VIII isoseismal, a zone of 100 km in length surrounding Kangra and Dharmashala, the second one is the epicentral intensity distribution shows the maximum intensities greater than or equal to VIII around two regions Kangra and Dehradun, which are separated by about 200 km, hence rupture was said to be 280 km long fault zone along NW-SE trending boundary of Himalaya.

5.6.3 1934 Bihar-Nepal Earthquake

The 1934 Bihar-Nepal earthquake with Ms = 8.3 is the most recent mega disaster in the Himalaya. It struck the Himalaya at 14:24:22 (local time) on 15 January, 1934. The earthquake did not result from movement along the MBT since no coseismic surface rupture was observed on this fault (Dunn et al. 1939), moreover the intensity X is closely associated with a ‘slump belt’ and a zone of soil liquefaction. The meizoseismal zone lies primarily on south of MBT and the region of intensity greater or equal to VIII extends about 300 km along the strike of the Himalaya and 250 km perpendicular to the strike (Seeber et al. 1981). The Lesser Himalaya, Sub-Himalaya, and the foredeep have been covered by this event in transverse direction. Based on the intensities and damages, the epicenter was located in the Lesser Himalaya east of Kathmandu at 27.6°N and 87.1°E (Pandey and Molnar 1988). Macroseismic intensities and subsidence of foreland revealed from leveling data suggest that the earthquake ruptured a 250–300 km along-strike segment of the arc (Bilham et al. 1998). Three distinct areas of high intensity have been discovered i.e. large alluvial plains associated with ground failure, liquefaction, and slumping, narrow belt of suffering from high acceleration and the Kathmandu valley (Dunn et al. 1939). The 1833 event might have ruptured about the same arc segment as the 1934 earthquake (Bilham 1995). Although epicenter was about 300 km north east of the Kathmandu valley, because of amplified waves, massive damage was observed in the southeastern part of the valley (Rana 1935).

5.6.4 1950 Assam-Tibet Earthquake

The 15 August, 1950 Assam earthquake is the most recent great Himalayan Earthquake located in the extreme eastern and most remote portion of Himalayan front although epicenter of the earthquake probably lie in China. Some part of rupture zone underlies Himalaya as per the indication of numerous aftershocks beneath Himalaya in eastern Assam (Chen and Molnar 1977). The earthquake occurred at 28.38°N and 96.76°E having surface wave magnitude of Ms = 8.4 and moment magnitude Mw = 8.6, (Kanamori 1977). On the basis of instrumental data, epicenter was located beyond the surface termination of Himalayan arc, in the Mishi Mountains that bound the Assam basin towards the east-northeast and trend northwest (Seeber and Armbruster 1981). The relocation study of aftershocks of 1950 earthquake by Molnar and Pandey (1989) confirms that all the aftershocks lie beneath the Himalaya in a zone extending about 250 ± 50 km west of the epicenter of the mainshock in east-west direction with 100 km width in north-south direction and thus the rupture occurred on a gently NNE dipping thrust fault. About 1,530 people were killed by this earthquake.

5.6.5 2005 Pakistan Earthquake

On October 8, 2005, at 8:50 AM, a magnitude Mw = 7.6 earthquake struck the Himalayan region of northern Pakistan and Kashmir. The epicenter was located approximately 19 km north northeast of Muzaffarabad, the capital of Pakistan-administered part of Kashmir, known as Azad Jammu Kashmir (AJK). Pakistani government’s official toll as of November 2005 stood at 87,350, though the death toll could have reached up to 100,000. About 138,000 people were injured and over 3.5 million rendered homeless. As many as 780,000 buildings were either damaged beyond repair or destroyed and many more were rendered unusable for extended periods of time. Massive landslide was particular feature of this event, a very dense, high frequency band of landslides was triggered along the fault rupture trace in the mid slope areas; however, it quickly dissipated with distance away from the fault rupture zone. Almost all landslides were shallow, disaggregated slides, with two of them larger than 0.1 km2. The earthquake occurred within the Hazara-Kashmir syntaxis of the Himalayan fold belt. The main identified feature in this zone is the Balakot-Bagh fault (Hussain 2005), which is likely source of the earthquake. The reported focal depth for this event ranges from 13 km (MSSP) to 20 km (USGS) to 26 km (IGS). Beyond the narrow width of rupture zone, the signs of damage appeared to be fairly minor. The remarkable damage in more distant locations such as Abbotabad (35 km from rupture zone), Islamabad (64 km), and Lahore (>250 km), delineate the effect of local site or poor construction rather than direct intense shaking from earthquake. Muzaffarabad suffered great damage (IX-X on MMI scale), and the city of Balakot was almost totally destroyed (X on MMI scale).

