applied sciences

Article Acoustic Emission/Seismicity at Depth Beneath an Artificial Lake after the 2011 Tohoku Earthquake

Hirokazu Moriya

Graduate School of Engineering, Tohoku University, 6-6-04, Aramaki Aza Aoba, Aoba-ku, 980-8579, ; [email protected]; Tel.: +81-22-795-7996

 Received: 20 July 2018; Accepted: 16 August 2018; Published: 20 August 2018 

Abstract: Acoustic emission (AE)/seismicity activity increased near the city of Sendai, Japan, after the 11 March 2011 Tohoku earthquake in a newly seismically active region near the Nagamachi-Rifu fault, which caused a magnitude 5.0 earthquake in 1998. The source of this activity was around 12 km beneath an artificial lake. At the same time, activity on the Nagamachi-Rifu fault nearly ceased. More than 1550 micro-earthquakes were observed between 11 March 2011 and 1 August 2012, of which 63% exhibited similar waveforms and defined 64 multiplets. It appears that crustal extension of about 2 m during the Tohoku earthquake and additional extension of about 1 m during the following year changed the stress field in this region, thus generating micro-earthquakes and controlling their frequency. However, it has been presumed that crustal movement during the Tohoku earthquake did not affect the direction of principal stress, and that these events induced repeated quasi-static slips at asperities and the resultant micro-earthquakes.

Keywords: acoustic emission swarm; 2011 Tohoku earthquake; repeating earthquake; multiplet; crustal movement

1. Introduction Acoustic emission (AE)/seismicity is important for evaluating the stability of geological structures; for instance, the delineation of fracture systems from AE activity in geothermal reservoirs is indispensable for thermal energy extraction [1]. The stability evaluation of fractures during hydraulic stimulation in geothermal reservoirs is necessary to avoid inducing large AE events, where the relationship between large AE events and the pore-pressure distribution in reservoirs is studied [2]. The stability of geological structures is also important from the viewpoint of disaster prevention. The dynamic behavior of geological structures has been studied by many seismologists who have examined repeating earthquakes and slow slip at seismic zones [3–5]. The 11 March 2011 Tohoku earthquake and its many aftershocks brought widespread destruction to eastern Japan [6–8]. This seismic activity occurred on both offshore and inland faults. Although thrust events dominated seismicity before and during the Tohoku earthquake, eastward movement of the shallow crust of the overriding tectonic plate caused extension that relaxed the subduction-related horizontal compressional stress locally, resulting in normal movement on some inland faults after the Tohoku earthquake [9]. Close observations of seismic activity are indispensable for the monitoring of the stability of geological structures. This paper describes a seismic swarm that occurred beneath an artificial lake near the plane of an active fault in Sendai city, Japan, after the Tohoku earthquake. It has long been known that changes in the water level of reservoirs can trigger substantial earthquakes, and it is shown that the seismic activity was sensitive to the small change of tectonic stress caused by the water level of the artificial lake.

Appl. Sci. 2018, 8, 1407; doi:10.3390/app8081407 www.mdpi.com/journal/applsci Appl. Sci. 2018, 8, 1407 2 of 9

2. StudyAppl. Sci. Area 2018, 8, x 2 of 9

The2. Study study Area area is in the western part of Sendai city, , Japan, about 150 km west of the TohokuThe study earthquake area is in epicenterthe western (Figure part of 1Sendaia). Here, city,a Miyagi river terracePrefecture, is present Japan, about on both 150 km sides west of the Hiroseof River.the Tohoku A double earthquake arch dam, epicenter built (Figure across a1a). tributary Here, a toriver the terrace Hirose is River present in 1961,on both created sides aofreservoir the 2 7 3 namedHirose Lake River. Okura, A double with a arch 1.6 kmdam,surface built across area a and tributary a storage to the capacity Hirose ofRiver 2.8 in× 1961,10 m created. The a water levelreservoir is normally named maintained Lake Okura, between with a 1.6 240 km and2 surface 270 m area above and a sea storage level capacity (masl), of which 2.8 × corresponds107 m3. The to waterwater depths level of is 0normally to 30 m maintained (Miyagi Prefecture, between 240 http://www.pref.miyagi.jp/snd-dam/data/okdam- and 270 m above sea level (masl), which corresponds cyosui.htmto water). depths The Nagamachi-Rifu of 0 to 30 m (Miyagi fault Prefectu extendsre, approximatelyhttp://www.pref.miyag 15 kmi.jp/snd-dam under Sendai/data/okdam- city, striking northeast–southwestcyosui.htm). The andNagamachi-Rifu dipping about fault 45 ◦extendsto the northwestapproximately [10 ,1115]. km Estimates under Sendai of the city, stress striking field made by thenortheast–southwest multiple inverse methodand dipping [12] indicateabout 45° that to the no directionrthwest of[10,11]. minimum Estimates principal of the stressstress isfield vertical made by the multiple inverse method [12] indicate that the direction of minimum principal stress is and the direction of maximum principal stress varies from east–west to northwest–southeast. vertical and the direction of maximum principal stress varies from east–west to northwest–southeast.

