Triggered Seismicity Research Articles

The role of water lubrication in critical state fault slip

The conventional driving mechanisms including elevating the pore pressure or far-reaching poroelastic effects, may not necessarily explain the injection-triggered seismicity that can be up to 35–90 km far away from injection wells. In this experimental study, we examined the mechanism of water lubrication for the triggered seismicity where the fluid pressure was negligible compared to the in-situ stresses. The results show that fault slip can occur upon saturation when the ratio of the applied shear stress to the fault peak shear strength exceeds a critical threshold, which is defined to quantitatively assess the critical stress state for faults. The microscale slip test using atomic force microscopy reveals the mechanism responsible for the significant reduction of friction between granite surfaces from dry to wet state. A nanometer-thick water layer was attached on the granite surfaces, causing the shear plane shifting from the direct granite/granite interface to the water layer, and meanwhile the adhesive force between granite surfaces was significantly reduced by water intercalation, both of which played the dominate role in triggering fault slip. In practice, slip may be associated with in-situ fluid flow into far-field faults, which causes the increase of water and lubricates the fault zone.

Soft-sediment deformation structures in a Permo-carboniferous glacio-marine setting, Talchir Formation, Dudhi Nala, India

The Permo-carboniferous glacio-marine Talchir Formation of Dudhi Nala, West Bokaro Coal Basin, India, hosts soft-sediment deformation structures that originated from liquefaction and concomitant fluidization of unconsolidated sediments at or close to the sediment–water interface. Since liquefaction of water saturated sediments may be initiated by different endogenic or exogenic triggering agents, identification of the trigger needs careful analysis of the deformation structure, as well as depositional mechanism and environment of deposition of the host sediments. In-depth analysis of the studied liquefaction- and fluidization-induced soft-sediment deformation structures and sedimentary attributes of the host sediments of the studied succession unequivocally stand against the role of strong wave and tidal action, rapid sediment loading, mass-flow and subaqueous slides related shear stress, and water or gas seepage as the triggering agent. However, development of the soft-sediment deformation structures, close to sediment–water interface, closeness to syn-sedimentary fault, flowage along the tilt towards the fault, confinement within undeformed beds, episodic character and close similarities with structures that are formed during recent earthquakes, are consistent with seismic triggering for liquefaction. The studied shallow marine sediments record a sequel of climatic amelioration, glacial retreat, shallow marine sedimentation, glacio-isostatic rebound and related syn-sedimentary faulting and development of soft-sediment deformation structures.

The July 2019 Ridgecrest, California, Earthquake Sequence Recorded by Creepmeters: Negligible Epicentral Afterslip and Prolonged Triggered Slip at Teleseismic Distances

AbstractWe report sequential triggered slip at 271–384 km distances on the San Andreas, Superstition Hills, and Imperial faults with an apparent travel-time speed of 2.2 ± 0.1 km/s, following the passage of surface waves from the 4 July 2019 (17:33:49 UTC) Mw 6.4 and 6 July 2019 (03:19:53 UTC) Mw 7.1 Ridgecrest earthquakes. Slip on remote faults was not triggered instantaneously but developed over several minutes, increasing in duration with distance. Maximum slip amplitudes varied from 10 μm to 5 mm within minutes of slip nucleation, but on the southernmost San Andreas fault slip continued for two months and was followed on 16 September 2019 by a swarm of microearthquakes (Mw≤3.8) near Bombay Beach. These observations add to a growing body of evidence that fault creep may result in delayed triggered seismicity. Displacements across surface faults in the southern epicentral region and on the Garlock fault in the months following the Ridgecrest earthquakes were negligible (<1.1 mm), and they are interpreted to characterize surface strain adjustments in the epicentral region, rather than to result from discrete slip on surface faults.

In Situ Stress Orientation From 3 km Borehole Image Logs in the Koyna Seismogenic Zone, Western India: Implications for Transitional Faulting Environment

