Increase In Seismicity Rate Research Articles

Investigating the Impacts of a Wet Typhoon on Microseismicity: A Case Study of the 2009 Typhoon Morakot in Taiwan Based on a Template Matching Catalog

AbstractRecent studies suggested that transient and long‐term stress changes caused by Earth's surface processes (e.g., extreme weather events, annual variations on groundwater storages) can affect earthquake activities in the subsurface. However, these studies may be limited by the completeness of standard earthquake catalogs, especially during or right after extreme weather events. Here we apply the template matching method to build a more complete earthquake catalog in Taiwan spanning seven months before and 12 months after 2009 typhoon Morakot, which brought the highest rainfall in southern Taiwan in the past 60 years and triggered numerous landslides. We then use the enhanced catalog to investigate possible influences of typhoon‐driven Earth's surface processes (atmospheric pressure, precipitation, and erosion) on local seismicity. We find that the seismicity rate of a 40‐day earthquake sequence in northeastern Taiwan was reduced significantly right after the passage of typhoon Morakot's eye center. In the typhoon‐triggered landslide zone in southern Taiwan, we find a slight increase in background seismicity rate in the next year after Morakot, matching the results of a recent study. However, we do not observe a clear change in the Gutenberg‐Richter b‐value in this zone, which is different from the recent study. Station outages during and right after Morakot prevents us from better understanding short‐term precipitation effect on local seismicity. Overall, except for a reduction in seismicity rate near the typhoon's low‐pressure eye center in northeastern Taiwan, we do not observe other clear seismicity changes that can be attributed to surface changes induced by typhoon Morakot.

Can Hydrocarbon Extraction From the Crust Enhance or Inhibit Seismicity in Tectonically Active Regions? A Statistical Study in Italy

A number of oil- and gas-producing leases have been operating in Italy in the last decades, many of which are located in the surroundings of tectonically active regions. Identifying human-induced seismicity in areas with high levels of natural seismicity is a difficult task for which virtually any result can be a source of controversy. We implemented a large-scale analysis aiming at tracking significant departures of background seismicity from a stationary behavior around active oil and gas development leases in Italy. We analyzed seismicity rates before and after hydrocarbon peak production in six oil-producing and 43 gas-producing leases, and evaluate the significance of possible seismicity rate changes. In a considerable number of cases seismicity rate results stationary. None of the observed cases of seismicity rate increase after the peak production is statistically significant (at as.l.= 0.05). Conversely, considering cases of seismicity rate decrease after peak production, our results suggest that the seismicity rate reduction is statistically significant (s.l.= 0.05) around one oil-producing lease (Val d’Agri, Basilicata) and around a cluster of gas-producing leases in Sicily. Our results put in evidence correlated changes between the rates of shallow seismicity and hydrocarbon production in these areas, which are then identified as hotspots requiring more detailed research; assessing actual causal relationships between these processes will require further physically-based modelling. If a physical causative link between these processes exists, then the observed seismicity rate reduction could either be due to increased seismicity during the progressive increase in production before reaching its maximum, or to an actual seismicity rate reduction after that peak. Considering that there is evidence of seismicity occurring before the start of hydrocarbon production, which contrasts with the evident reduction of events observed after the peak production, we think it likely that the seismicity inhibition is a plausible hypothesis. Using a simple model we also calculate Coulomb stress changes in planes optimally oriented for failure, and we show that under some conditions the inhibition of seismicity is feasible in at least one of these cases. We conclude that more efforts to study the mechanisms and the possible consequences of anthropogenically-driven seismicity inhibition are required.

