Evolution of crustal deformation in the northeast–central Japanese island arc Insights from fault activity Received 6 October 2015 Revised 15 August 2016 Accepted 30 September 2016 DO I 10 1111/iar 12[.]
Received: October 2015 Revised: 15 August 2016 Accepted: 30 September 2016 DOI 10.1111/iar.12179 RESEARCH ARTICLE Evolution of crustal deformation in the northeast–central Japanese island arc: Insights from fault activity Ayumu Miyakawa | Makoto Otsubo Geological Survey of Japan, Tsukuba 305‐8567, Japan Abstract Correspondence Ayumu Miyakawa, Geological Survey of Japan, AIST, Tsukuba Central 7, 1‐1‐1 Higashi, Tsukuba 305‐8567, Japan Email: miyakawa‐a@aist.go.jp We evaluated fault activity in northeast–central Japan based on fault orientation, regional stress field, and slip tendency analysis for active and non‐active faults (i.e faults for which Quaternary activity has not been identified) Slip tendency is generally higher along active faults than non‐ active faults, although a high slip tendency was observed along some non‐active faults, indicating their potential to become active The potential for fault activity along non‐active faults can be modeled using the temporal evolution from non‐active to active during long‐term crustal deformation The density of potentially active faults varies spatially across the study areas and reflects the temporal evolution of crustal deformation in northeast–central Japan KEY W ORDS active fault, eastern margin of the Japan Sea, Niigata–Kobe Tectonic Zone, non‐active fault, slip tendency, stress inversion | I N T RO D U CT I O N during the late Pliocene to early Quaternary (e.g Sato, 1994; Taira, 2001; Tsutsumi, Sato, & Yamaji, 2012) Previous surveys have revealed Understanding the evolution of crustal deformation in subduction prominent deformation zones in Japan dating from the late Pliocene to zones is important for the reconstruction of geodynamic processes present, including the eastern margin of the Japan Sea (EMJS) and the Long‐term deformation in subduction zones is thought to be Niigata–Kobe Tectonic Zone (NKTZ) (Sagiya, Miyazaki, & Tada, 2000; associated with fault activity (e.g Mazzotti, Henry, & Le Pichon, Okamura, 2002) (Figure 1) Okamura (2002) reports that structural 2001) Furthermore, it is well known that one of the dominant controls deformation, such as reverse faulting and folding, are localized along on crustal deformation is tectonic stress First‐order tectonic stress is former rifts as a result of inversion tectonics along the EMJS Data from concordant in its principal orientations with relative plate motion a continuous Global Navigation Satellite System (GNSS) array indicate directions (Zoback, 1992) The agreement of instantaneous plate that the NKTZ is a region of high strain rate (> 0.1 ppm/year) along motions with long‐term plate motions determined from marine the coast of the Japan Sea and in the northern Chubu and Kinki districts magnetic anomalies as old as about 106 years (e.g Argus & Gordon, of central Japan (Sagiya et al., 2000) The key difference between the 1990) suggests that first‐order tectonic stress has been associated EMJS and the NKTZ is the time frame over which deformation can be with plate motion for as long as ca 10 years Hence, a study of the observed, over geological (long) and geodetic (short) time frames for long‐term evolution of crustal deformation should consider fault the former and the latter, respectively Both time frames are critical activity in the context of the long‐term (ca 106 years) stress field for understanding the temporal and spatial evolution of crustal defor- The Japanese island arc has been intensively studied in terms of geodynamics and crustal deformation along a subduction zone where mation in Japan Here we use fault activity as a proxy for the geodynamic evolution of the crust in northeast–central Japan intraplate deformation and earthquakes are common (e.g Taira, 2001) Studies of active faults have revealed not only the activity of (Figure 1) The stress regime of Japan’s present‐day island arc system individual faults but also the temporal evolution of fault activity in is dominated by east–west compression that is thought to have begun Japan For example, Doke et al (2012) compiled chronological data This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited © 2017 The Authors Island Arc published by John Wiley & Sons Australia Ltd Island Arc 2017;e12179 https://doi.org/10.1111/iar.12179 wileyonlinelibrary.com/journal/iar of of MIYAKAWA AND OTSUBO FIGURE Map showing the tectonic setting of the Japanese Islands The Pacific and Philippine Sea plates are subducting beneath the North American and Amur plates, respectively, and black arrows represent their relative plate motions (Demets, Gordon, & Arugus, 2010) Orange areas mark the intense deformation zones in the eastern margin of the Japan Sea (EMJS) (Okamura, 2002) The green area marks the Niigata–Kobe Tectonic Zone (NKTZ) (Sagiya et al., 2000) Blue dots are the epicenters of earthquakes with magnitudes > (1 January 1997 to 10 March 2011 at depths ≤ 30 km) from the NIED F‐net catalog (http://www.fnet.bosai.go.jp/top.php? LANG=en) The black rectangle outlines the study area for active faults in Japan and determined that the initiation of faulting spans the Pliocene and Pleistocene The range in the ages of fault initiation indicates that present‐day active faults were not simultaneously activated at the onset of the present‐day stress regime (Doke et al., 2012), but instead that fault activity