EERI Special Earthquake Report -July 1999 34°NW -rr ~ the youngest, appears to be seis- Strike 282. micalfy more active. Several damag- 9. <> <><>19<65 <> Dip Slip 95. ing earthquakes have occurred along <>~angsa 33 p these thrust faults, and there are con- ~ Strike 97" tinuing debates on the current O 81. Dip seismogenic potential of these fault Slip 89. systems. The M6.5 Uttarkashi earth- 32 ~--O g>"V I Fau/t-P/aneSo/utlon (source: USGS) quake of 1991 I centered about 70 km northwest of Chamoli, is considered ~O I OU to be associated with this fault, as is 1991 O the Chamoli quake. Observations of 31 rkQshlo recent deformation in the epicentral 'Okhlmath 8 8 ~USGS Epicenter region also support this association. MBT TehrQ 0 .0 . o .IMD Epicenter o Dehra Dun .-~ <> An Active Fold? 30 1803 <> The Lesser Himalayan sequence ~~G lying between the MCT and the MBT OI\O~O \' shows stacking of various groups of 291 ...~- 01 75 76 77 78 79 80 81 82°E rocks characterized by south-vergent Figure 2: Spatial pattern of seismicity in Garhwal Himalaya during 1684- thrusts, which were later folded into 1985 with respect to two of the major thrusts MCT and MaT [Khattri et al., major scale synforms and antiforms. 1989, Proc. Indian Acad. Sci. (Earth. Planet. Sci.), 91-109]. The subsets The geologic maps of the area MCT -I, II, III are not marked. Shaded circle indicates the maximum felt area indicate the presence of an anticlinal structure very close to Chamoli. of the 1803 earthquake. Insert: Fault-plane solution of the main shock at During the postearthquake field Chamoli. investigation, signatures of recent deformation associated with this Kanpur (440 km southeast of the Geologic and Tectonic anticline were observed. A sharp epicenter), Shimla (220 km north- Setting contact of MCT I with recent deposits west) and Delhi (280 km southwest). The Himalayan mountain range, an is identified near Chamoli on the Maximum death and damage oc- outcome of the compressional pro- southern flanks of the anticline along curred in the district of Chamoli, cesses ensuing from the India-Asia the Alaknanda River. Thick deposits where about 63 persons died and collision (70-40 Ma), has been under- of colluvium (boulders and pebbles over 200 were injured. About 2,595 going extensive crustal shortening intercalated with coarse sand) occur houses collapsed and more than along the entire 2400-km-Iong north- at the foot of the steeper limb of the 10,850 were partially damaged. In ern edge of the Indian plate. A series fold. The colluvium may have been all, about 1,256 villages were af- of thrust planes is known to have remobilized on an incipient slope due fected. A few buildings at the distant formed as a result of these processes. to the development of the growing city of Delhi sustained nonstructural Three principal thrust planes in the fold. This contact is interpreted to be damage. No instances of liquefac- Himalayan region are the Main the surface expression of an active tion were reported. Linear cracks in Central Thrust (MCT), the Main fold. the ground were seen in some loca- Boundary Thrust (MBT) and the Main tions in the affected area. Frontal Thrust (MFT). Two of the major These observations are significant The earthquake was followed by thrusts and the regional seismicity are because the contact of the thrust intense aftershock activity; this in- shown in Figure 2. plane occurs close to the epicentral zone of the Chamoli earthquake. Al- Gluded at least three events of M >5 The MCT is believed to have resulted though the models for many earth- Most of the aftershocks were 10- from an intra-crustal thrust that quakes, including that at Uttarkashi, Gated to the east of Chamoli. The brought up the mid-crustallevel rocks suggest the rupture along MCT I , fault plane solution obtained from of the Higher Himalayan Crystallines geologic evidence for active faulting USGS (Figure 2, insert) indicates a to the Lesser Himalaya. Tectonically, it in this region is sparse. Observations pure thrust mechanism with two represents a ductile shear zone at in the epicentral region of the Cham- nodal planes striking at 282° and depth, comprising a duplex zone with oli earthquake may provide guide- 97°. The first one is preferred three distinct sub-thrusts: MCT I, MCT lines to identify active faults and folds because it conforms to the field II and MCT Ill, from south to north. Of in the Himalaya. observations. these, MCT I,the southernmost and 2 EERI Special Earthquake Report -July 1999 potential area for a great earthquake (Figure 1). An alternate view is that the strain in this stretch was partially or totally released by the moderate earthquakes in the past. In this con- text, the