Uplift and Seismicity Driven by Magmatic Inflation at Sierra Negra Volcano, Galápagos Islands

AbstractAlthough episodes of surface uplift and elevated seismicity precede many volcanic eruptions, their temporal evolution is often complex, and apparently in contradiction to simple trends predicted by mechanical deformation models. Here, we use continuous global positioning system and seismic data recorded at Sierra Negra volcano, Galápagos Islands, to show how the edifice responded to stress changes driven by magma accumulation in a shallow sill. The rate of uplift varied during the 13 years and 6.5 m of inflation before the 2018 eruption. The number of earthquakes per unit of uplift increased exponentially with total uplift as the differential stress increased. Accordingly, the temporal seismicity rate varied in time as a function of both the total uplift and the uplift rate. The Gutenberg‐Richter b‐value decreased as a function of total uplift. In the final six months before the eruption, a sequence of large (M > 4) earthquakes regulated the state of stress on the fault, each being followed by 2–3 days of postseismic quiescence, and retarding the increase in seismicity rate. These earthquakes did not affect the overall uplift rate. Subsidence of 8.5 m accompanied the 2‐month eruption. On resumption of uplift, the number of earthquakes per unit of uplift was very low, and the b‐value high, reflecting the relaxed stress state of the fault system. These observations show that crustal deformation becomes increasingly brittle at higher stress states, and supports theoretical models based on elastic‐brittle mechanics. They suggest that joint interpretation of deformation and seismicity is key for forecasting future eruptions in similar volcanic settings.

A Significant Increase in Interplate Seismicity near Major Historical Earthquakes Offshore Martinique (FWI)

ABSTRACT Understanding the processes that may be at the origin of major earthquakes in subduction zones is highly challenging, especially in the case of slowly converging areas such as the Lesser Antilles subduction zone. Our study reveals a recorded increase in seismicity rate and cumulative seismic moment over the last two decades offshore Martinique island and, particularly, in the presumed rupture area of the major historical 1839 earthquake. This sustained seismicity is shared between extensive intermediate depth activity since the 2007 Mw 7.4 earthquake that occurred about 20 km north of Martinique and a compressive seismic cluster located in the seismogenic zone of the subduction zone. We also observe a downward migration of the seismicity along the plate interface up to a 60–65 km depth and a recorded increase in the magnitudes of the earthquakes. All these observations may indicate ongoing changes along the plate boundary interface near the area of the historical major rupture of 1839, although more detailed statistical studies and additional data are necessary to confirm this trend.

Fluid‐Earthquake and Earthquake‐Earthquake Interactions in Southern Kansas, USA

AbstractAn increase in injection activity associated with energy production in southern Kansas starting in 2013 has been linked to the occurrence of more than 130,000 earthquakes (M −1.5 to 4.9) between 2014 and 2017. Studies suggest that the dramatic increase in seismicity rate is related to wastewater injection into the highly permeable Arbuckle formation. Most of the seismicity is located in the underlying crystalline basement, for which hydrological properties and specific fault geometries are unknown. Additionally, some earthquake clusters occurred relatively far (tens of kilometers) from the main injection wells. Therefore, the effect of pore pressure diffusion may be insufficient to explain the relation between the volume of injected fluids and the spatiotemporal evolution of seismicity. Combining physical models (static stress and poroelasticity) and a statistical cluster analysis applied to a high‐resolution relocated catalog, we analyze the evolution of seismicity in southern Kansas. We find that pore pressure changes (Δp) and Coulomb stress changes (ΔCFS) due to fluid diffusion smaller than 0.1 MPa are enough to initiate seismic sequences, which then evolve depending on their distance from the major injection wells. However, we find that earthquake sequences have different seismogenic responses to Δp and ΔCFS in terms of triggering threshold. In regions located close to disposal wells (Harper area) our cluster analysis suggests that both earthquake interactions and fluid diffusion control the evolution of seismicity. On the other hand, at greater distances (Milan area), where clustering behavior suggests greater earthquake interactions, we find that coseismic ΔCFS are larger than Δp.

