Hydromechanics of the Koyna–Warna Region, India INMACULADA DURA ´ -GO ´ MEZ 1 and PRADEEP TALWANI 1 Abstract—Continuous reservoir-induced seismicity has been observed in the Koyna–Warna region in western India following the beginning of impoundment of Koyna and Warna Reservoirs in 1961 and 1985, respectively. This seismicity includes 19 events with M C 5.0 which occurred in 7 episodes (I–VII) between 1967 and 2005 at non-repeating hypocentral locations. In this study, we examined the first six episodes. The seismicity occurs by diffusion of pore pressures from the reservoirs to hypocentral locations along a saturated, critically stressed network of NE trending faults and NW trending fractures. We used the daily lake levels in the two reservoirs, from impoundment to 2000, to calculate the time history of the diffused pore pressures and their daily rate of change at the hypocentral locations. The results of our analysis indicate that Episodes I and IV are primarily associated with the initial filling of the two reservoirs. The diffused pore pressures are generated by the large (20–45 m) annual fluctuations of lake levels. We interpret that critical excess pore pressures [ *300 kPa and [ *600 kPa were needed to induce Episodes I–III and Episodes IV–VI, respectively, suggesting the presence of stronger faults in the region. The exceedance of the previous water level maxima (stress memory) was found to be the most important, although not deter- mining factor in inducing the episodes. The annual rise of 40 m or more, rapid filling rates and elevated dp/dt values over a filling cycle, contributed to the rapid increase in pore pressure. Key words: Reservoir-induced seismicity, Koyna reservoir, pore pressure diffusion, stress memory, strain hardening. 1. Introduction Continuous reservoir-induced seismicity (RIS) has been observed since 1963 in the Koyna–Warna region (KWR) after the impoundment of the Koyna Reservoir (in western India) in 1961 and Warna Reservoir in 1985. The seismicity covers a volume *30 km long, *20 km wide and *10 km deep. Through 2006, more than 100,000 earthquakes have been recorded, including more than 170 events with M C 4 and 19 events with M C 5 (henceforth refer- red as M 5 events) with non-repetitive epicenters (GUPTA, 2002;GUPTA et al., 2007). Over the years, the KWR has been the subject of numerous studies (see e.g., in the last two decades, GUPTA, 1992, 2002; TALWANI et al., 1996;TALWANI, 1997b, 2000;RASTOGI et al., 1997;MANDAL et al., 1998;GUPTA et al., 2002, 2005, 2007;SATYANARAYANA et al., 2005;KUMAR et al., 2006;SARMA and SRINAGESH, 2007;CHADHA et al., 2008). A few studies have been aimed at understanding the mechanism of RIS at the KWR (e.g., SIMPSON et al., 1988;KALPNA and CHANDER, 2000;TALWANI 1995, 1997a;RAJENDRAN and HARISH, 2000;PANDEY and CHADHA, 2003). A unique observation in the region is the contin- uing occurrence of M 5 earthquakes, 44 and 20 years after the impoundment of the Koyna and Warna Res- ervoirs, respectively. In this paper, we combine our improved understanding of both the seismotectonic framework of the KWR and the mechanism of RIS in order to explain the RIS and hydromechanics of the KWR, by a study of the pore pressure history associ- ated with the spatial and temporal evolution of M 5 earthquakes. We chose M 5 events because of their non-repeating hypocentral locations, each of which was associated with a *10 km 2 rupture area. Under- standing the origin of these earthquakes is challenging and fascinating, and is the subject of this study. We first present the revised seismotectonic frame- work of the KWR in light of results of studies of recent seismicity and geophysical investigations. We suggest the association of various fractures and faults with the M 5 earthquakes. Our analysis suggests that the satu- rated fractures in the KWR are critically stressed (Sect. 2). Next we describe our understanding of the pre- dominant role of pore pressure diffusion (CHEN and 1 Department of Geological Sciences, University of South Carolina, Columbia, SC 29208, USA. E-mail: talwani@geol.sc.edu Pure Appl. Geophys. Ó 2009 Birkha ¨user Verlag, Basel/Switzerland DOI 10.1007/s00024-009-0012-5 Pure and Applied Geophysics