AbstractKnowledge of the in situ stress regime is critical to investigate the genesis of recurrent triggered seismicity over the past five decades in the Koyna seismogenic zone. Orientations of in situ horizontal stresses are determined for the first time from analyses of image logs in a 3 km deep scientific borehole KFD1 in the area. KFD1 passed through 1,247 m thick Deccan Traps and continued 1,767 m in the granitic basement. Stress‐induced wellbore failures, breakouts and drilling‐induced tensile fractures, are identified in the acoustic and microresistivity images of the granitic basement. Additionally, tightly constrained focal mechanisms of 50 earthquakes of M ≥ 3.6, reported in literature, are inverted to constrain the stress regime. Salient results are as follows: (i) wellbore breakouts and drilling‐induced tensile fractures constrain N9°W ± 17° orientation for maximum horizontal principal stress (SHmax); (ii) consistency with the regional NNW‐SSE orientation of SHmax from inversion of earthquake focal mechanisms shows that the borehole data are representative for the Koyna region; (iii) breakout rotations at multiple depths below 2,100 m indicate that the borehole may have passed through localized fault damage zones; (iv) consistent strike azimuths of steeply dipping fractures with SHmax orientation indicate favorable conditions for strike‐slip to normal transitional faulting environment; and (v) stress inversion of 50 well‐determined earthquake focal mechanisms supports transitional faulting environment in the Koyna seismogenic zone. Thus, stress orientation and fracture information from borehole data, together with stress regime constrained from inversion of earthquake focal mechanisms, shed new light on the faulting environment in the region.

Reconstruction of river valley evolution before and after the emplacement of the giant Seymareh rock avalanche (Zagros Mts., Iran)

Abstract. The Seymareh landslide, detached ∼10 ka from the northeastern flank of the Kabir-kuh fold (Zagros Mts., Iran), is recognized worldwide as the largest rock slope failure (44 Gm3) ever recorded on the exposed Earth surface. Detailed studies have been performed that have described the landslide mechanism and different scenarios have been proposed for explaining the induced landscape changes. The purpose of this study is to provide still missing time constraints on the evolution of the Seymareh River valley, before and after the emplacement of the Seymareh landslide, to highlight the role of geomorphic processes both as predisposing factors and as response to the landslide debris emplacement. We used optically stimulated luminescence (OSL) to date lacustrine and fluvial terrace sediments, whose plano-altimetric distribution has been correlated to the detectable knickpoints along the Seymareh River longitudinal profile, allowing the reconstruction of the evolutionary model of the fluvial valley. We infer that the knickpoint migration along the main river and the erosion wave propagation upstream through the whole drainage network caused the stress release and the ultimate failure of the rock mass involved in the landslide. We estimated that the stress release activated a mass rock creep (MRC) process with gravity-driven deformation processes occurring over an elapsed time-to-failure value on the order of 102 kyr. We estimated also that the Seymareh damming lake persisted for ∼3500 years before starting to empty ∼6.6 ka due to lake overflow. A sedimentation rate of 10 mm yr−1 was estimated for the lacustrine deposits, which increased up to 17 mm yr−1 during the early stage of lake emptying due to the increased sediment yield from the lake tributaries. We calculated an erosion rate of 1.8 cm yr−1 since the initiation of dam breaching by the Seymareh River, which propagated through the drainage system up to the landslide source area. The evolutionary model of the Seymareh River valley can provide the necessary constraints for future stress–strain numerical modeling of the landslide slope to reproduce the MRC and demonstrate the possible role of seismic triggering in prematurely terminating the creep-controlled time-to-failure pathway for such an extremely large case study.

Seismic and geologic controls on spatial clustering of landslides in three large earthquakes

Abstract. The large, shallow earthquakes at Northridge, California (1994), Chi-Chi, Taiwan (1999), and Wenchuan, China (2008), each triggered thousands of landslides. We have determined the position of these landslides along hillslopes, normalizing for statistical bias. The landslide patterns have a co-seismic signature, with clustering at ridge crests and slope toes. A cross-check against rainfall-induced landslide inventories seems to confirm that crest clustering is specific to seismic triggering as observed in previous studies. In our three study areas, the seismic ground motion parameters and lithologic and topographic features used do not seem to exert a primary control on the observed patterns of landslide clustering. However, we show that at the scale of the epicentral area, crest and toe clustering occur in areas with specific geological features. Toe clustering of seismically induced landslides tends to occur along regional major faults. Crest clustering is concentrated at sites where the lithology along hillslopes is approximately uniform, or made of alternating soft and hard strata, and without strong overprint of geological structures. Although earthquake-induced landslides locate higher on hillslopes in a statistically significant way, geological features strongly modulate the landslide position along the hillslopes. As a result the observation of landslide clustering on topographic ridges cannot be used as a definite indicator of the topographic amplification of ground shaking.