Seismicity Rate Change at the Coso Geothermal Field Following the July 2019 Ridgecrest Earthquakes

ABSTRACT Many geothermal and volcanic regions experience remote and regional triggering following large earthquakes. The transient or permanent changes in stresses acting on faults and fractures can lead to changes in seismicity rates following either the passage of teleseismic waves or the permanent change in stresses following regional events. One such region of prevalent triggering is the Coso Geothermal Field (CGF) in eastern California, which is located roughly 30 km to the north of the 2019 Mw 7.1 Ridgecrest epicenter. Previous regional earthquakes have not only seemingly caused increase in seismicity rate surrounding the CGF, but also showed an absence of such rate increases in the CGF itself. To test whether seismicity rates in the CGF were dissimilar to the surrounding area following the Mw 7.1 Ridgecrest earthquake, I carry out seismicity rate change calculations using a catalog of seismicity compiled using a local seismic network and find that the behavior at CGF is identical to the surrounding area. Comparisons of seismicity rate changes calculated using a regional-network-derived catalog, and the local-network-derived catalog show that for a moderate, regional earthquake (2009 Mw 5.2 Olancha, California), the local network catalog reveals a change in seismicity rate whereas the regionally network catalog shows no significant changes. The differences are possibly related to incomplete sampling of seismicity using the regional network due to the existence of a shallow brittle–ductile transition centered on the CGF. The CGF, thus, is prone to triggering from both teleseismic and regional earthquakes.

Temporal patterns of induced seismicity in Oklahoma revealed from multi-station template matching

Over the past decade, Oklahoma became the most seismically active region of the mid-Continental USA as a result of industry operations. However, seismic network limitations and completeness of earthquake catalogs have restricted the types of analyses that can be performed. By applying multi-station template matching on the 23,889 cataloged earthquakes in Oklahoma and Southern Kansas between late-2008 and 2016, we increased the number of detected earthquakes to 209,409 events. While the improved catalog produced an order of magnitude events than the original catalog, the frequency-magnitude distribution remains similar to the original catalog. We found that the coefficient of variation of interevent times in small spatial bins tends to spatially correlate with the location of M ≥ 4 earthquakes. The improved catalog reveals the pervasiveness of swarm-like patterns in seismicity across the entire study region. The rapid increase in seismicity rate of these swarms in 2013 coincided with a reduction in the calculated p values (power law decay rates) before and after larger events. We also used the catalog to revisit the temporal patterns in the four M ≥ 5 sequences, finding more active foreshock behavior than previously recognized and variations in aftershock behavior. When compared against poroelastic stress models for the Pawnee and Fairview sequences, the catalog shows an improved correlation with stress that accounts for variable-rate injection, supporting the conclusion that injection rate is an important contributor to seismic hazard.

Compaction-Induced Permeability Loss’s Effect on Induced Seismicity During Reservoir Depletion

Producing fluid out of subsurface reservoirs causes stress changes that can lead to induced seismicity. In the fully coupled theory of poroelasticity, these stress changes are caused by the gradient of pore pressure which acts as an internal force in the momentum balance equation. This implies that, if a larger pore pressure gradient is required to produce the fluid, larger stresses will be induced, likely leading to a higher seismicity rate. Based on Darcy’s Law, the gradient of pore pressure required to achieve a certain fluid production rate is inversely proportional to the permeability. The lower the permeability, the larger the pore pressure gradient required. Consequently, the permeability reductions that are seen due to compaction during fluid production may be significant in terms of induced seismicity because this will increase the pore pressure gradient required to produce a certain amount of fluid. Therefore, using a poroelastic flow model and an existing seismicity model, it is shown that compaction-induced permeability loss has the effect of increasing the stress changes associated with the production of a certain amount of fluid and thereby indirectly increases the seismicity rate, even when the compaction remains in the near-elastic range. In the event that compaction begins occurring inelastically, the permeability decreases and seismicity rate increase are much more drastic. Finally, it is shown that the optimal orientation for a horizontal well, if inelastic compaction near the wellbore is to be avoided, is parallel to $$S_{Hmax}$$ . These results have implications for reservoir management in fluid producing reservoirs.