FigureFigure 1. 1.(a ()a) Map Map of northeastern northeastern Honshu, Honshu, Japan, Japan, with inset with showing inset location showing of study location area. of The study solid area. The solidstar marks star the marks epicenter the of epicenter the Tohoku of earthquake; the Tohoku (b) Map earthquake; showing epicenters (b) Map of showing micro-earthquakes epicenters of from 5 February 2005 to 1 August 2012; (c) Vertical cross section showing hypocenters of micro- micro-earthquakes from 5 February 2005 to 1 August 2012; (c) Vertical cross section showing earthquakes projected onto line A–B. Open stars show the location of the magnitude 5.0 earthquake hypocenters of micro-earthquakes projected onto line A–B. Open stars show the location of the of 15 September 1998. magnitude 5.0 earthquake of 15 September 1998. At Sendai city, very high seismic intensities (greater than 6 on the Japan Meteorological Agency Atscale) Sendai were city, recorded veryhigh during seismic the Tohoku intensities eart (greaterhquake. thanGPS 6(Global on the JapanPositioning Meteorological System) data Agency scale)documented were recorded crustal during movements the Tohoku of more earthquake. than 5 m GPSat the (Global coast and Positioning about 2.5 System)m in Sendai data city. documented The crustalNagamachi-Rifu movements of fault more is known than 5 to m be at theactive coast in the and Sendai about area 2.5 [10,13,14]. m in Sendai Seismic city. activity The Nagamachi-Rifu increased near this fault after the Tohoku earthquake, and more than 1550 events were observed between 11 fault is known to be active in the Sendai area [10,13,14]. Seismic activity increased near this fault March 2011 and 1 August 2012. A micro-earthquake is physically the same to as an AE and the term after the Tohoku earthquake, and more than 1550 events were observed between 11 March 2011 and 1 August 2012. A micro-earthquake is physically the same to as an AE and the term “AE” is generally Appl. Sci. 2018, 8, 1407 3 of 9