has evolved over the past several million years In this study, we define a “non‐active fault” as a fault that has not been recognized as being active in the Quaternary, whereas we define an “active fault” as a fault recognized as being active in the Quaternary We sometimes overlook the potential for fault activity, as shown by recent large hazardous earthquakes in Japan that occurred along the causative fault that had not been identified prior to the earthquake (e.g the 2008 Iwate–Miyagi Earthquake; Ohta et al., 2008) Recent activity on a fault may be overlooked in the case of faults with small cumulative displacement (Doke et al., 2012) In this study, such faults are called “overlooked active faults” (Figure 2) Furthermore, chrono- FIGURE Definitions of faults as used in this study and relationships between them Observed faults are divided into active (pink area) and non‐active faults (blue areas) according to their activity in the Quaternary (e.g The Research Group for Active Faults of Japan, 1991) The non‐active faults are divided into three categories: overlooked active faults (light blue) for which the record of activity is unknown, future active faults (blue) that are inactive but may be activated in the near future (i.e < 106 years), and completely inactive faults (dark blue) that are not expected to be active in the future The overlooked and future active faults are considered potential active faults The length of the bars is proportional to the number of faults in each category (Figure 7) logical data for a fault can indicate that it has been inactive in the late Quaternary, but it may be activated in the near future (here called a temporal and spatial evolution of crustal deformation in northeast– “future active fault”) (Figure 2) Therefore, a potentially active fault is central Japan defined as a fault that is either an overlooked or future active fault This study uses a physical method based on Mohr–Coulomb criteria to determine fault activity for both active and non‐active faults, | DATASET AND METHODS as opposed to the chronological methods employed by Doke et al (2012) Our method uses fault orientation data and the proximal stress We evaluated fault activity in northeast–central Japan (i.e the Tohoku, field, meaning that it can be applied in evaluating present‐day and Chubu, and Kinki areas; Figure 1) including the EMJS and NKTZ using future fault activity without the need for data on the timing of fault slip tendency analysis (Morris, Ferrill, & Henderson, 1996) based on the activity This study evaluates fault activity to gain insight into the orientation of faults and the regional stress field We considered both of MIYAKAWA AND OTSUBO active and non‐active faults We obtained the locations and orienta- and Kanto–Chubu to Kinki area Hence, the long‐term effect of the tions of 317 active faults in the study area from the Active Fault Data- Tohoku earthquake is thought to be limited in our target area, and base of Japan, compiled by the National Institute of Advanced we consider the stress state obtained in our analysis to be suitable Industrial Science and Technology (https://gbank.gsj.jp/activefault/ for evaluation index_e_gmap.html) We also obtained fault orientations of 354 non‐ We used slip tendency analysis (Morris et al., 1996) to evaluate active faults from a database compiled by Kosaka, Kanaori, Chigira, the likelihood of fault slip For a cohesionless fault, the slip tendency and Yoshida (2011) based on the Geological Maps of Japan (e.g., on a planar fault surface is defined as the ratio of shear stress to normal Geological Survey of Japan, 1989), comprising representative stress: Ts = τ (μσn)−1, where Ts is the slip tendency (Morris et al., 1996), structures, rather than all existing faults, in each region τ is the shear stress acting on the fault plane, σn is the normal stress We estimated the regional stress state in the study area The area acting on the fault plane, and μ is the friction coefficient of the fault that the estimated stress state covers is the same as the area of the plane The likelihood of slip is high when the slip tendency is high, Geological Maps of Japan from which the non‐active fault and vice versa The stress field derived by the inversion of earthquake database was constructed The regional stress field was estimated focal mechanisms does not provide the magnitude of the stress using the damped inversion method (Hardebeck & Michael, 2006) However, the slip tendency can be computed by assuming that the applied to earthquake focal mechanisms in equal‐sized around envelope of Coulomb frictional sliding depending on the friction 1° (longitude) × 20′ (latitude) regions The damped inversion method coefficient of the fault plane is tangential to the Mohr circle (Lisle & produces a regional‐scale model of stress orientations, with the Srivastava, 2004) Finally, the value of Ts varies from to by normal- minimum complexity needed to fit the data The regional stress was izing the values to the friction coefficient (μ) We used a constant inverted using a damping parameter (e = 2.25) on the basis of the friction coefficient throughout all our analyses Therefore, the trade‐off between model length and data variance (Hardebeck & friction coefficient value does not affect the results of our slip Michael, 2006) We used publicly available focal mechanism data from tendency analysis the National Research Institute for Earth Science and Disaster Prevention (NIED) of Japan The data correspond to 922 earthquakes that occurred