Seismicity induced by hydraulic fracturing and wastewater disposal in the Appalachian Basin, USA: a review

Eastern Ohio is an area of North America where a significant increase in seismicity rate was noted in the early 2010s. This increase has been associated with intensification of unconventional gas extraction performed in the Appalachian Basin and has been directly linked to two processes: hydraulic fracturing and disposal of the associated wastewater. In this paper, we review the recent seismicity in the Appalachian Basin including various episodes of induced seismicity that were temporally and spatially linked to operational activity, and we have performed some comparable analyses on the most recent sequences. The activities have not been as pervasive as other areas of North America, such that the cases are typically isolated and provide opportunities to study the seismogenic process in detail. The observed seismicity is concentrated in a narrow corridor that extends north–south in eastern Ohio and into central West Virginia, perhaps due to differences in operational targets and geologic variations. Ohio appears to have a higher prevalence of seismicity induced by wastewater disposal than surrounding states, but this is based on limited number of cases. Ohio also has an order of magnitude higher prevalence of seismicity induced by hydraulic fracturing than surrounding states, and prior work has suggested this is due to the targeting of the deeper Utica–Point Pleasant formation in Ohio that is closer to basement rocks than the Marcellus formation in West Virginia or Pennsylvania. In areas where hydraulic fracturing has induced seismicity, the percentage of stimulated wells that produce detectable seismicity is approximately 10–33%. Detailed studies of induced seismicity via double-difference relocation and focal mechanism analysis have revealed a series of linear fault segments, none of which correspond to previously mapped faults. Yet, the remarkable coherence in their orientation suggests these were preexisting, optimally oriented, and critically stressed. These fault orientations reveal a consistent regional stress field that only varies over a narrow azimuthal range from ~ 50° to 74°. The strongest observed seismic events in Ohio appear to occur in the Precambrian basement and indicate that these rocks have the maturity needed to produce M > 2 earthquakes and hence the greater potential hazard.

Investigating Dynamic Triggering of Seismicity by Regional Earthquakes: The Case of the Corinth Rift (Greece)

AbstractDynamic triggering has been commonly observed after large teleseismic events, but the physics behind it is still under debate. To broaden observations, we here focus on the dynamic triggering by regional earthquakes, that is, by events with magnitude lower than 6.2 at distances smaller than 600 km. The western part of the Corinth Rift (Greece) is characterized by intense seismic swarms and is therefore adapted to study such responses. The microseismicity rates before and after the transient perturbations are high enough to analyze 30 regional earthquakes out of the 59 occurring in 2013. More than 40% of those 30 events, including earthquakes with magnitude as small as 4.5, are associated with a significant seismicity rate increase. The triggerability primarily depends on the amplitude of the seismic waves. However, triggering is mainly observed when the seismic perturbations are orthogonal to the faults, which suggests that fluid pressurization is likely involved.

EVIDENCE OF PRECURSORY PATTERNS IN AGGREGATED TIME SERIES

We investigate temporal changes in seismic activity observed in the West Corinth Gulf and North-West Peloponnese during 2008 to 2010. Two major earthquake sequences took place in the area at that time (in 2008 and 2010). Our aim is to analyse Greek seismicity to attempt to confirm the existence or non-existence of seismic precursors prior to the strongest earthquakes. Perhaps because the area is geologically and tectonically complex, we found that it was not possible to fit the data well using a consistent Epidemic Type Aftershock Sequence (ETAS) model. Nor could we unambiguously identify foreshocks to individual mainshocks. Therefore we sought patterns in aggregated foreshock catalogues. We set a magnitude threshold (M3.5) above which all the earthquakes detected in the study area are considered as “mainshocks”, and we combined all data preceding these into a single foreshock catalogue. This reveals an increase in seismicity rate not robustly observable for individual cases. The observed effect is significantly greater than that consistent with stochastic models, including ETAS, thus indicating genuine foreshock activity with potential useful precursory power, if sufficient data is available, i.e. if the magnitude of completeness is sufficiently low.