Appl. Sci. 2018, 8, x 3 of 9 used for much smaller seismic events with smaller magnitude, and this paper discusses the stability of geological“AE” is structuresgenerally used using for micro-earthquakes much smaller seismic in the events same with manner. smaller magnitude, and this paper discusses the stability of geological structures using micro-earthquakes in the same manner. 3. Tectonic Conditions before and after the Tohoku Earthquake 3. Tectonic Conditions before and after the Tohoku Earthquake On 15 September 1998, an earthquake of magnitude (M) 5.0 occurred at a 12.4 km depth, on the deepestOn part 15 of September the Nagamachi-Rifu 1998, an earthquake fault [14]. of The magnitud mechanisme (M) of5.0 this occurred earthquake at a 12.4 and km its depth, aftershocks on the was reversedeepest faulting, part of and the Nagamachi-Rifu the P-axes of the fault events [14]. described The mechanism a line orientedof this earthquake east–west and to northwest–southeast;its aftershocks was reverse faulting, and the P-axes of the events described a line oriented east–west to northwest–southeast; the maximum principal stress direction in the deep part of the active fault was northwest–southeast. the maximum principal stress direction in the deep part of the active fault was northwest–southeast. On the basis of GPS data from an observation point about 6 km southeast of the dam, the On the basis of GPS data from an observation point about 6 km southeast of the dam, the Geospatial Information Authority of Japan (GSI) reported crustal movement of 2.55 m eastward and Geospatial Information Authority of Japan (GSI) reported crustal movement of 2.55 m eastward and 16 cm16 cm ground ground subsidence subsidence during during the the Tohoku Tohoku earthquakeearthquake (e.g., [15]). [15]). The The GSI GSI also also measured measured eastward eastward crustalcrustal movements movements of of 1.85 1.85 m m in in Tendo Tendo city,city, 3030 kmkm toto the west, and and 3.98 3.98 m m in in Higashi- Higashi-Matsushima city,city, 70 70 km km to to the the east east (GSI, (GSI, http://www.gsi.go.jp/index.html http://www.gsi.go.jp/index.html; Figure; Figure 1a).1 Froma). From these thesedata plus data the plus theassumption assumption of of an an elastic elastic rock rock mass mass in inthe the area, area, the the strain strain change change resulting resulting from from the the Tohoku Tohoku −5 earthquakeearthquake was was estimated estimated to be be 2.13 2.13 × ×10−105 in thein study the studyarea. Eastward area. Eastward crustal movement crustal movement around aroundSendai Sendai continued continued after afterthe earthquake, the earthquake, reaching reaching about about 1 m 1during m during the thefollowing following year year (GSI; (GSI; http://www.gsi.go.jp/cais/chikakuhendo40012.htmlhttp://www.gsi.go.jp/cais/chikakuhendo40012.html). ). It appearsIt appears that that the the direction direction of of maximum maximum principalprincipal stress stress in in the the study study area area was was roughly roughly east–west east–west bothboth before before and and after after the the Tohoku Tohoku earthquakeearthquake [7]. [7]. However, However, at at the the Kamaishi Kamaishi mine, mine, ≈170≈ km170 northeast km northeast of of Sendai,Sendai, horizontal displacement displacement of of 3.3 3.3 m m was was obse observedrved during during the the Tohoku Tohoku earthquake earthquake [16]. [16 In]. situ In situ measurementsmeasurements of tectonicof tectonic stress stress at theat the mine mine documented documented a large a large change change in the in stressthe stress field field in the in shallow the crustshallow in the crust two yearsin the after two theyears Tohoku after the earthquake, Tohoku earthquake, with the principal with the stress principal decreasing stress decreasing in the east–west in the east–west direction and increasing in the north–south direction [16]. direction and increasing in the north–south direction [16]. The evidence suggests that although the Tohoku earthquake caused large changes of the stress The evidence suggests that although the Tohoku earthquake caused large changes of the stress field field in some parts of the shallow crust, changes in the orientation of the stress field at seismogenic in some parts of the shallow crust, changes in the orientation of the stress field at seismogenic depths were depths were small in the study area, which implies that the stress field in the study area still favors small in the study area, which implies that the stress field in the study area still favors reverse faulting. reverse faulting. 4. AE Activity before and after the Tohoku Earthquake 4. AE Activity before and after the Tohoku Earthquake A largeA large number number of micro-earthquakesof micro-earthquakes were were recorded recorded in in western western Sendai Sendai city city from from 5 February5 February 2005 to 12005 August to 1 2012,August a period 2012, spanninga period thespanning time of the the time Tohoku of the earthquake Tohoku earthquake (Figure1). The (Figure distribution 1). The of hypocentersdistribution defines of hypocenters two areas defines of AE two activity, areas largely of AE activity, corresponding largely corresponding to seismic events to seismic before events and after thebefore Tohoku and earthquake, after the Tohoku respectively earthquake, (Figures respectively2 and3). (Figures 2 and 3).

Figure 2. Vertical cross section showing hypocenters of micro-earthquakes before the Tohoku Figure 2. Vertical cross section showing hypocenters of micro-earthquakes before the Tohoku earthquake projected onto the vertical plane defined by line A–B (location in Figure 1b). The dashed earthquake projected onto the vertical plane defined by line A–B (location in Figure1b). The dashed line shows the downward projection of the Nagamachi-Rifu fault. The red star represents the source line shows the downward projection of the Nagamachi-Rifu fault. The red star represents the source location of the earthquake of magnitude (M) 5.0 occurred on 15 September 1998. location of the earthquake of magnitude (M) 5.0 occurred on 15 September 1998. Appl. Sci. 2018, 8, 1407 4 of 9 Appl. Sci. 2018, 8, x 4 of 9

FigureFigure 3.3.Vertical Vertical cross cross section section showing showing hypocenters hypocenters of micro-earthquakes of micro-earthquakes after the Tohokuafter the earthquake Tohoku projectedearthquake onto projected the vertical onto plane the vertic definedal plane by line defined A–B. The by line red starA–B. represents The red star the sourcerepresents location the source of the earthquakelocation of the of magnitude earthquake (M) of magnitude 5.0 occurred (M) on 5.0 15 Septemberoccurred on 1998. 15 September 1998.