Evaluation of Seismicity Triggering: Insights from the Coulomb Static Stress Changes after the 30 August 1968 Dasht-e-Bayaz Earthquake (Mw = 7.1), Eastern Iran

In this research the Coulomb stress changes due to earthquake sequence that began in the 1936 to 1997 earthquakes in Eastern Iran and their triggering effect on the rupturing of adjacent faults were investigated. Obtained results revealed a well triggering relation between main shocks and later events on the Dasht-e-Bayaz and Abiz faults, respectively. The calculated stress maps indicate that positive and negative zones of Coulomb stress changes with increasing lobes of 0.1‒1 bar are found in the off-faults tips, located at rupture terminates while the decreasing Coulomb stress lobes lie in the high angle relative to the rupture plane or relieve stress along the ruptures. Cross sections drawn respect to the source and received faults confirm the results. Also, the stress perturbations resulted from successive earthquakes and the spatial patterns of the following earthquake distribution have a good consistency. This correlation shows that the population of secondary faults and subsequent earthquakes can be encouraged in the future by coseismic Coulomb stress changes due to mainshock and background loading.

Seismic clustering in the Sea of Marmara: Implications for monitoring earthquake processes

Quantifying regional earthquake cluster style is essential for providing a context for studies of seismicity patterns and earthquake interactions. Here, we identify clusters of seismicity in the Sea of Marmara region of the North Anatolian Fault, NW Turkey, using a recently derived high-resolution seismicity catalog and the nearest-neighbor earthquake cluster approach. The detected earthquake clusters are utilized for (1) determining spatial distribution of mainshock and aftershock rates and estimating the proximity to failure on different fault segments, (2) identifying fault sections having earthquake repeaters, and (3) finding areas with enhanced foreshock activity. About 6%, 70% and 24% of the events are identified as foreshocks, mainshocks and aftershocks, respectively, with the largest concentration of aftershocks and foreshocks located along the Western High and the Cinarcik Fault, respectively. The method successfully identifies regions where previous studies reported earthquake repeaters as indicator for fault creep and suggests additional repeater areas in the Gulf of Gemlik. The largest proportion of mainshocks with associated foreshocks and aftershocks are along the Western High and Cinarcik Fault segments, potentially indicating that these segments are closer to failure and have increased susceptibility to seismic triggering. Continuing studies can contribute to monitoring possible preparation phase of a large (M > 7) earthquake in the Marmara region near the Istanbul Metropolitan region.

Variable spatio-temporal clustering of microseismicity in the Hellenic Subduction Zone as possible indicator for fluid migration

This study investigates the variability of spatio-temporal microseismicity clustering and the occurrence of mutual triggering of events along the subduction interface in south-eastern Aegean as indication for fluid flow along and above the plate interface. We quantify spatial, temporal and spatio-temporal microseismicity clustering from the outer to the inner forearc and at intermediate depths. Waveform similarity indicates a decreasing of spatially clustered events from the outer and central towards the inner forearc and at intermediate depths. Highly similar events (cross-correlation >0.9), used as proxy for spatial clustering, decrease from the outer (30.2%) and central forearc (34.9%), towards the inner forearc (20.5%) and at intermediate depth (6.9%). Such highly similar events show increasing median inter-event times from the outer and central towards the inner forearc and at intermediate depth: 0.35, 0.34, 16.45, and 70 days, respectively. The Epidemic-Type-Aftershock-Sequences (ETAS) model, employed to investigate microseismicity temporal clustering, indicates an increase of the percentage of independent events from the outer (32%) and central (46%) forearc, to the inner forearc (93%) and at intermediate depth (93%). Hence, ETAS results suggest that mutual triggering of events is significant in the outer and central forearc, and it is almost absent in the inner forearc and at intermediate depths. Autocorrelation analysis, investigating spatio-temporal clustering, shows the tendency of earthquakes to occur close in space and time in the outer and central forearc, while in the inner forearc, and especially at intermediate depth, earthquakes are more homogeneously and randomly distributed. Combining the results from spatial, temporal and spatio-temporal analysis, we suggest that the different spatio-temporal patterns hint at systematic variations in the presence of migrating fluids on active faults close to failure. Triggering of seismicity is significant in the outer and central forearc, indicating fluid flow from the subduction interface, and it is diminishing towards the inner forearc. At intermediate depths, the nearly complete absence of mutual triggering of earthquakes indicates that there is little evidence for migration of fluids on active faults close to failure. Because intermediate depth seismicity in the Hellenic subduction zone occurs at P-T conditions where dehydration reactions are expected, fluids released by dehydration reactions within the slab are very likely migrating directly into the overlying mantle without triggering earthquakes.