mechanisms of large earth- quakes in the region and their relation to strain accumulation on the MCT need to be understood in greater detail. Co-Seismic Phenomena The Chamoli earthquake was accom- panied by severe ground deforma- tions. Fissures, landslides and changes in the groundwater flow were Figure 3: Ground fissure at Telcom Hill near Gopeshwar reported from several locations. Well- h I developed ground cracks, trending Historic and Current muc arger. roug hly In. th e eas t -wes t d 'Irec t 'Ion an d Seismicity The limited recorded data from this showing lateral movement of up to region suggest a northwest-oriented about 12 cm, were observed in Go- Historic and instrumental data sug- zone of moderate activity (Figure 2). peshwar, Chamoli and Bairagna. gest fewer large earthquakes in the Two M >6 earthquakes within a span Attempts were made to trench across region compared to the rest of the of eight years, a larger event 196 some of these ground fissures at Himalaya (Figure 2). One earth- years ago, and the clustered micro- Telecom Hill, Gopeshwar (Figure 3), quake, probably of M > 7, is report- seismicity indicate strain accumulation but none of the trenches could be ex- ed to have occurred in this region in the region, The 700-km-Iong seis- tended downward more than a meter on September l' 1803. Several mic gap between the rupture zones of because of debris and boulders, In villages are reported to have been two great earthquakes (the 1905 one of these trenches, a poorly de- buried by the rockfalls and land- Kangra [M8.6] and the 1934 Bihar fined thrust plane was detected. Ori- slides caused by that earthquake. M8.4]), generally referred to as the I entation of these discontinuous The Badrinath temple located -40 "Central Gap," is considered to be a ground fissures conforms to the strike km north of Chamoli was severely I iO 8km damaged. Even though the earth- JPShi~ quake's location remains uncer- -Rock fall t Telecom Hill USG,:iCeneter tain, intensity reports suggest that the 1803 event may have been in D Okhimath . the same region affected by the 30~301 r O 0 current earthquake. Birahi ~ ;l/"1 c []J 0 sett!ements I The postearthquake survey of the Chamoli earthquake examines two ~- temples-one at Gopeshwar and 30"25' c MSK IntensitV1 the other at Makkumath, built dur- o V ~moli ing the 7 to12th centuries. These O \11 Qadi show evidence of severe damage O \111 c;;cr during the 1803 event. Many parts OWl/ D c c 3fJ'20' of these two temples have been Aftershocks a~dprayag I reconstructed, as indicated by the cM<3 I/MD EPiCe;e1 [J3~M<4 inscriptions on their wall stones. Gauchar """"""'- D' D4~M<5 ~ That these temples suffered only DM?5 / minor vertical cracks during the i,/ 30°15' current earthquake, in spite of 79"05' 79°15' 79°25' 79°35' being located in the meizoseismal Figure 4: Intensity variation during the main shock and location of aftershocks area, indicates that the magnitude in the affected area. Shaded portion shows the trend of the fault from the fault- of the 1803 event may have been plane solution, which is consistent with the damage distribution. 3 ~ EERI Special Earthquake Report -July 1999 - of the thrust front (MCT I). The higher elevation on the hill slopes. The well-developed. The Uttarkashi earth- majority of these east-west-oriented intensity VII observed at Makkumath, quake took place in October immedi- fissures, particularly those devel- located on a river terrace about 20 km ately after the monsoons, which lead oped in the well-consolidated debris, from the epicenter, is possibly another to a much higher incidence of slope may be a manifestation of the fault example of site amplification due to failure and foundation movement. movement. Ground cracks at several alluvial CoVer. Finally, construction practices in the places also developed as part of Chamoli area are much better than The area has a number of analog slope failure, and these pose threats those in Uttarkashi before 1991. to the down-slope settlements. strong-motion accelerographs oper- ated by the University of Roorkee. Cracks were seen in asphalt roads Building Behavior at some locations, indicating the Strong-motion records were obtained possibility of failure due to ground at Gopeshwar (9 km from the epicen- Many of the villages in the affected slippage. At several sites, large- ter), Joshimath (27 km), Okhimath (25 area are not connected by passable scale earthquake-induced landslides km) and Tehri (80 km). The peak roads and are accessible only after and rock falls were observed. Those ground accelerations in the two hori- considerable trekking. In addition to near Gopeshwar, Chamoli and Gadi zontal and the vertical directions at these villages.