TheThe AEAE eventsevents beforebefore thethe TohokuTohoku earthquakeearthquake occurredoccurred atat depthsdepths betweenbetween 88 andand 1414 kmkm onon thethe Nagamachi-RifuNagamachi-Rifu fault.fault. However, after the earthquake, this activity ceased and a newnew activeactive zonezone appearedappeared to the west, generating generating hundreds hundreds of of ev eventsents between between 11 11 March March 2011 2011 and and 1 August 1 August 2012 2012 at atdepths depths of of8 8to to 14 14 km. km. Activity Activity in in the the study study area area did did not not increase increase immediately after the TohokuTohoku earthquake,earthquake, butbut beganbegan aa gradualgradual riserise onon 33 AprilApril andand totaledtotaled 1111 eventsevents (maximum(maximum magnitudemagnitude 2.3)2.3) byby 2121 April.April. AfterAfter 24 24 April, April, AE AE activity activity increased increased further, further, and and by 1 by August 1 August 2012, 2012, more thanmore 1550 than AE 1550 events AE hadevents occurred. had occurred. It is known It is known that smaller that smaller earthquakes earthquakes (e.g., aftershocks) (e.g., aftershocks) occur afteroccur aafter previous a previous large earthquakelarge earthquake in the in same the same area ofarea the of main the main shock, shock, and theand frequency the frequency of aftershocks of aftershocks decreases decreases with with the reciprocalthe reciprocal of time of time after after the mainthe main shock shock (Omori’s (Omori’s law). law). However, However, clear clear mainshock mainshock and and aftershocks aftershocks are notare identifiednot identified after theafter Tohoku the Tohoku earthquake earthquake in Figure in4 ;Figure thus, the4; thus, typical the decay typical of earthquake decay of earthquake swarms is difficultswarms tois difficult discern. to The discern. relationship The relationship between the between event activitythe event and activity the seasonal and the variationsseasonal variations in water levelin water in other level yearsin other were years investigated. were investigated. However, However, the clear the relationship clear relationship as seen in as Figure seen in4 wasFigure not 4 observedwas not observed before the before Tohoku the earthquake Tohoku earthquake and after and the studiedafter the period, studied and period, the seismic and the activity seismic was activity very lowwas invery 2017. low Figure in 2017.5 shows Figure the 5 Gutenberg-Richtershows the Gutenberg- relationRichter before relation and after before 11 Marchand after 2011. 11 TheMarchb-value 2011. wasThe 1.0b-value before was 11 March1.0 before and 11 0.92 March after and 11 March, 0.92 after indicating 11 March, that indicating larger events that became larger events more frequent became aftermore the frequent Tohoku after earthquake. the Tohoku earthquake. AmongAmong thethe micro-earthquakesmicro-earthquakes after 11 MarchMarch 2011,2011, wewe identifiedidentified eventsevents thatthat sharedshared similarsimilar waveformswaveforms (Figure (Figure6), known6), known as repeating as repeating earthquakes earthquakes or multiplets or multiplets [ 17–19]. By[17–19]. calculating By coherencescalculating forcoherences all possible for all pairs possible of events, pairs and of events, using coherenceand using greatercoherence than greater 0.98 as than the 0.98 criterion as the for criterion similarity, for wesimilarity, identified we 985 identified similar 985 events similar (63% events of those (63% observed), of those which observed), defined which 64 multiplets. defined 64 Travel-time multiplets. differencesTravel-time of differences P- and S-waves of P- wereand determinedS-waves were by cross-correlatingdetermined by cross-correlating the P- and S-components the P- and with S- theircomponents master events.with their master events. TheThe meanmean andand standardstandard deviationdeviation ofof thethe travel-timetravel-time differencesdifferences betweenbetween P-P- andand S-wavesS-waves werewere 0.0240.024 andand 0.04 s, respectively. SourceSource locationslocations werewere notnot corrected according toto travel-time differences becausebecause thethe numbernumber ofof seismicseismic stationsstations waswas insufficientinsufficient toto properlyproperly constrainconstrain them.them. However,However, thethe similaritysimilarity ofof P-P- andand S-waveS-wave traveltravel timestimes suggestssuggests thatthat thethe multipletmultiplet sourcessources werewere closeclose togethertogether onon adjacentadjacent fractures.fractures. DuringDuring thethe periodperiod followingfollowing thethe TohokuTohoku earthquake,earthquake, eighteight earthquakesearthquakes largerlarger thanthan MM 3.03.0 werewere observedobserved atat moremore thanthan 4040 stations,stations, makingmaking itit possiblepossible toto estimateestimate theirtheir sourcesource mechanismsmechanisms byby usingusing P-waveP-wave polaritypolarity distributions.distributions. Figure7 7aa showsshows thethe faultfaultplane plane solution solutionof of the the largest largest of of these these events events (M(M 3.43.4 on on 3 June3 June 2011), 2011), for whichfor which only P-waveonly P-wave polarities polarities that were that clearly were detected clearly bydetected visual inspectionby visual wereinspection used towere determine used to the determine nodal planes. the nodal The planes other seven. The other of these seven events of these had the events same had mechanism, the same mechanism, indicating that the fault planes strike north-northeast–south-southwest and dip at about 65°, and that the slip included both left-lateral strike-slip and reverse components. The accuracy of Appl. Sci. 2018, 8, 1407 5 of 9 indicating that the fault planes strike north-northeast–south-southwest and dip at about 65◦, and thatAppl.Appl. the Sci. Sci. slip 2018 2018, included8, 8x, x both left-lateral strike-slip and reverse components. The accuracy5 of of5 of9 the 9 source locations was not enough to identify the orientation of the fault plane which caused the micro-earthquakes.thethe source source locations locations If was thewas not source not enough enough locations to to identify identify were the approximatedthe orientation orientation of toof the the a plane,fault fault plane plane the planewhich which wascaused caused dipping the the micro-earthquakes. If the source locations were approximated to a plane, the plane was dipping aroundmicro-earthquakes. 45 degrees to the If eastthe andsource striking locations nearly were north-south, approximated and to this a plane, orientation the plane was similarwas dipping to one aroundaround 45 45 degrees degrees to to the the east east and and striking striking nearly nearly north-south, north-south, and and this this orientation orientation was was similar similar to to one one of the nodal planes in Figure7a. The orientation of nodal planes for the mainshock of the 1998 M 5.0 ofof the the nodal nodal planes planes in in Figure Figure 7a. 7a. The The orientation orientation of of nodal nodal planes planes for for the the mainshock mainshock of of the the 1998 1998 M M 5.0 5.0 earthquake (Figure7b) [ 14] was similar to those of the eight M > 3 events (Figure7a), but included a earthquakeearthquake (Figure (Figure 7b) 7b) [14] [14] was was si milarsimilar to to those those of of the the eight eight M M > 3> events3 events (Figure (Figure 7a), 7a), but but included included a a right-lateral strike-slip component. right-lateralright-lateral strike-slip strike-slip component. component.