Linking structures with the genesis and activity of mud volcanoes: examples from Emilia and Marche (Northern Apennines, Italy)

Mud volcanism is known to be strictly linked to tectonic structures, since they have the ability to trap hydrocarbon and other fluids, and eventually induce them to migrate from a deep reservoir (e.g. sited into an anticline core, where large overpressures may be generated), to the surface. A central theme is how fluids can migrate upward, and which is the role in this process of near structures (i.e. structures directly controlling the mud volcano system) and far structures (i.e. faults located far away from the mud volcano system). In this study, we investigate the role of both types of structures in the genesis and evolution of mud volcanoes. In particular, we investigate six mud volcano case studies from the Emilia-Romagna and Marche Pede–Apennine margin, in Italy, through integration of fieldwork, interpretation of available seismic reflection profiles and aerial photos. The results of these analyses support an intimate link of the investigated mud volcanoes with anticline structures. We discuss two different fluid migration settings, particularly (i) mud volcanoes emplaced on outcropping anticlines, and (ii) mud volcanoes located on top of buried structures, discerning when fluids are likely to exploit anticline-related fracture sets, or secondary structures and porosity. Finally, we speculate on how far structures, via the seismic triggering, may play a role in the occurrence of historical eruption of some of the investigated mud volcanoes.

Field investigation and numerical simulation of the seismic triggering mechanism of the Tahman landslide in eastern Pamir, Northwest China

Landslides are one of the most common geological hazards in many parts of the world, which have brought catastrophic consequences to mankind. Seismic shaking can be an important trigger factor generally resulting in large-scale landslides. For some paleolandslides triggered by earthquakes, despite the lack of historical records, certain paleoseismic parameters can be back-analyzed. In this paper, the discrete element method (DEM) was applied to back-analyze the peak ground acceleration (PGA) triggering the Tahman landslide, a paleolandslide located in eastern Pamir, northwest China. Firstly, detailed field geological and geomorphological investigation was conducted to reveal the natural and seismic characteristics of the landslide. Subsequently, discrete element modeling was carried out. Simulation results suggested that the Tahman rock slope was stable under natural conditions. Owing to arid climate and water shortage of the study area, the influence of water was ignored. Additionally, from the perspective of the occurrence time of the landslide and geomorphologic feature of the deposits, the landslide was not caused by glacier activity and freeze–thaw events, and thus, must have been triggered by an earthquake. A peak ground acceleration (PGA) of 0.55 g was found to be a critical value for triggering landslides, and a PGA of 0.76 g could best simulate the kinematic process of a landslide when the simulation results were calibrated against the current geomorphological features of the landslide deposits. It can be concluded from the present study that: (1) PGA amplification in the rock slope was related to the topography during earthquake excitation; (2) the effective duration of the seismic wave had a considerable impact on the back-analyzed PGA, which decreased with increasing effective duration; and (3) the long runout of the Tahman landslide was due to its large volume and the interaction of the slide mass with moraines.

Borehole Microseismic Imaging of Hydraulic Fracturing: A Pilot Study on a Coal Bed Methane Reservoir in Indonesia

Over the last decade, microseismic monitoring has emerged as a considerable and capable technology for imaging stimulated hydraulic fractures in the development of unconventional hydrocarbon resources. In this study, pilot hydraulic-fracturing treatments were operated at a coal-bed methane (CBM) field in Indonesia to stimulate the flow and increase the reservoir's permeability while the monitoring system was set in a single near-vertical borehole. Locating event sources accurately is fundamental to investigating the induced fractures, but the geometry of a single downhole array is a challenging data processing task, especially to remove ambiguity of the source locations. The locating procedure was reviewed in 3 main steps: (i) accurate picking of P- and S-wave phases; (ii) inclusion of P-wave particle motion to estimate the back azimuth; (iii) guided inversion for hypocenter determination. Furthermore, the seismic-source moment magnitudes were calculated by employing Brune's model. Reliable solutions of locations were obtained as shown statistically by uncertainty ellipsoids and a small misfit. Based on our results, both induced and triggered seismicity could be observed during the treatments and therefore conducting intensive monitoring is important. The triggered seismicity is an undesired activity so disaster precautions need to be taken, in particular for preventing reactivation of pre-existing faults.