5.6.6 2011 Sikkim-Nepal Border Earthquake

A magnitude 6.9 earthquake occurred in Sikkim-Nepal border on September 18, 2011 at 6:25 PM (Nepal standard time). The epicentre (27.72°N, 88.06°E) of the earthquake was 272 km east of Kathmandu, while the focal depth was 19.7 km. The earthquake was complex due to two events occurring within a short interval of time. The moderate magnitude earthquake triggered a large number of landslides causing massive damage infrastructures. The severe effect was in the Indian state of Sikkim and neighboring West Bengal, Bihar from India and Nepal, Bangladesh, Bhutan, and Tibet (China) were also affected in some extent. The maximum shaking intensity is estimated to be around VI + in MSK scale (EERI 2012). Series of aftershocks were recorded; however, two of the significant were of M4.5 and M5.0 within 75 min. Landslides, rock falls, and mudslides were responsible for loss of lives and damage to infrastructures, as well as associated economic losses. The Buddhist monasteries and temples were extensively damaged during this event; constructed with random rubble masonry with mud mortar. Most of the multi-storied reinforced concrete (RC) buildings were non-engineered and sustained considerable damage during this event. Sixty people died in Sikkim and another eighteen people died in West Bengal and Bihar.

The location of ruptured areas produced by these earthquakes shows several seismic gaps along the mountain range. The major gap has been identified between the location of the Kangra (1905) and Bihar-Nepal earthquakes (1934). It is believed that this region has not experienced such an earthquake since the last great earthquake. This portion of the arc thus stands as a “central seismic gap” (Fig. 5.5), a potential location for the next large earthquake, unless the seismotectonic behavior of that area is very different from that of the rest of the Himalayan arc. In such a large to great overdue earthquake event, the rapidly urbanizing intermontane valley that filled up with hundreds of meter soft sediments in the Himalaya, may amplify shock wave’s energy bringing massive devastation in the region even to southern megacities located nearby active fault lines. In the following section of the chapter, therefore, seismic site effects assessment in a part of Kathmandu valley is presented aiming to provide general scenario of ground amplification during the great earthquakes. Since the Kathmandu valley shares similar tectonic as well as geological features to that of the other valleys in the Himalaya, the results shown here may be significant to understand seismic hazard in the Himalayan intermontane valleys.

5.7 Seismic Site Effects in Kathmandu Valley

5.7.1 Basic Concept of Site Effects

Local sub-surface geology, geomorphology and the geotechnical characteristics of the soil strata have a strong influence on seismic ground motion. After the 1906 San Francisco, U.S.A. and 1923 Kanto earthquakes, Japan, it has been widely perceived that sub-surface geology and earthquake damage has strong correlation. These observations have led to the development of new discipline of science “seismic site effects” in Geotechnical Earthquake Engineering since 1930. The term “seismic site effects” has been used to describe the influence of local geology on the strong ground motion and are widely recognized as an important factor of earthquake risk. It has also become one of the major issues in the field of earthquake engineering because several observations made after the damaging earthquakes (e.g. Nigata and San Francisco 1964; Irpinia 1980; Città Del Messico 1985; Kobe 1995 etc.) revealed that the local amplification of ground motion has great influence on non-uniform damage pattern, particularly in soft alluvial valley.