FigureFigure 4. 4.(a )( aNumber) Number of of micro-earthquakes micro-earthquakes per per day day and and (b ()b magnitudes) magnitudes and and cumulative cumulative number number of of Figure 4. (a) Number of micro-earthquakes per day and (b) magnitudes and cumulative number of seismicseismic events events (N) (N) during during the the 360 360 days days before before and and 600 600 days days after after the the Tohoku Tohoku earthquake earthquake along along with with seismic events (N) during the 360 days before and 600 days after the Tohoku earthquake along with the thethe water water level level of of Lake Lake Okura. Okura. Th The recorde record ends ends on on 1 August1 August 2012. 2012. water level of Lake Okura. The record ends on 1 August 2012.

FigureFigureFigure 5. 5.Histograms 5.Histograms Histograms showingshowing showing the the Gu Gutenberg-Richter Gutenberg-Richtertenberg-Richter relation.relation. relation. Appl. Sci. 2018, 8, 1407 6 of 9 Appl. Sci. 2018, 8, x 6 of 9 Appl. Sci. 2018, 8, x 6 of 9

Figure 6. Example of a set of waveforms constituting a multiplet. Amplitudes are normalized to the Figure 6. Example of a set of waveforms constituting a multiplet. Amplitudes are normalized to the maximumFigure 6. Example amplitude of aof set each of waveform.waveforms Theconstituting vertical axis a multiplet. represents Amplitudes the number are of normalized events. to the maximum amplitude of each waveform. The verticalvertical axis represents the number of events.events.