between January 1997 and 10 March 2011 All of the RESULTS | foci used in the analysis were located at depths of 0–30 km (Figure 1) We excluded areas where the number of focal mechanisms was < 10, 3.1 | Stress fields in Japan in order to ensure optimal inversion stability We obtained the stress The late Quaternary to present‐day stress field is well defined through- field in the 34 regions and calculated the stress regime using the out Japan Seno (1999) derived the maximum horizontal stress trajec- methods of Simpson (1997) (Figure 3) The change in stress state tories, as representing long‐term regional stress fields of Japanese resulting from a large event such as the Tohoku earthquake in 2011 islands during the late Quaternary, on the basis of the distribution of can be a major issue in terms of the stress state in the study area active faults (Huzita, 1980), focal mechanisms of earthquakes (Ukawa, However, the stress field did not change in central Tohoku and 1982) and in situ stress measurements (Tsukahara & Ikeda, 1991) The Kanto–Chubu after the Tohoku earthquake, whereas it did change in regional stress field is also constrained from earthquake focal mecha- northern Tohoku and southern Tohoku near Iwaki City (Yoshida, nisms (e.g Townend & Zoback, 2006; Terakawa & Matsu’ura, 2010; 2012) The majority of our study area is in the central Tohoku area Yukutake, Takeda, & Yoshida, 2015) In general, the regional stress field reported in these previous studies is consistent with the results of Seno (1999), and reveals some spatial heterogeneity across Japan Our results are consistent with the regional stress fields defined in these previous studies (Figure 3) Overall, the direction of maximum compressive stress (σ1) is E–W, and the stress regime gradually changes from reverse to strike‐slip faulting from northeast to southwest Japan The σ1 direction in northern Japan is W–E to WNW–ESE with a reverse fault regime The σ1 direction in central Japan is WNW–ESE with a mixed reverse to strike‐slip fault regime, apart from the region southern Chubu where the NNW–SSE σ1 direction can be locally affected by the collision of the Izu–Bonin Arc 3.2 | Slip tendencies of active and non‐active faults The slip tendencies of each fault, whether active or non‐active, are calculated based on the fault databases and the regional stress we estimated above (Figure 4) The slip tendencies of active faults are FIGURE Inverted stress field of the study area Bars indicate the azimuth of the maximum compressional stress (σ1) Bar colors represent the stress regime following Simpson (1997) generally high across Japan, whereas few non‐active faults have high slip tendencies (Figure 5a,e) The average slip tendency by region is consistently high for active faults, but varies across regions for of MIYAKAWA AND OTSUBO FIGURE The slip tendency of (a) active and (b) non‐active faults in the study area The colors of the faults represent the values of slip tendency Only the faults for which the slip tendency was calculated are shown (a) (b) (c) (d) (e) (f) (g) (h) FIGURE Histograms of (a, e) the total and region‐by‐region slip tendency (ST) for (a–d) active and (e–f) non‐active faults in central–northeast Japan Data are plotted for (b, f) the Kyoto and Osaka regions of the Kinki area, (c, g) Takayama region of the Chubu area, and (d, h) Shinjo and Sakata regions of the Tohoku area as shown in Figures and FIGURE Average slip tendencies for (a) active and (b) non‐active faults in central–northeast Japan Color and size of circle indicate the average slip tendency and the number of active and non‐active faults, respectively non‐active faults (Figure 6) The lowest average slip tendencies for Japan Regional patterns of slip tendency are apparent from histograms non‐active faults were observed in northeast Japan, and the average for three representative areas: the Shinjo region in southeastern slip tendencies gradually increase to the southwest towards central Tohoku, the Takayama region in northern Chubu, and the Kyoto region of MIYAKAWA AND OTSUBO in central Kinki (Figure 5) In the Shinjo region, slip tendencies are high even though geologic evidence of recent fault activity is not for all active faults (> 0.7) and low for non‐active faults (< 0.7) The slip recognized tendencies of active faults are generally high in the Takayama and The presence of potentially active faults with little or no record of Kyoto regions, although some are low The slip tendencies of Quaternary activity can be modeled in terms of the temporal evolution non‐active faults are generally low in central and southeastern Japan, from non‐active to active fault via three stages (Figure 7) Deformation although some have high values of the crust in a given region may begin immediately after the onset of a new state of stress (Stage I) However, the reactivation of non‐active faults usually requires several million years (Doke et al., 2012) Over DISCUSSION | this time, non‐active faults will be reactivated in order of their likelihood of failure, as indicated by their slip tendency values 4.1 Comparison of the slip tendencies of active and non‐active faults | (Stage II) In other words, non‐active faults with a high slip tendency might be potentially active faults in a region that undergoes evolution (Stage I or II) The shift from non‐active to active faults in the region The difference between the slip tendencies of active and non‐active progresses over time (Stage III) faults reflects the difference in their activities High slip