The July 20, 2017 M6.6 Kos Earthquake: Seismic and Geodetic Evidence for an Active North-Dipping Normal Fault at the Western End of the Gulf of Gökova (SE Aegean Sea)

On July 20, 2017 22:31 UTC, a strong Mw = 6.6 earthquake occurred at shallow depth between Kos Island (Greece) and Bodrum (Turkey). We derive a co-seismic fault model from joint inversion of geodetic data (GNSS and InSAR) assuming that the earthquake can be modelled by the slip of a rectangular fault buried in an elastic and homogeneous half-space. The GNSS observations constrain well most of the model parameters but do not permit to discriminate between south- and north-dipping planes. However, the interferograms, produced from C-band ESA Sentinel 1 data, give a clear preference to the north-dipping plane. We also map surface motion away from the satellite along the Turkish coast (from Bodrum towards the east) which reached about 17 cm onshore islet Karaada. The best-fit model is obtained with a 37° north-dipping, N283°E striking normal fault, in agreement with the published moment tensor solutions. The resolved slip vector is dominantly normal with a slight component of left-lateral motion (15°). The surface projection of the seismic fault outcrops in the Gokova ridge area, a well-developed bathymetric feature inside the western Gulf of Gokova. Our geodetic model fits the pattern of the shallow, north-dipping aftershocks obtained from rigorous relocation of all available recordings in the region (about 1120 events; relocated mainshock is at 36.955°N, 27.448°E; depth at 9.2 km ± 0.5 km). The relocated aftershocks also indicate clustering at both ends of the rupture and seismicity triggering mainly towards the east and the north, within 2 weeks following the mainshock. We also analysed regional GPS data (interseismic velocities) and obtained an extension rate of 3.2 mm/yr across the Gokova rift, along a direction N165°E.

Dynamic triggering of microseismicity in Southwest China following the 2004 Sumatra and 2012 Indian Ocean earthquakes

We conduct a systematic detection of microseismicity around Yunnan Province, Southwest China following the 2004 Mw9.1 Sumatra and 2012 Mw8.6 Indian Ocean earthquakes. Based on seismicity analysis on China Earthquake Network Center (CENC) catalog and spectral analysis of continuous waveforms, we find clear evidence of triggered seismicity in Yunnan immediately after the 2004 mainshock, including a magnitude 4.8 earthquake near Tengchong volcanic region. In comparison, we do not find clear evidence of instantaneous triggering following the 2012 mainshock. We further apply a matched filter technique to detect microearthquakes in Tengchong volcanic region 15 days before and 15 days after the 2012 mainshock. The seismicity rate remains stable immediately after the 2012 mainshock, and increases in the next few days, indicating a possible delayed triggering, although similar changes can be seen at other times long before and after the mainshock. In addition, we find possible evidence of remote triggering near the city of Yibin in Southern Sichuan Province, Southwest China, near the sites with ongoing shale gas development. We also examine other magnitude >8 earthquakes from the Sumatra region and do not find any additional evidence of dynamic triggering around Yunnan. We suggest that the absolute value of dynamic stress change is the most important factor to explain why the 2004 Sumatra event triggered a widespread seismicity rate increase around Yunnan, while other great events around Sumatra did not.

A New Earthquake Sequence at Koyna–Warna, India, and Its Implication for Migration of the Reservoir Triggered Seismicity

A New Earthquake Sequence at Koyna–Warna, India, and Its Implication for Migration of the Reservoir Triggered Seismicity

Landslides near Enguri dam (Caucasus, Georgia) and possible seismotectonic effects

Abstract. The Enguri dam and water reservoir, nested in the southwestern Caucasus (Republic of Georgia), are surrounded by steep mountain slopes. At a distance of 2.5 km from the dam, a mountain ridge along the reservoir is affected by active deformations with a double vergence. The western slope, directly facing the reservoir, has deformations that affect a subaerial area of 1.2 km2. The head scarp affects the Jvari–Khaishi–Mestia main road with offsets of man-made features that indicate slip rates of 2–9 cm yr−1. Static, pseudostatic and Newmark analyses, based on field and seismological data, suggest different unstable rock volumes based on the environmental conditions. An important effect of variation of the water table is shown, as well as the possible destabilization of the slope following seismic shaking, compatible with the expected local peak ground acceleration. This worst-case scenario corresponds to an unstable volume on the order of up to 48±12×106 m3. The opposite, eastern slope of the same mountain ridge is also affected by wide deformation affecting an area of 0.37 km2. Here, field data indicate 2–5 cm yr−1 of slip rates. All this evidence is interpreted as resulting from two similar landslides, whose possible causes are discussed, comprising seismic triggering, mountain rapid uplift, river erosion and lake variations.