(a) (b) (a) (b) Figure 7. (a) Waveforms of eight M > 3 earthquakes after the Tohoku earthquake and fault plane Figure 7. (a) Waveforms of eight M > 3 earthquakes after the Tohoku earthquake and fault plane solutionFigure 7. for(a the) Waveforms largest event of (M eight 3.4, M 3 >June 3 earthquakes 2011). Open aftercircle theand Tohoku closed circ earthquakele represent and negative fault plane and solution for the largest event (M 3.4, 3 June 2011). Open circle and closed circle represent negative and positivesolution P-wave for the largest polarities, event respectively; (M 3.4, 3 June (b) 2011). Fault Openplane circlesolution and for closed the 1998 circle M represent 5.0 earthquake negative on andthe positive P-wave polarities, respectively; (b) Fault plane solution for the 1998 M 5.0 earthquake on the Nagamachi-Rifupositive P-wave polarities,fault, which respectively; was reproduced (b) Fault following plane solution the result for by the Umino 1998 M et 5.0al. earthquake[14]. on the Nagamachi-Rifu fault, which was reproduced followingfollowing thethe resultresult byby UminoUmino etet al.al. [[14].14]. 5. Discussion 5. Discussion East–west crustal extension during the Tohoku earthquake relaxed the subduction-related horizontalEast–west compressional crustal extension stress in during the study the area. Tohoku The extension earthquake decreased relaxed normal the subduction-relatedsubduction-related stresses on east- andhorizontal west-dipping compressional fault planes stress and, in the because study sheararea. The stress extension on the reverse decreased decreased faults normal normal decreased, stresses stresses they on on wereeast- east- dynamicallyand west-dipping stabilized fault against planes shearand, becauseslip. Thus, shear the st stressnumberress on of the normal reverse faulting faults events decreased, increased they wereafter thedynamically Tohoku earthquakestabilized against [7], and shear earthquakes slip. Thus, on the the number Nagamachi-Rifu of normal faulting fault nearly events ceased increased after after 11 Marchthe Tohoku 2011, asearthquake horizontal [7], compressional and earthquakes stress onon the faultNagamachi-Rifu surface decreased. fault nearly ceased after 11 March 2011, as horizontal compressional stress on the fault surface decreased. Appl. Sci. 2018, 8, 1407 7 of 9 dynamically stabilized against shear slip. Thus, the number of normal faulting events increased after the Tohoku earthquake [7], and earthquakes on the Nagamachi-Rifu fault nearly ceased after 11 March 2011, as horizontal compressional stress on the fault surface decreased. After the Tohoku earthquake, seismic activity increased in a region 6 km west of the source of the 1998 earthquake (Figure3), apparently on newly activated fractures. Fault plane solutions (Figure7a) indicate that the newly activated fractures dip either east or west and slipped with left-lateral strike-slip and reverse mechanisms. Seismic activity on these fractures, which were previously under critical conditions for reverse faulting, would be expected to decrease when horizontal compressional stress was relaxed during the Tohoku earthquake and lowered the shear stress on the fault plane. However, seismic activity increased after the earthquake, so we must seek other contributing factors. One possible explanation is a relative increase of principal stress in the north–south direction [16]. This would have created conditions favoring left-lateral strike-slip faulting. Fault plane solutions after the Tohoku earthquake (Figure7a) include both left-lateral and reverse components, whereas for the 1998 earthquake (Figure7b), although a left-lateral component was present, reverse faulting was dominant. This result suggests that a relative increase of principal stress in the north–south direction may have also increased the left-lateral component, but the Tohoku earthquake did not dramatically change the contrast between the principal east–west and north–south stresses. This observation implies that the differential stress was greater before the Tohoku earthquake, and that the Tohoku earthquake did not significantly affect the stress field in the study area. The triggering of earthquakes is sensitive to changes in stress when conditions are critical for shear slip. For example, tidal changes are a likely factor in triggering certain seismic events [20] even though the stress change due to earth tides is no greater than about 103 Pa [21]. On the other hand, the water level of Lake Okura changes by about 20 m, which corresponds to a change in overburden pressure of 200 × 103 Pa. This increase in the overburden stress would increase the pore pressure in the fractures under the lake. AE events can be triggered by increased pore pressure due to fluid diffusion (or pressure diffusion) associated with increases of water level in a lake or reservoir, and it is known that small changes of pore pressure can trigger earthquakes [1,22]. Earthquakes triggered by this mechanism have been reported at several kilometers depth beneath a reservoir in the Koyna district of India [23]. Here, micro-earthquakes started when the water level of Lake Okura approached its maximum level of 270 masl, and the rate of events peaked at 25/day just after that water level was reached. Although those facts do not establish a correlation, fluctuations of the lake level should not be precluded as a possibility in explaining the swarm. The increase of seismic activity after the Tohoku earthquake can be explained as follows. Before the earthquake, strain energy was accumulating in fractures such that they reached a critical condition for shear slip before 11 March 2011, but cohesion on the fracture surfaces kept the fault locked. Given that crustal movement at this time was a few centimeters per year, cohesive strengthening would have been aseismic. In the process that increases frictional resistance to sliding, generally referred to as healing, micromechanical processes (e.g., increases of asperity contact areas) result in cohesive strengthening [24]. Crustal extension during the Tohoku earthquake changed the direction of shear stress on fractures and weakened the contacts across fracture planes, thus activating shear slip at asperities that were under critical conditions for shear slip. It appears that quasi-static slips occurred repeatedly in the newly activated fracture zone after the Tohoku earthquake, causing AE events as great as M 3.4. This interpretation is supported by the observation of many AE events with similar waveforms (accounting for 63% of located micro-earthquakes) and the decrease after 11 March of the gradient of the high-magnitude tail of earthquake magnitude distributions (Figure5).