Generally the site effects measure the amplification of the seismic wave energy when propagate various layers having different impedance contrast. The recording earthquake motion at any site generally provides information on source activation (rupturing of faults), propagation path of seismic energy and effect of local geology on the wave-field at the recording site as shown in Fig. 5.6 with illustration of wave propagation from the fault to the ground surface. The physical meaning of the site response can be explained using the physical amplitude r(t), which represents acceleration, velocity or displacement that is recorded at a site and is written as

$$ r(t)=e(t)*p(t)*s(t) $$
(5.1)

where e(t) is the source signal, p(t) is the function that characterize the wave propagation from source to site and s(t) represents the effect of local site condition on ground motion and is simply termed as site effects. The Eq. (5.1) can be written in frequency domain as

Fig. 5.6
figure 6

Schematic illustration of the wave propagation from fault to ground surface

Full size image
$$ R\left(\kern0.15em ,f\right)=E\left(\kern0.15em ,f\right).P\left(\kern0.15em ,f\right).S\left(\kern0.15em ,f\right) $$
(5.2)

The parameters in Eq. (5.2) are the Fourier transform of the time dependent functions of the Eq. (5.1). Thus the site effects, physically, is a combined effects of the source signal, wave propagation, and interaction of wave field on soil condition.

The site effects usually provide the effect of local geology in the modulation of seismic wave at each and every recording site where the local geology is composed of soft sediments and underlying bedrock. The main factors that affect the site effects are soil thickness, lateral discontinuities like faults, fracture, lineaments etc dynamic and physical properties of the soil and underlying bed rocks with topography. Since the soils are the result of extensive processes of weathering, erosion and deposition because of which they generally acquire different physical properties due to which they interact differently with the seismic shaking resulting into different level of amplification. The topography of the site has also significant effect on site response. For example, in case of convex topography, higher amplification is observed at the crest as compared to the foot but in concave topography amplification varies at the lateral parts. The effect of local geology on ground motion depends on frequency, intensity, and angle of incidence of incoming wave. Therefore, site effects are a result of various factors that affect alone or sometime in combination with other parameters. Since the study is based on 1D model, the effects of valley geometry, and topography are not considered here.

5.7.2 Geology of Kathmandu Valley

Geologically, Kathmandu valley lies in the Lesser Himalayan zone. It is an intermontane basin with young fluvio-lacustrine sediments (Pliocene to Quaternary age) up to 500 m thick (Yoshida and Igarashi 1984). The surrounding mountains are the Shivapuri in the north and Phulchoki in the south. The valley periphery is characterized by outcropping bedrock whereas the central portion of valley has gotten the thick fluvio-lacustrine unconsolidated sediments; moreover the distribution of sediments across Kathmandu valley is not uniform (Fig. 5.7a). The basement rock in Kathmandu valley consists of Phulchoki Group and Bhimphedi Group of the Kathmandu Complex (Stocklin and Bhattarai 1977). The southern part of the valley consists of hill terraces formed during late Pliocene to middle Pleistocene (Yoshida and Igarashi 1984), with the sediment exposure along the terraces. The southern part is formed from the Tarebhir Formation, Lukundol Formation, and Itaiti Formation (Sakai 2001). The central part of the valley consists of Bagmati Formation, Kalimati Formation, and Patan Formation. The Bagmati Formation was active before the lake formation in valley and is taken to be responsible for deposition of sediments in most part of the valley. The black clayey central portion is called the Kalimati Formation, with dark grey carbonaceous and diatomaceous beds of the open lacustrine facies (Sakai 2001). The Patan Formation is distributed in and around Kathmandu and Patan city, consisting of fine to medium sand and silt intercalated with clay and fine gravels in some places. The northern and northeastern part of Kathmandu valley consists of fluvio-deltaic or fluvio-lacustrine origin mostly sandy facies called as Gokarna Formation and Thimi Formation (Yoshida and Igarashi 1984; Sakai 2001) (Fig. 5.7b). The Kathmandu valley rocks are intersected by a number of faults systems. The Chandragiri fault and the Chovar fault acting on southern part of Kathmandu valley are considered to be active faults cutting the colluvial slopes and the terraces of the late Pleistocene age (Sakai 2001).