Detection of earthquake swarms at subduction zones globally: Insights into tectonic controls on swarm activity

AbstractEarthquake swarms are characterized by an increase in seismicity rate that lacks a distinguished main shock and does not obey Omori's law. At subduction zones, they are thought to be related to slow‐slip events (SSEs) on the plate interface. Earthquake swarms in subduction zones can therefore be used as potential indicators of slow‐slip events. However, the global distribution of earthquake swarms at subduction zones remains unclear. Here we present a method for detecting such earthquake sequences using the space‐time epidemic‐type aftershock‐sequence model. We applied this method to seismicity (M ≥ 4.5) recorded in the Advanced National Seismic System catalog at subduction zones during the period of 1995–2009. We detected 453 swarms, which is about 6.7 times the number observed in a previous catalog. Foreshocks of some large earthquakes are also detected as earthquake swarms. In some subduction zones, such as at Ibaraki‐Oki, Japan, swarm‐like foreshocks and ordinary swarms repeatedly occur at the same location. Given that both foreshocks and swarms are related to SSEs on the plate interface, these regions may have experienced recurring SSEs. We then compare the swarm activity and tectonic properties of subduction zones, finding that swarm activity is positively correlated with curvature of the incoming plate before subduction. This result implies that swarm activity is controlled either by hydration of the incoming plate or by heterogeneity on the plate interface due to fracturing related to slab bending.

The Effects of Varying Injection Rates in Osage County, Oklahoma, on the 2016Mw 5.8 Pawnee Earthquake

ABSTRACT The 2016 M w 5.8 Pawnee earthquake occurred in a region with active wastewater injection into a basal formation group. Prior to the earthquake, fluid injection rates at most wells were relatively steady, but newly collected data show significant increases in injection rate in the years leading up to earthquake. For the same time period, the total volumes of injected wastewater were roughly equivalent between variable‐rate and constant‐rate wells. To understand the possible influence of these changes in injection, we simulate the variable‐rate injection history and its constant‐rate equivalent in a layered poroelastic half‐space to explore the interplay between pore‐pressure effects and poroelastic effects on the fault leading up to the mainshock. In both cases, poroelastic stresses contribute a significant proportion of Coulomb failure stresses on the fault compared to pore‐pressure increases alone, but the resulting changes in seismicity rate, calculated using a rate‐and‐state frictional model, are many times larger when poroelastic effects are included, owing to enhanced stressing rates. In particular, the variable‐rate simulation predicts more than an order of magnitude increase in seismicity rate above background rates compared to the constant‐rate simulation with equivalent volume. The observed cumulative density of earthquakes prior to the mainshock within 10 km of the injection source exhibits remarkable agreement with seismicity predicted by the variable‐rate injection case.

Evolution of seismicity near the southernmost terminus of the San Andreas Fault: Implications of recent earthquake clusters for earthquake risk in southern California

AbstractThree earthquake clusters that occurred in the direct vicinity of the southern terminus of the San Andreas Fault (SAF) in 2001, 2009, and 2016 raised significant concern regarding possible triggering of a major earthquake on the southern SAF, which has not ruptured in more than 320 years. These clusters of small and moderate earthquakes with M ≤ 4.8 added to an increase in seismicity rate in the northern Brawley seismic zone that began after the 1979 Mw 6.5 Imperial Valley earthquake, in contrast to the quiet from 1932 to 1979. The clusters so far triggered neither small nor large events on the SAF. The mostly negative Coulomb stress changes they imparted on the SAF may have reduced the likelihood that the events would initiate rupture on the SAF, although large magnitude earthquake triggering is poorly understood. The relatively rapid spatial and temporal migration rates within the clusters imply aseismic creep as a possible driver rather than fluid migration.

Earthquake clusters in NW Peloponnese.