6. Conclusions About 40 days after the 2011 Tohoku earthquake, a micro-earthquake swarm began west of Sendai, Japan, centered about 8–14 km deep beneath an artificial lake near the down-dip projection of an active fault that caused an M 5 earthquake in 1998. Observations suggest that the new fractures were Appl. Sci. 2018, 8, 1407 8 of 9 seismically activated, even though fracturing should have been stabilized by a decrease of principal stress in the east–west direction that was caused by the Tohoku earthquake. It is typically presumed that crustal movement during the Tohoku earthquake did not affect the direction of principal stress; however, a relative increase of the principal stress magnitude in the north–south direction in the study area may have caused a change in the direction of shear stress on fracture surfaces, thus weakening the contacts at asperities. These events induced repeated quasi-static slips at asperities and the resultant AE events. The magnitudes and frequency of the micro-earthquakes indicate that these fractures were capable of causing micro-earthquakes with magnitudes of about 4. The results of this study indicate that ongoing monitoring of the earthquake swarm is important for mitigation of damage caused by micro-earthquakes at these newly activated inland faults. The findings of this study may be helpful as a case history to understand the stability of geological structures in at the kilometer scale.

Author Contributions: H.M. has performed the works of conceptualization, methodology, writing–original draft preparation, writing–review and editing. Funding: This research received no external funding. Acknowledgments: The earthquake waveform data and source location data used in this study were acquired from the Hi-net website of the National Research Institute for Earth Science and Disaster Prevention (NIED), Japan. Conflicts of Interest: The authors declare no conflicts of interest.

References

1. Mukuhira, Y.; Moriya, H.; Ito, T.; Asanuma, H.; Häring, M. Pore pressure migration during hydraulic stimulation due to permeability enhancement by low-pressure subcritical fracture slip. Geophys. Res. Lett. 2017, 44, 3109–3118. [CrossRef] 2. Moriya, H.; Naoi, M.; Nakatani, M.; van Aswegen, G.; Murakami, O.; Kgarume, T.; Ward, A.K.; Durrheim, R.J.; Philipp, J.; Yabe, Y.; et al. Delineation of large localized damage structures forming ahead of an active mining front by using advanced acoustic emission mapping techniques. Int. J. Rock Mech. Min. Sci. 2015, 79, 157–165. [CrossRef] 3. Talwani, P. On the nature of reservoir-induced seismicity. Pure Appl. Geophys. 1977, 150, 473–492. [CrossRef] 4. Nakajima, J.; Uchida, N. Repeated drainage from megathrusts during episodic slow slip. Nat. Geosci. 2018, 11, 351–356. [CrossRef] 5. Ariyoshi, K.; Uchida, N.; Matsuzawa, T.; Hino, R.; Hasegawa, A.; Hori, T.; Kaneda, Y. A trial estimation of frictional properties, focusing on aperiodicity off Kamaishi just after the 2011 Tohoku earthquake. Geophys. Res. Lett. 2014, 41, 8325–8334. [CrossRef] 6. Hirose, F.; Miyaoka, K.; Hayashimoto, N.; Yamazaki, T.; Nakamura, M. Outline of the 2011 off the Pacific coast of Tohoku Earthquake (Mw 9.0)—Seismicity: Foreshocks, mainshock, aftershocks, and induced activity. Earth Planets Space 2011, 63, 513–518. [CrossRef] 7. Ide, S.; Baltay, A.; Beroza, G.C. Shallow dynamic overshoot and energetic deep rupture in the 2011 Mw 9.0 Tohoku-Oki Earthquake. Science 2011, 332, 1426–1429. [CrossRef][PubMed] 8. Sato, M.; Ishikawa, T.; Ujihara, N.; Yoshida, S.; Fujita, M.; Mochizuki, M.; Asada, A. Displacement above the hypocenter of the 2011 Tohoku-Oki earthquake. Science 2011, 332, 1395. [CrossRef][PubMed] 9. Yoshida, K.; Hasegawa, A.; Okada, T.; Iinuma, T.; Ito, Y.; Asano, Y. Stress before and after the 2011 great Tohoku-oki earthquake and induced earthquakes in inland areas of eastern Japan. Geophys. Res. Lett. 2012, 39, L03302. [CrossRef] 10. Nakamura, A.; Asano, Y.; Hasegawa, A. Estimation of deep fault geometry of the Nagamachi-Rifu fault from seismic array observations. Earth Planets Space 2002, 54, 1027–1031. [CrossRef] 11. Nakajima, J.; Hasegawa, A.; Horiuchi, S.; Yoshimoto, K.; Yoshida, T.; Umino, N. Crustal heterogeneity around the Nagamachi-Rifu fault, northeastern Japan, as inferred from travel-time tomography. Earth Planets Space 2006, 58, 843–853. [CrossRef] 12. Yamaji, A. The multiple inverse method: A new technique to separate stresses from heterogeneous fault-slip data. J. Struct. Geol. 2000, 22, 441–452. [CrossRef] 13. Umino, N.; Ujikawa, H.; Hori, S.; Hasegawa, A. Distinct S-wave reflectors (bright spots) detected beneath the Nagamachi-Rifu fault, NE Japan. Earth Planets Space 2002, 54, 1021–1026. [CrossRef] Appl. Sci. 2018, 8, 1407 9 of 9