Fig. 5.7
figure 7

(a) Geological map of the Kathmandu valley (b) Schematic geological cross-section along N-S (after Sakai 2001)

Full size image

5.7.3 Geotechnical Characterization of Kathmandu Valley

Geotechnically, Kathmandu valley is characterized by the soft fluvio-lacustrine soft sediments with thickness variation from place to place. Based on the gravity measurements in the Kathmandu valley, Moribayashi and Maruo (1980) estimated the maximum thickness of 650 m. In addition, several drilling data have revealed that more than 300 m thick sequence of muddy and sandy sequence (Katel et al. 1996) in the valley. At the central part of the Kathmandu City, Bhrikutimandap, a drillhole reaches the basement rock at a depth of 550 m. Katel et al. (1996) compiled the drillhole data and presented fence diagram that clearly shows the extensive distribution of the black clay soils rich in organic content. In this study, we have considered 49 boreholes maximum 30 m thick engineering soils below which model considers engineering bed rocks. Based on the observed shear wave velocity in some bore holes and N-values of Standard Penetration Test, the average shear wave velocity has been found to be varying from 148 to 297 m/s. A representative borehole log with shear wave velocity is shown in Fig. 5.8. The low plasticity silt with some portion of gravel in underlying layer is found to constitute the PI as 15. However, the plasticity of 23 is found to be occurring in the medium clay slit. As per the grain size distribution in the study area clays are dominant in most of the sites and a small portion of sandy facies is present in the northern part. As there is no data on the dynamic properties of the soil of Kathmandu valley, for the modelling purpose, we used experimental soil curves given by Vucetic and Dobry (1991) for different PI values.

Fig. 5.8
figure 8

Representative borehole log with shear wave velocity

Full size image

5.7.4 Methodology

5.7.4.1 EERA

Site effect basically relies on the geotechnical and geological databases; the sufficient information is to be collected for the reliable results. A number of softwares are available for the ground response analysis. The SHAKE (Schnabel et al. 1972) was for the first time introduced to incorporate the site effect due to the fact that the ground motions in those sites of layered soil deposits were found to larger than predicted ones (Seed and Idriss 1970). It is subjected for computing response in horizontally layered soil-rock encountered with transient and vertically travelling shear waves and is primarily based upon the wave propagation solution of Kanai (1951), Roesset and Whitman (1969), and Tsai and Housner (1970). SHAKE considers that cyclic behavior could be simulated with an equivalent linear model, and is modified many on occasions (e.g. Frequency-dependent equivalent strain) as per the features and convenience. In 1998, Equivalent Linear Earthquake Site Response Analyses (EERA) was developed in FORTRAN 90 with the same basic concept as SHAKE; EERA is based on the modern implementation of established concepts of equivalent linear earthquake site response analysis and moreover its advantage hinges with the dynamic array dimensioning and matrix operations in FORTRAN 90 (Bardet et al. 2000). EERA is, therefore, used for assessing the stratigraphic site response analysis because of its user friendliness and convenience over calculations.

5.7.4.2 Input Motion

The Uttarkashi earthquake (M w  = 6.9) of 20 October 1991 is used as the input motion in this study. The acceleration time history (Fig. 5.9a) was recorded at the epicentral distance of 34 km at Uttarkashi station with the maximum acceleration of 0.31 g at 0.72 s. It is chosen as the input motion due to the fact that it was measured on bedrock, and there is no good data base of strong ground motion in Nepal. Moreover, Uttarkashi earthquake is consistent with the trend of earthquakes in Nepal Himalaya i.e. thrust type. The response spectra of the input motion shows peak spectral acceleration of 1.2984 g at 0.26 s (Fig. 5.9b).

