There is now hard evidence that impounding of water by even small dams of 1 sq km with just 10 metre variation in water level can cause deformation in the neighbouring region. Such deformation may add to the stress of a nearby fault, which if already critically stressed, can trigger an earthquake.
Even very small reservoirs (1 sq. km) such as Dharoi on Sabrmati River impounding water seasonally with water-level variations of just 10 metres can cause deformation in the neighbouring region. The vertical displacement by such small dam was as much 9.2 mm. Such deformation may add to the stress of a nearby fault, which if already critically stressed, can trigger an earthquake.
Based on data from Global Positioning System and satellite, a team of researchers led by Vineet Gahalaut of the National Centre for Seismology, Ministry of Earth Sciences, New Delhi found crustal deformation in all four reservoirs. The four reservoirs are: Koyna in Western Ghats, Tehri in Lesser Himalaya of Garhwal region on the confluence of Bhagarathi and Bhilangana Rivers, Ukai on Tapti River, and Dharoi in central part of India on Sabrmati River. These are of varying sizes and are located in different geological zones in India.
In two papers (here and here) published recently in the journal Bulletin of the Seismological Society of America, the researchers conclude that there is “no lower threshold on the size of the reservoir to cause deformation” in the surrounding area.
A 6.3 magnitude Koyna earthquake that struck in December 1967 killing about 180 people and injuring many occurred about five years after the dam was built. Innumerable smaller earthquakes occur each year in the region surrounding the Koyna dam. The researchers have been studying the Koyna dam for the last 51 months since the beginning of 2013 using five GPS located close to the seismic zone of the region.
“There are no reports of any earthquake from the other three reservoirs. But that does not mean there are no earthquakes. It is just that we don’t have adequate seismic monitoring network to collect data in these places,” says Dr. Gahalaut. “The results of our study underscore the need for a seismic network to monitor earthquakes in all reservoirs.”
“The strength of the fault reduces when the reservoirs are full. Impounding of water causes stress, and the water that percolates from the reservoir lubricates the fault thus reducing the frictional force leading to reduced strength of the fault or even aiding the fault to fail causing an earthquake,” he says.
In addition to deformation caused by seasonal loading of the reservoir, tectonic movement of crustal plates causes stress build-up. The fault fails causing an earthquake when the combined stress by the two factors crosses the critical point.
The researchers found that the eastern block of the Koyna-Warna fault zone is moving about 0.7 mm per year faster than the western block. Such anomalous motion of the blocks adds stress to the fault leading to earthquake. Differences in the rock composition could be causing the differential movement of the blocks. “We don’t know how much variation in rock composition is needed for the differential motion between the blocks. We need to study what is causing this differential motion,” Dr. Gahalaut says.
The team found seasonal variation in displacement in the case of Koyna-Warna seismic zone. The horizontal displacement was 0.5-2 mm and vertical was 9-15 mm. These displacements are more than the movement of the eastern block (about 0.7 mm per year). However, unlike the tectonic stress caused by plate movement, the displacement caused by seasonal reservoir loading weakens with distance from the dam.
According to the study, in the fault region at a depth of about 6 km the horizontal displacement is “insignificant” while the vertical displacement is less than 2 mm per year. The tectonic stress will more or less be the same at depths that can generate earthquakes.