Field Report on Liquefaction and Ground Deformation Survey

The MAEC team of C.P. and Kusala Rajendran and Martitia Tuttle with the assistance of Mahesh Thakkar of Lalan College in Bhuj recently conducted a liquefaction and ground deformation survey in the meizoseismal area of the Republic Day earthquake in Gujarat, India. We documented liquefaction features and surface deformation in the epicentral area and up to 50 km towards the south, west, and north, and 130 km towards the northwest near the Allah Bund. Liquefaction features include lateral spreads, sand blows, sand blow craters. In the epicentral area  between Choberi and Bachau, many water pipes and wells were broken, some due to lateral spreading with displacements ranging up to 1.3 m. Like other reconnaissance teams, we found no evidence of the primary surface rupture. However, we noted deformation, including buckling, flexuring, and folding, characteristic of compressional tectonics.

At a site north of Budharmora (N 23° 20.575' and E 70° 11.605') where the French team had excavated a deformed linear feature that was uplifted by 0.5 m (this may be the same structure identified by the OYO-RMS team as a possible surface rupture and interpreted by Seeber and the NSF-SCEC team as a slump), we surveyed the topographic profile across a disturbed zone, which is 140 m across and at least 400 long, and measured several small sand blows as well as displacements of a N-S oriented water pipe across the zone. The pre-event ground surface slopes about 1° towards the north. Ground deformation at the site includes 0.35-m and 1-m wide grabens, extensional cracks and back-rotated blocks, and sand blows up to 3.5 m long, 50 cm wide, and 4 cm thick in the upslope portion of the disturbed zone (Photo 1) and two uplifted linear features and related cracks in the downslope portion of the zone. Local residents described 1-m high water spouts issuing from the sand fissures following the earthquake. Due to the low slope, the disturbed zone is interpreted as a lateral spread characterized by extensional features at the head and compressional features at the toe. The water pipe, that crosses the lateral spread and is roughly parallel to the downslope and slip direction, is broken and separated laterally by 1.35 m in the scarp area and broken and overlapping both laterally and vertically by at least 0.8 m at the toe.

Photo 1. Upslope portion of lateral spread at Budharmora. Photograph by M. Tuttle.

At at village south of Manfara (N 23° 28.555' and E 70° 21.195'), a zone of N80°E oriented fissures and intervening conjugate cracks crosses a field east of the village and the tarred road to Chobari, and runs beneath a small building near the village well (Photo 2). The relationship of this zone to the "Manfara fault" as described by Seeber and the NSF-SCEC team is not known at this time. The ground surface which slopes about 1° toward the north-northwest is displaced downward by 10-15 cm on the north side of the cracks within the zone. The small house, like the blocks between the cracks, are back rotated (Photo 3). According to local residents, water and sand had vented in the field south of the zone where small sand blows are still present. We excavated a trench across the zone and found sand dikes filling the fissures and cracks below the surface but no obvious displacements in the rather homogeneous soil (Photo 4). Liquefaction was involved in ground failure at this site; however, the orientations of the fissures and cracks also suggest that N-S oriented compressive force played a role in the failure.

Photo 2. Zone of ground fissures and cracks at village south of Manfara. Photograph by M. Tuttle.

Photo 3. Ground fissures and back-tilted house at village south of Manfara. Photograph by M. Tuttle.

Photo 4. Sand dikes found filling fissures and cracks at village south of Manfara. Photograph by M. Tuttle.

North of Bharodia village (N23° 34.385 E 70° 24.792), we measured an uplifted zone that is 80 m long and 10 m wide and oriented N30°W. Crops on the portion of the agricultural field that was uplifted by about 70 cm are suffering due to the resultant change in the moisture regime of the soil (Photo 5). A zone of ground fissures extends at least another 2 km towards the northwest. Deformational features at this and other sites will be described in more detail in a later report.

Photo 5. Uplifted mound north of Bharodia village. Photograph by M. Tuttle.