Fig. 5.9
figure 9

(a) Acceleration time history (b) Response spectra of the Uttarkashi earthquake

Full size image

5.7.5 Results

The equivalent linear site response analyses of fluvio-lacustrine deposits of Kathmandu valley has been performed by analyzing 49 borehole logs across the valley. The study shows interesting results that are consistent with the observed damages in the past great earthquakes. The maximum peak spectral acceleration has been obtained between 1.2725 and 1.2826 g in Thamel, Swoyambhu, Boudha, south of Chabhil, Nakhhu and south of Khumaltar (Fig. 5.10). The range of the amplification factor is found to be varying from 1.980 (Sundhara) to 7.788 (Thamel) (Fig. 5.11). The predominant period in the study area has been computed in the range of 0.27 (e.g. Boudhha, Dhapasi, Dhumbarahi) to 0.61 s (northeast of Thamel and New Baneshwor) (Fig. 5.12).

Fig. 5.10
figure 10

Peak spectral acceleration (in g) map of the study area

Full size image
Fig. 5.11
figure 11

Soil Amplification map of the study area

Full size image
Fig. 5.12
figure 12

Predominant period (in second) map of the study area

Full size image

The bedrock exposure in Swoyambhu area has led to be lowered damage during 1934 earthquake, in the same way; the amplification factor of 3.789 estimated in this research seems to be comparatively lower than other fluvio-lacustrine deposit sites. The Thamel site has been estimated to be the maximum amplified site in this study, with the amplification factor of 7.788 at the frequency of 2.6 Hz. The maximum peak spectral acceleration has been estimated to be 1.2826 g. In this area, the boreholes incorporate the deposition of low plasticity silt. The average shear wave velocity in Thamel area has been estimated to be about 222 m/s. The predominant period is at the range of 0.32 s. Those sites with higher peak spectral acceleration are also experiencing the higher peak spectral velocity.

Moreover, the boreholes from Bouddha area has shown the amplification factor of 5.896 at the frequency of 2.4 Hz, with the maximum spectral acceleration of 1.2826 g at predominant period of 0.27 s, where average shear wave velocity is about 265 m/s. The soil configuration in Bouddha area is similar to that of the Thamel, except that traces of pebbles are observed at higher depth. The Chabhil site has been estimated to be 5.59 times amplified at the frequency of 2.0 Hz. The predominant period has been estimated to be 0.30 s with peak spectral acceleration 1.2826 g. The shear wave velocity is computed as 297 m/s. The similar case of higher amplification factor has been estimated for Maharajgunj too. The amplification factor is found to be around 7 at the frequency of 2.4 Hz. The predominant period and peak spectral acceleration are computed as 0.35 s and 1.2743 g respectively. The top layer of the soil has been covered with the filling material and the underlying layers are abundantly composed of silt and sandy facies with low to medium plasticity. The lower strata are composed of medium to fine sand with traces of pebbles. The average shear wave velocity has been estimated to be 204 m/s for this. Such higher amplification factor and the low predominant period have delineated the susceptibility of higher damage during earthquakes.

The Dhapasi and Dhumbarahi site (east of Maharjgunj) have the amplification factor of about 6 at the frequency of 2.4 Hz. The peak spectral acceleration 1.28 g is computed at the predominant period of 0.27 s and the shear wave velocity is about 265 m/s respectively. These sites have medium to coarse sand with pebbles and gravel underlain by the silty sand and fine sand. Similarly, the amplification factor for the Gyaneshwor site, south of Chabhil, has been estimated to be about 6 at the frequency of 2.6 Hz. The peak spectral acceleration, predominant period, and the average shear wave velocity are 1.2826 g, 0.30 s, and 239.00 m/s respectively. The area is mainly characterized by sandy silt underlain by clayey silt of medium plasticity.