Statistical parameters of seismicity induced by the impoundment of the Three Gorges Reservoir, Central China

Induced seismicity has been observed in the Three Gorges Reservoir (TGR) area, central China, since its impoundment in May, 2003. Data from a local seismic network in the TGR area and a temporal network allow us to obtain a detailed picture of the spatio-temporal distribution of earthquakes in the Badong area, in which earthquakes are clustered. Statistical parameters, namely the seismic b-value, spatial correlation length, and fractal dimension of the hypocentre distribution, as an integrated set, are investigated and taken as indicators to divide the seismicity in the study area into typical stages. Epidemic-type aftershock sequence modelling, which is useful for extracting possible fluid signals from earthquake occurrences, is applied to the data set. The temporal variations in the parameters indicate a long-term increasing and diffusing stress until the present. Thus, much more attention should be given to seismic activity in the TGR area. The largest earthquake (M5.1), which was followed by aftershocks, a large fraction of which were Omori-type, primarily resulted from a triggered fault movement. In the M4.3 earthquake sequence, both Omori-type seismic triggering and reservoir triggering are statistically important. After the Omori-type aftershocks (which accounted for a small fraction (<20% in total) of earthquakes) are removed, the forced seismic rate is correlated with the water level of the reservoir, indicating that most of the small earthquakes were induced by the reservoir.

Fully coupled hydro–mechanical controls on non-diffusive seismicity triggering front driven by hydraulic fracturing

The spatio–temporal evolution of fluid-injection-induced seismicity is often bounded by a triggering front that expands away from the injection point in space and time. For some injection scenarios, the triggering front is thought to be directly linked to pore pressure diffusion, but in the case of hydraulic fracturing, the stress interaction of the growing tensile fracture with natural joints may be more significant. In order to explore the concept of a triggering front in this context, we use a fully coupled hydro–mechanical finite-discrete element (FDEM) approach to simulate microseismicity induced by hydraulic fracture growth. The medium contains a network of randomly oriented pre-existing fractures that are activated based on the Mohr–Coulomb failure criterion. As expected, the primary triggering front is defined by the envelope of microseismicity that tracks the hydraulic fracture, although more distal events are triggered by mechanical stress changes beyond the bounds of the triggering front. However, these distal events are approximately synchronous with initiation of the hydraulic fracture and are attributed to far-field elastic perturbations associated with the stress wave spread in the medium. A field example indicates that patterns of seismicity that emerge from our simulations have characteristics similar to observed microseismicity during hydraulic fracturing.

Interpretations of Reservoir‐Induced Seismicity May Not Always Be Valid: The Case of Seismicity during the Impoundment of the Kremasta Dam (Greece, 1965–1966)

The ‘Kremasta seismic sequence’ in western Greece is one of the most commonly cited examples of Reservoir Induced Seismicity (RIS). Here, we show that this ‘sequence’ is a result of normal tectonic activity and that only some small, unrelated microseismic events are reservoir induced. Shortly after the beginning of the impoundment of the Kremasta Dam in 1965, the then newly established seismic monitoring network in Greece recorded two Ms ≥ 6.0 events and numerous small shocks spread over a 120 km wide region. These were interpreted as a single seismic sequence (namely the Kremasta seismic sequence), and assumed to be reservoir induced. We revisit the epicenter locations of these events and interpret them in the framework of the regional tectonic context and the local hydrogeology. Placing these events into the local context shows that they represent an amalgamation of separate, ordinary (tectonic) seismic sequences. Further, the regional rocks are highly fragmented by small faults and the spatial distribution of seismic events is not consistent with a model of stress transfer from reservoir loading. In addition, it is not likely that events at such long (> 20-30 km) distances from the reservoir could be induced by an initial reservoir load head of 30 m. Whilst the larger magnitude events are tectonic, after impoundment local residents reported an unusual frequency of small microseismic events felt only within 10 km of the dam. We provide evidence that these are a result of the collapse of numerous shallow karstic cavities adjacent and beneath the reservoir due to increased water load (locally 100-150 m depth). This study has significant implications for interpretation of seismic triggering mechanisms in other regions: earthquake occurrence within the proximity of reservoirs during and after impoundment time cannot be assumed to be RIS unless supported by seismological, geological and hydrogeological evidence.

Rate of Change in Lake Level and Its Impact on Reservoir Triggered Seismicity

Rate of Change in Lake Level and Its Impact on Reservoir Triggered Seismicity