Here we give some examples of the work done in the project, please see the publications for more detail.

Earth structure and swarm activity in the Nordland region of Norway
A 3D tomographic velocity model was developed for the Nordland area in northern Norway based on a refined earthquake catalogue. The 3D model was used to relocate seismicity, but also for the ray tracing involved in fault plane determination. The models reveal a sharp transition from relatively thin crust beneath the southern Lofoten to a much thicker crust beneath the mainland. The thickness of crust south of the Lofoten has previously been disputed, our model shows that the crust indeed is thin here. On the mainland, we find anomalies that suggest fluid saturated fracture zones that collocate with earthquake swarms.

The earthquake swarms in the Nordland area were further investigated. Fault plane solutions showed that the earthquake activity is due to relatively shallow extension, which is possibly the result of glacial isostatic adjustment or sediment redistribution. Deep learning tools were applied to improve completeness of the earthquake catalogues and precise relative earthquake locations could be obtained. Significant differences were found between swarms in two adjacent areas, Jektvik and Rana. Most interestingly, the Jektvik activity is modulated by snow load in the winter. We find that activity occurs on specific fault segments that have similar orientation and migration occurs from central to outer segments. In contrast, the Rana activity does not appear to be modulated and occurs on spots, possibly fault irregularities or intersections, rather than elongated segments.

Long duration, non-volcanic and non-tectonic Palghar earthquake swarm in the stable continental region of India
The stable continental region of India plate is seismically more active than other similar regions on the Earth. It also experiences earthquake swarms, but not many of them have been monitored with a local dense seismic network. A very intense swarm of minor and micro earthquakes occurred in the Palghar region near the west coast of India, which is so far the best monitored earthquake swarm with a dense network in India. It started immediately after the seasonal monsoon in November 2018 as north south oriented clustered activity and is continuing even after three years (as in September 2021). The focal mechanisms of the earthquakes are predominantly normal on north-south oriented steep planes. In November 2019 another north-south oriented cluster of earthquakes initiated, about 5 km east and parallel to the first cluster.

The seismic moment release due to these earthquakes is strongly correlated with the seasonal monsoon rainfall with a delay of up to 3-5 months. Our analysis suggests that the Coulomb stress change caused by the earthquakes of the first cluster encouraged earthquake occurrence in the second cluster and pore pressure increase due to the seasonal rainfall promoted occurrence of earthquakes in both clusters. InSAR analysis reveals subsidence of ~3 cm between November 2018 to May 2019 in the swarm region. The earthquakes of the two clusters and their corresponding seismic moment is strongly correlated with the seasonal monsoon. We found that the contribution of pore pressure at depth due to rainfall is quite substantial. The occurrence of tightly clustered shallow focus earthquakes causing subsidence in the overall compressive regime of stable continental region of the Indian plate, possibly implies some shallow subsurface process of precipitated water migration leading to collapse of subsurface cavities and may not be linked with the tectonics of the region. We propose that aseismic slip driven by the fluid migration at shallow depth is responsible for the swarm.

 Earthquake swarm of Palghar from January 30, 2019 to October 31, 2020.

Rapid filling and longer retention of water during Pulichintala reservoir impoundment triggering earthquakes in the stable continental region of India
Pulichintala reservoir on Krishna river is located in the southeastern Indian shield on the northern extent of the Palnad sub-basin of Proterozoic Cuddapah Basin. First stage of reservoir impoundment started in 2013 and the second stage of impoundment started in 2019 when the height of the dam was increased.  Soon after the second stage of impoundment earthquakes started occurring which continued for 2 months with maximum magnitude as 4.7. Two distinct clusters with shallow depth of up to 5 km were identified which follow the faults/shear zone in the region (Figure 1). The estimated focal mechanisms with predominant strike slip motion on steep planes and the derived stress state are consistent with the regional stress and plate motion of the Indian plate. We analysed the influence of reservoir impoundment on the inferred seismogenic faults, derived for 4.7 magnitude earthquake, and found that the reservoir load and the induced pore pressure indeed promoted failure, thus triggering earthquakes. We ascribe the earthquake triggering to rapid filling, longer and higher retention of high-water level during the reservoir impoundment in 2019-2020 which caused stress change of 4-8 kPa which was at least two order magnitude higher than the typical tectonic loading rate of in the stable continental region of Indian Shield.

Figure Location of Pulichintala reservoir in the Cudappah basin, earthquakes and estimated focal mechanisms.

An Application Software using Green’s function based stress diffusion soluitons due to time varying finite surface loads for analysing the effect of reservoir impoundment on seismicity
In this research, several codes are written utilizing Green’s function solution of poroelastic equations and the frictional failure criterion in Matlab. All these codes are embedded into a single user-friendly software application program to analyze the fault stability caused by any surface reservoir. The application takes in the reservoir water-level time series data and reservoir parameters as input that are required for computing the induced stresses and stress-induced and diffusion pore pressure. These computations are done in just a single click using this application and can be applied to any surface reservoir to get the results that are essential in analysing the seismicity near the reservoir. The computation starts with running the main application program, which is designed and developed using the MATLAB App Designer wherein the main interface layer which is the main window pops up consisting of different panels with different programs running in the background where one can enter the required input, get the results in the form of induced stress and stress-induced and diffusion pore pressure and plot the graph. More information will be available on CSIR-NGRI website (www.ngri.org.in).