The Jawalakhel area has been estimated to be amplified 5.47 times at the frequency of 2.00 Hz, with the peak spectral acceleration of 1.28 g at the predominant period of 0.37 s for average shear wave velocity 208 m/s. Lithologically, area is composed of medium to coarse sand with traces of pebbles, which is underlain by the soft clayey silt, subsequently the low plasticity silt is found in the geotechnical investigation. The Thapathali site, north of Jawalakhel, has been estimated as the amplification factor of 6.190 at the frequency of 1.60 Hz. The spectral acceleration of 1.2746 g has been obtained at 0.46 s and the average shear wave velocity for the area is 158 m/s. The dominance of sandy silt and clayey silt has led to higher amplification though the shear wave velocity is not much higher at Thapathali. However, the amplification factor for Dillibazzar area has been estimated to be around 3 at the frequency of 0.80 Hz.

The Sundhara site has given the least amplification factor of 1.908 at the frequency of 0.40 Hz. The average shear wave velocity is estimated to be 156 m/s. The low shear wave velocity may be inhibiting the lower amplification factor here but the peak spectral acceleration has been estimated to be the maximum as 1.2826 g. The predominant period for Sundhara site is estimated to be 0.58 s. The geotechnical investigation has furnished the information regarding the soil composition to be silty sand underlain by clayey silt for this site. The Balkumari, Kuleshwor, Swoyambhu, and Chhampee sites have the amplification factors lying between 3 and 4. Sandy silt overlain by clayey silt is found in these areas so less amplification is observed there. Balaju, Sukedhara, Sankhamul, Patandhoka, New Road, Manbhawan, and Gwarko sites have the amplification factors lying between 4 and 5. The silt ranging from low to medium plasticity is found through the geotechnical investigation. The average shear wave velocity for these sites has been estimated to be the average as compare to other sites so the average amplification factor has been obtained.

5.7.6 Discussions

5.7.6.1 Seismic Site Effects

An attempt has been made to evaluate ground response analysis of the soft fluvio-lacustrine deposits (e.g. clay silt and sand) of the Kathmandu valley, one of the large intermontane valley in the Himalaya. Although effect of valley size, geometry and effect lateral variation of soil strata are important factors in such analysis, we consider simple 1D model for the purpose because the ratio of depth to width of the valley is within the permissible limits. In the study area, it is found that local subsurface geology has strong influence on seismic site response. The maximum peak spectral acceleration (around 1.28 g) in the study area has been obtained in the sites having dominantly silt, e.g. Swoyambhu, Thamel, Boudha and northeast of Sundhara, Nakkhu and south of Khumaltar. The soils, in these sites, are characterized by either low plasticity silt, or even clayey silt. The minimum values are obtained for the sites having predominantly clay and silty clay, where the plasticity is found to be medium overcoming the low plasticity silt. Hence the inference could be drawn that, for the higher shear wave velocity, the silt (clay or sand mixed) with low to medium plasticity in the Kathmandu valley, could suffer from occurrence of high spectral acceleration (e.g. Kalimati, Maharajgunj etc). The difference between the maximum and minimum peak spectral acceleration is quite less, showing the severe situation in future earthquakes. It is observed that the sites with sandy silt have given the greater amplification ratio; moreover, those sites situated beside the natural water course and dominated by low plasticity silt are estimated to be amplified in greater extent, e.g., Thamel, Bodha, Chabhil, Tribhuvan International Airport area. As the entire site response is greatly characterized by the impedance contrast, along with the very small variation in density vertically, the dependence of site response towards the shear wave velocity is observed to be high. This has proven that the geotechnical aspect of soil to be significant during the large earthquake. Whenever the shear wave velocity is obtained to be smaller, the amplification is estimated to be lesser, and vice versa. The subdivision of soil profiles is directly dependent over the shear wave velocity, so the amplification has been affected by the shear wave velocity in greater extent. In general, higher the shear wave velocity, the higher is the amplification factor and vice versa for the Kathmandu valley.