Within 15 km of the epicenter, we observed sand blows that range up to 60 m long, 10 m wide, and 14 cm thick (Photo 6). From the area of wetted ground surrounding the sand blows, it appears that a large volume of water vented to the surface especially in the Rann north and west of Chobari. At 50 km south of the epicenter in the mud flats near Kandla Port, sand blows ranged up to 12 m long, 7 m wide, and 14 cm thick (Photo 7). Southwest of Khadir Island, about 45 km northwest of the epicenter, sand blows are up to 13 m long, 12 m wide, and 3 cm thick (Photo 8). We found no evidence of surface rupture or sand venting along a linear channel immediately south of Khadir Island that we had noticed on satellite images. About 48 km west of the epicenter in the Banni Plain near Lodai, sand blows range up to 15 m long, 12 m wide, and 5 cm thick. The thickest sand blow we observed was located near Umedpur about 45 km west-northwest of the epicenter (Photo 9). At this site, the largest sand blow was 33 m long, 32 m wide, and 26 cm thick and had a water-filled crater that was 10 m by 5 m (Photo 10). Other sand blows at the site measured 14.5 m by 11.5 m and 18.5 m by 8 m, in plan view. Two dry craters (2.4 m by 1.8 m and 1.6 m by 1.5 m) and associated broken ground and clast zone extending 26 m west of the craters suggest that gas release may have been involved in ground failure at this site. A few small sand blows occur along the Allah Bund about 100 km northwest of the epicenter. These features range up to 1.6 m long, 0.8 m wide, and 1.8 cm thick. Only vented water spots (no sand) from 5 cm to 10 m in diameter occur just south of the Allah Bund. However, another 30 km to the north, sand blows as large as 7 m long, 2 m wide, and 6 cm thick, as well as craters on the order of 1.5 m in diameter, are present.  In one of the craters, two lower sand layers could be traced around the crater walls and are interpreted as liquefaction features related to the 1819 Kutch earthquake and a previous event (reported by the Rajendrans in their earlier studies of the Allah Bund).  A black clay which is 3 m thick at the Allah Bund and thins towards the north may be controlling the distribution of liquefaction features in this area.

Photo 6. Moderate-size sand blow in Rann northwest of Chobari. Photograph by M. Thakkar.

Photo 7. Sand blow in mud flats used for salt production southwest of Kandla Port. Photograph by M. Tuttle.

Photo 8. Sand blow in Rann southwest of Khadir Island. New salt marks area of vented water following the earthquake. Photograph by M. Tuttle.

Photo 9. Excavation of 26 cm thick sand blow adjacent to large water-filled crater at Umedpur. Photograph by M. Tuttle

Photo 10. Sand blow and water-filled crater at Umedpur. Photograph by M. Tuttle.

In general, the liquefaction features in Gujarat appear to be much smaller than features in the New Madrid seismic zone.  For example, most of the Gujarat sand blows that we documented are less than 60 m long, 10 m wide, and 15 cm thick. In contrast, sand blows in the New Madrid region that formed during the 1811-1812 earthquakes are commonly 100 m long, 30 m wide, and 0.5-1 m thick.  Unlike sand blows induced by the recent earthquake in Gujarat, New Madrid sand blows are often composed of multiple sedimentary units ranging from about 30 to 60 cm in thickness, each unit related to a very large earthquake.  Feeder dikes of the Gujarat sand blows that we measured are 0.2-10 cm wide, with one dike ranging up to 25 cm. It is conceivable that a few sand dikes related to lateral spreading in the epicentral area (e.g., Budharmora) may be as wide as 1 m.  Feeder dikes of New Madrid sand blows are commonly 0.5-2 m wide and can range up to 10 m.  During the Republic Day earthquake, a number of sand blow craters formed.  A crater near Dharmsala is about 4 m in diameter (pers. comm., Arch Johnston) and several craters near the Pakistan border are about 1.5 m in diameter.  Sand blow craters such as these are not common in the New Madrid region but did form during the 1886 Charleston, South Carolina, earthquake of Mw 7.3 (Dutton, 1886).

Little is known about the liquefaction susceptibility of the subsurface sediments in the Gujarat region. However, Holocene floodplains, lacustrine, playa, and esturine deposits like those in the Gujarat region are thought to have a moderate likelihood of liquefaction (Youd and Perkins, 1978). Depth to the water table and cohesionless sediments in this arid region may affect liquefaction potential. A comparison of liquefaction induced by the Republic Day and New Madrid earthquakes would benefit from in situ geotechnical testing.

We plan to identify other, perhaps larger, liquefaction features in the meizoseismal area and to further map the distribution of liquefaction features using satellite imagery. There are reports of liquefaction in the Little Rann, south of Ahmadabad, and even in Hyderabad, Pakistan; however, these reports remain to be verified in the field. Identification of distal sites of liquefaction for the Mw 7.7 India earthquake will serve to calibrate magnitude-distance relations (Ambraseys, 1988) and liquefaction severity index (Youd and Perkins, 1987) for intraplate earthquakes.
Last modified: Mar 08, 2001 8039 visits (1 today, 1 this week, 59 this month, 888 this year)