Stress drops and scaled energies of micro-earthquakes at Koyna-Warna, India: Implications for self-similarity scaling

The scale-(in)dependency of source physics of micro -earthquakes with increasing seismic moment is one of the most debated fundamental question in the seismology. In this study, we have investigated the stress drops and scaled energies to see the existence of scale-(in)variance of source properties with increasing earthquake size. The analyzed 982 earthquakes (ML 0.5-5.1) are recorded by borehole and surface networks in the Reservoir Triggered Seismicity (RTS) site of Koyna-Warna, India. An empirical Green's function spectral ratio technique has been applied to eliminate the propagation effects, then stress drops are estimated using the Madariaga's model. The radiated seismic energies are estimated using the time -domain approach of velocity integral of the wave -train of an event. The stress drops and scaled energies are varying from similar to 0.01 to 16 MPa and 10(-7) to 10(-5) respectively. Their median trend shows a transitional characteristic at the seismic moment of similar to 1013 Nm. For larger earthquakes (Mw > similar to 2.5), the median trend of static stress drops and scaled energy seems to be constant with increasing seismic moment. Whereas, for smaller earthquakes (Mw <similar to 2.5), it linearly increases with increasing seismic moment. Thus, our observation indicates a break -down of self -similarity scaling, suggesting a gradual change in dynamics of earthquake rupture with the increasing seismic moment. This could imply that smaller earthquakes are likely to propagate in a rough fluctuating stress field, typical of a region with RTS. Conversely, moderate or larger earthquakes tends to propagate in a relatively more stable correlated stress field. In a variable stress field, slip increases less with rupture size compared to a uniform stress field. Given that stress drop on a rupture area correlates with slip amount, this may explain the transitional patterns in stress drop with varying event size.