5.7.6.2 Comparison with Past Earthquake

The 1934 Bihar-Nepal and 1988 Udayapur earthquakes were the major events that caused a huge damage in Nepal. As the 1988 earthquake did not severely affected Kathmandu valley, here we compare our results only with Bihar-Nepal earthquake, which killed at least 8,519 people in Nepal. Although epicenter was about 200 km north east of the Kathmandu valley, because of amplified waves, massive damage was observed in the southeastern part of the valley (Rana 1935). An intensity up to X was assigned for the soft sediments covered areas of the valley. According to Dunn et al. (1939) there were nearly 66,440 houses in 1920; 12,397 (nearly 19 %) houses were completely destroyed by the earthquake, and 25,658 (38 %) were badly fractured (Rana 1935). During this earthquake, the areas covered with the fluvio-lacustrine deposits have experienced the massive damages (Dunn et al. 1939). Particularly, southern part of the valley; e.g. Bhaktapur, Lubhu suffered massive damage whereas part of Kathmandu city (e.g. Thamel, Kalimati, Sundhara, Baneshwor) and Patan city had witnessed significant destruction (Rana 1935). However, the areas with bedrock exposure, e.g. Pashupati, Swoyambhu and Kirtipur were least affected. This damage pattern is consistent with the amplification factor, peak spectral acceleration, and predominant period obtained from this research. The highly amplified areas coincide with the massively damaged areas of 1934 earthquake. Further, lesser predominant period of the sites signify the vibration resonance with buildings, which is approximately equal to 0.1* No. of storey, so the low predominant period might be the possible cause of large destruction during the great earthquake of 1934. Thus, the damage pattern and computed results have clearly indicated that the amplification of the ground motion was the main culprit of the devastation during 1934 great earthquake.

5.7.7 Summary of Findings

Seismic site response analysis of densely populated fluvio-lacustrine deposit of Kathmandu valley is carried out using EERA. The study area covers the Kathmandu Metropolitan and Lalitpur Sub-Metropolitan cities. The earthquake ground motion of the Uttarkashi earthquake, a thrust type earthquake, has been used as input motion as the bedrock motion due to the fact that the similarity of the earthquake patterns in Nepal Himalaya too. The spatial variation of geotechnical database has been considered with great care and the site response across the soft soil deposit of these two cities has been estimated in terms of the peak response spectra, amplification factor, and predominant period. This research has estimated the variation of peak spectral acceleration from 1.2725 to 1.2826 g indicating a higher peak spectral acceleration throughout the study area. The maximum peak spectral acceleration in the study area has been obtained in the sites having silt dominance. The soils, in these sites, are characterized by either low plasticity silt, or even clayey silt. The key governing factors for the higher amplification factor are higher shear wave velocity and silty facies of clay. It is also observed that the sites with sandy silt and silty clay have given the greater amplification ratio; moreover, those sites situated beside the natural water course and dominated by low plasticity silt are estimated to be amplified in greater extent. The amplification factor and the peak spectral acceleration are found to be consistent with the damage pattern of 1934 Bihar-Nepal great earthquake, i.e. the severe damage wherever occurred was estimated to be highly amplified with the higher peak spectral acceleration. In the same way, the predominant period is found to be varying between 0.27 and 0.61 s. This predominant period could be implemented for the approximate analysis of soil structure resonance phenomena. The geotechnical earthquake engineering investigation in this research varies from 20 to 30 m depth, the boreholes reaching up to the bedrock may be used in future to better evaluate the ground response. The availability of database in Kathmandu valley is not sufficient to carry out 2D site response analyses, so the seismic arrays are to be designed and installed for precise assessment of seismic site effects adopting 2D approach.