Clusters commonly occur as main shock – aftershock (MS-AS) sequences but also as earthquake swarms, which are empirically defined as an increase in seismicity rate above the background rate without a clear main shock. A delcustering algorithm is employed to identify clusters from a complete catalog of earthquakes that occurred in the area of NW Peloponnese (Greece) during 1980-2007. In order to distinguish these clusters we calculate the skewness and kurtosis of seismic moment release for each cluster, since swarm-like sequences generally have lower skew value of moment release history than MS-AS. The spatial distribution of b-value was calculated for the entire catalog as for the declustered one, in order to correlate them with seismicity behavior of the region. Finally, the pre-stress field of Achaia 2008 earthquake was calculated aiming to associate the stress accumulation with the occurrence of the identified clusters

Injection‐induced seismicity in Carbon and Emery Counties, central Utah

AbstractUtah is one of the top producers of oil and natural gas in the United States. Over the past 18 years, more than 4.2 billion gallons of wastewater from the petroleum industry has been injected into the Navajo Sandstone, Kayenta Formation, and Wingate Sandstone in Carbon and Emery Counties, central Utah, where seismicity has increased during the same period. Previous studies have attributed this seismicity to coal mining. Here, we present evidence for wastewater injection being a major cause of the increased seismicity. We show that, in the coal mining area, seismicity rate increased significantly 1–5 years following the wastewater injection, and the earthquakes, mostly with magnitudes <3.0, are concentrated in areas seismically active prior to the injection. Using simple analytical and numerical models, we show that the injection in central Utah can sufficiently raise pore pressure to trigger seismicity within 10–20 km of the injection wells, and the time needed for the diffusion of pore pressure may explain the observed lag of seismicity increase behind the commencement of injection. The b‐value of these earthquakes increased following the wastewater injection, which is consistent with these events being injection‐induced. We conclude that the marked increase in seismicity rate in central Utah is induced by both mining activity and wastewater injection, which raised pore pressure along preexisting faults.

Matched-filter detection of the missing pre-mainshock events and aftershocks in the 2015 Gorkha, Nepal earthquake sequence

The 25 April 2015 Mw 7.8 Gorkha, Nepal earthquake occurred at the bottom edge of the locked portion of the Main Himalayan Thrust, where the Indian plate underthrusts the Himalayan wedge. The earthquake was followed by a number of large aftershocks, but not preceded by any foreshocks within ~3weeks according to the NEIC or ISC catalog. However, due to the limited station coverage of the local seismic network, a large portion of events may not be reported in routine catalogs. Here, we employ the matched filter technique to recover the undocumented earthquakes for the period beginning 80days before through 30days after the 2015 Gorkha earthquake. We detect twice as many aftershocks as those listed in the ISC catalog. We observe an along-strike aftershock expansion after the mainshock and before the largest Mw 7.3 aftershock that occurred on 12 May 2015. Repeating earthquakes are found to the ESE of the mainshock rupture area, suggesting that the expansion may be partially driven by afterslip. In addition, we observe a significant increase in seismicity rate ~3–4days prior to the mainshock, initiating ~6h after a M 5.2 earthquake, located ~240km to the NW of the mainshock. The increase of seismic activities occurs in a wide region around the Gorkha principal slip zone, indicating the effect of delayed dynamic triggering may contribute to the large-scale unloading process prior to the 2015 Gorkha mainshock.

Far-field triggering of foreshocks near the nucleation zone of the 5 September 2012 (MW 7.6) Nicoya Peninsula, Costa Rica earthquake

On 5 September 2012, a moment magnitude (MW) 7.6 earthquake occurred directly beneath the Nicoya Peninsula, an area with dense seismic and geodetic network coverage. The mainshock ruptured a portion of a previously identified locked patch that was recognized due to a decade-long effort to delineate the megathrust seismic and aseismic processes in this area. Here we conduct a comprehensive study of the seismicity prior to this event utilizing a matched-filter analysis that allows us to decrease the magnitude of catalog completeness by 1 unit. We observe a statistically significant increase in seismicity rate below the Nicoya Peninsula following the 27 August 2012 (MW 7.3) El Salvador earthquake (about 450 km to the northwest and 9 days prior to the Nicoya earthquake). Additionally, we identify a cluster of small-magnitude (<2.2) earthquakes preceding the mainshock by about 35 min and within 15 km of its hypocenter. The immediate foreshock sequence occurred in the same area as those earthquakes triggered shortly after the El Salvador event; though it is not clear whether the effect of triggering from the El Salvador event persisted until the foreshock sequence given the uncertainties in seismicity rates from a relatively small number of earthquakes. If megathrust earthquakes at such distances can induce significant increases in seismicity during the days before another larger event, this sequence strengthens the need for real-time seismicity monitoring for large earthquake forecasting.