14. Umino, N.; Okada, T.; Hasegawa, A. Foreshock and aftershock sequence of 1998 M5.0 Sendai, northeastern Japan, earthquake and its implications for earthquake nucleation. Bull. Seismol. Soc. Am. 2002, 92, 2465–2477. [CrossRef] 15. Ozawa, S.; Nishimura, T.; Suito, H.; Kobayashi, T.; Tobita, M.; Imakiire, T. Coseismic and postseismic slip of the 2011 magnitude-9 Tohoku-Oki earthquake. Nature 2011, 475, 373–376. [CrossRef][PubMed] 16. Sakaguchi, K.; Koyano, T.; Yokoyama, T. Change in stress field in the Kamaishi Mine, associated with the 2011 Tohoku-oki earthquake. In Proceedings of the 13th Japan Symposium on Rock Mechanics, Okinawa, Japan, 9–11 January 2013; pp. 513–517. (In Japanese) 17. Moriya, H.; Niitsuma, H.; Baria, R. Multiplet-clustering analysis reveals structural details within the seismic cloud at the Soultz geothermal field, . Bull. Seismol. Soc. Am. 2003, 93, 1606–1620. [CrossRef] 18. Nadeau, R.M.; McEvilly, T.V. Periodic pulsing characteristic microearthquakes on the San Andreas fault. Science 2004, 303, 220–222. [CrossRef][PubMed] 19. Igarashi, T. Spatial changes of inter-plate coupling inferred from sequences of small repeating earthquakes in Japan. Geophys. Res. Lett. 2010, 37, L20304. [CrossRef] 20. Chen, H.J.; Chen, C.C.; Tseng, C.Y.; Wang, J.H. Effect of tidal triggering on seismicity in Taiwan revealed by the empirical mode decomposition method. Nat. Hazards Earth Syst. Sci. 2012, 12, 2193–2202. [CrossRef] 21. Tsuruoka, H.; Ohtake, M.; Sato, H. Statistical test of the tidal triggering of earthquakes: Contribution of the ocean tide loading effect. Geophys. J. Int. 1995, 122, 183–194. [CrossRef] 22. Talwani, P.; Cobb, J.S.; Schaeffer, M.F. In situ measurements of hydraulic properties of a shear zone in northwestern South Carolina. J. Geophys. Res. 1999, 104, 14993–15003. [CrossRef] 23. Talwani, P. Seismogenic properties of the crust inferred from recent studies of reservoir-induced seismicity–Application to Koyna. Curr. Sci. 2000, 79, 1327–1333. 24. Scholz, C.H.; Engelder, T. Role of asperity indentation and ploughing in rock friction. Int. J. Rock Mech. Min. Sci. 1976, 13, 149–154. [CrossRef]

© 2018 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).