5.8 Ways Forward in Reducing Earthquake Risk

The foregoing discussion on seismicity, active tectonics and seismicity of Nepal Himalaya has clearly shown that Nepal is one of the earthquake prone areas on the globe. Basically, the MHT, which is exposed to the frontal part of the Himalaya as a MFT, is generating major earthquakes posing threats to millions of people living in the northern tip of the Ganga basin. The above-mentioned seismic effects analysis in the Kathmandu valley has clearly shown that the soft sediments filled valley would be amplified during the moderate to great earthquake. The other soft alluvial valleys e.g. Intermontane as well as Dun valleys in the entire Himalayan range has high probability of damages as the major urban settlements are rapidly but rampantly increasing exacerbating the seismic risk due to flawed construction in the region. In the last two decades Nepal has shown good progress on seismotectonic research, however, paleoseismological study has not been carried out extensively. Therefore, new efforts should focus paleoseismological investigation to understand recurrence interval of devastating earthquakes, which are directly useful to hazard assessment. The efforts initiated by Nepal Government, NGO, INGO and CBO are encouraging but access to these programs are limited only to people of urban and semi urban areas (Chamlagain 2009). Therefore earthquake risk reduction programs should also be launched remote areas. For the better management of earthquake risk government policy and plan should spotlight the following:

  1. 1.

    Investment on research (e.g., seismotectonics and paleoseismology), education, training, and human resource development in the area of engineering seismology.

  2. 2.

    Development and deployment of earthquake early warning system in the earthquake prone urban areas particularly located in the intermontane and dun valleys.

  3. 3.

    Seismic site effects analysis and Microzonation studies of the urban areas in large scale and linking of these maps with the land use and development plan and activities.

  4. 4.

    Establishment of Earthquake Risk Evaluation Centre (EREC) of highly educated scholars aiming to entire work from earthquake early warning system to risk management phase.

  5. 5.

    Revision and strict implementation of existing National Building Code throughout the country.

  6. 6.

    Coordination among the financial and legal institution, insurance companies, disaster related organization to formulate disaster related insurance policy.

  7. 7.

    Promotion of public awareness and preparedness programs intensively throughout the country.

Conclusion

The entire Himalayan belt, formed due to collision between the Indian and Eurasian Plates, is located in the active plate boundary zone characterized by frequent seismic activities. The belt is segmented almost along the east-west direction by several active faults having potential of generating devastating earthquakes as they did in the past. Most of the megacities and rapidly urbanizing areas are either located in the intermontane basin filled with soft sediments or located nearby the active faults. And the burgeoning population in such urban areas has led to non-engineered construction practices against the earthquakes putting the lives, infrastructures and hard-won economy at high risk of earthquake. Further, because of local geological condition, basin morphology and geotechnical features of the valleys, the earthquake risk in the city like Kathmandu located in the intermontane basin are much more vulnerable in terms of earthquake disasters. This studies has shown that soil type, geotechnical characteristics and shear wave velocity across the soil valley is responsible for the higher amplification of the seismic wave in the impending great earthquake in the soft sediments filled Kathmandu valley. The obtained predominant periods could be implemented for the approximate analysis of soil structure resonance phenomena. As many urban areas of the intermontane valley in the Himalaya (e.g. Hetaunda, Chitwan, Pokhara, Dang, and Surkhet valley in Nepalese Himalaya; Dehra Dun, Kashmir, Pinjau, Karewa valley in Indian Himalaya) are located in the “Central Seismic Gap” and share comparable geo-tectonic features, seismic sequence, soil stratigraphy, geotechnical characteristics etc, similar study may be significant to assess the seismic hazard, vulnerability and risk. An integrated approach comprising of paleoseismological studies, seismic microzonation, deployment of earthquake early warning system, development and enforcement of site specific building code, insurance policy along with preparedness directed awareness programs could be key measures in reducing earthquake risk in the intermontane basins.