Delayed Dynamic Triggered Seismicity in Northern Baja California, México Caused by Large and Remote Earthquakes

We analyze the seismicity in northern Baja California, Mexico, that occurred one month before and one month after the 11 March 2011 ( M w 9.1) Tohoku‐Oki, Japan, earthquake and for two other large and remote earthquakes, the 27 February 2010 central Chile ( M w 8.8) earthquake and the 11 April 2012 northern Sumatra earthquake ( M w 8.6). The northern region of Baja California exhibits high microseismic activity and moderate‐size earthquakes. The seismicity in this region is monitored by the seismic network Red Sismica del Noroeste de Mexico (RESNOM) operated by the Centro de Investigacion Cientifica y de Educacion Superior de Ensenada (CICESE). We use the RESNOM seismic catalog to search for changes in local seismic rates that occurred after the passing of surface waves generated by the three large earthquakes studied. When we compare one month of seismicity before and after the M w 9.1 Japan earthquake, in the biggest of the three events analyzed, we observe the absence of triggered seismicity in the northern Peninsular Ranges and an increase of seismicity south of the Mexicali Valley, where the Imperial fault jumps southwest and the Cerro Prieto fault continues south. We also observe an increase of seismicity rate in the Cucapah fault and in the San Pedro Martir fault regions. The Cucapah fault region, the most active of the three regions during the studied period, shows an increase of seismicity several days after the 2010 Chile ( M w 8.8) and the 2011 Japan ( M w 9.1) earthquakes. This observation can be interpreted as evidence of delayed seismicity and confirms the observation of Rubinstein et al. (2009) that earthquakes tend to be triggered in regions with high ambient seismicity rates. We observed delayed triggering of earthquakes generated by two of the three remote earthquakes analyzed. These observations can be explained by dynamic triggering of prolonged fault creep as proposed by Shelly et al. (2011). Online Material: Maps of seismicity and M c.

Forward modelling of seismicity rate changes in georeservoirs with a hybrid geomechanical–statistical prototype model

A key challenge for the development of Enhanced Geothermal Systems (EGS) is to forecast the probability of occurrence of seismic events that have the potential to damage man-made structures. Induced seismicity results from man-made time-dependent stress changes, e.g., due to fluid injection, shut-in and fluid or steam production. To accomplish a classical Probabilistic Seismic Hazard Assessment (PSHA) a catalogue of induced seismicity is required. In addition, PSHA does not return any practical recommendation for how to treat the reservoir geomechanically in order to lower the probability of occurrence of induced seismicity. Thus, we propose to link the simulated stress changes from forward geomechanical numerical reservoir models with the statistical rate-and-state approach of Dieterich (1994). Using this link we translate the modelled time-dependent stress changes into time-dependent changes of seismicity rates. This approach is general and independent of the incorporated geomechanical numerical model used. We exemplify our hybrid model approach using a geomechanical model that describes the stimulation of the well GPK4 at the EGS site in Soultz-sous-Forêts (France) including the shut-in phase. By changing the injection rate in the geomechanical model we generate various synthetic injection scenarios. With these scenarios we can study the effect on the seismicity rate and provide a recommendation for which injection experiment results in the least increase of seismicity rate. The results indicate an explicit coupling between the time-depending stress changes and the induced seismicity rate for each scenario. Even though the hybrid model cannot be used in general to derive absolute values of the rate of induced seismicity a priori (this is only possible if the geomechanical model can be calibrated against observed induced events), it serves as a tool to test the effect of stress changes on the induced seismicity rate. The approach described here is a prototype model illustrating the general workflow. In particular the geomechanical model can be replaced by any other type of reservoir description.