tendencies The slip tendency on active and non‐active faults in a region are reported for highly active faults (Miyakawa & Otsubo, 2015a; identifies the evolutionary stage in the development of each region Yukutake et al., 2015) The low slip tendency assigned to some active throughout our study area The Shinjo region contains a small number faults may reflect an underestimation of slip tendency due to uncer- of non‐active faults with a high slip tendency that are in Stage III, tainty in the fault orientation or stress field (Miyakawa & Otsubo, whereas the Kyoto region has many such faults that are in Stage II 2015a; Yukutake et al., 2015) Although some active faults have low (Figure 7) The Takayama region contains a number of non‐active faults slip tendencies, most have high slip tendencies (Figure 5) On the other with a high slip tendency, which record evidence of a stage hand, non‐active faults overall have low slip tendencies reflecting that intermediate between Stage II and Stage III (Figure 7) The they are presently inactive (Figure 5) These relationships are typically abundance of potentially active faults in central Japan (i.e the observed in the Shinjo region (Figure 5d,h) Chubu–Kinki area) reflects this region’s early stage of evolution from non‐active to active fault 4.2 | Temporal evolution from non‐active to active faults High slip tendency was observed for some non‐active faults (Figure 5), 4.3 | Spatial heterogeneity in the evolution from non‐active to active faults although evidence of Quaternary activity is not recognized along these The stage of evolution from non‐active to active fault inferred from the faults As mentioned above, a potentially active fault can be mistakenly slip tendency analysis also changes from northeast to central Japan, classified as a non‐active fault (Figure 2) Therefore, non‐active faults such as from Stage III in the Shinjo region to Stage II–III in the with high slip tendencies may be considered potential active faults, Takayama region and to Stage II in the Kyoto region This spatial FIGURE Schematic diagram of the temporal evolution from a non‐active to an active fault (a) Proportions of geological non‐active and active faults during each stage Bar lengths represent the total number for each region (b) The number of geological non‐active and active faults relative to slip tendency and, therefore, the likelihood of fault slip Trend lines with lighter colors represent faults of the previous stage Gray arrows indicate the increase or decrease in the number of faults of MIYAKAWA AND OTSUBO variation in the locations of potentially active faults and difference in is generally higher along active faults than non‐active faults, although the evolution stages results from the variable crustal deformation in a high slip tendency is observed along some non‐active faults, northeast to central Japan indicating their potential to become active The potential for fault The onset of fault reactivation that is favorable to the present activity along non‐active faults can be modeled using the temporal stress state varies from Tohoku (before 1.5 Ma) to Kinki (after evolution from non‐active to active fault during long‐term crustal 1.0 Ma) The stress state in the study area generally shows deformation The density of potentially active faults varies spatially WNW–ESE to Northwest–Southeast maximum compression directions across the study area and reflects the temporal evolution of crustal and a reverse faulting stress regime (Figure 3) The fault activity favorable deformation in northeast–central Japan to these stress states is Northeast–Southwest striking reverse faulting Our model of the evolution from non‐active to active faults indi- The timing of onset of Northeast–Southwest striking reverse faulting cates that potentially hazardous faults are not only those recognized varies from Tohoku to Kinki: before 1.5 Ma in the Tohoku area, ca as active faults, but also those defined as potentially active faults 1.0 Ma in the Chubu area, and after 1.0 Ma in the Kinki area (Doke et al., The physical methodology employed here establishes the degree of 2012) The origin of the Northeast–Southwest striking reverse faulting activity over the long‐term but cannot precisely constrain the timing since 1.5 Ma is thought to reflect a change in the subduction direction of fault activation The slip tendency analysis can reveal the pres- of the Philippine Sea Plate at 2.0 Ma, from NNW to WNW (Kamata & ence of non‐active faults with high slip tendency that might become Kodama, 1999; Doke et al., 2012) active faults Our results show the wide applicability of this method The evolution of fault reactivation favorable to the present stress state may explain geological and geodetic information on crustal for evaluating potentially active faults without requiring specific surveys (e.g trenching) to obtain chronological data deformation in northeast to central Japan There is a clear More detailed spatial and temporal considerations of these difference between the types of deformation in the EMJS and NKTZ faulting patterns requires further work, as our study has primarily deformation zones in Japan The structures produced by deformation, focused on broad regional trends Consideration of the effects of a such as reverse faulting and folding, are concentrated in the EMJS large seismic event are also necessary for more localized and short‐ (Okamura, 2002) The EMJS is located in the Tohoku area where term studies, although the effect of the 2011 Tohoku Earthquake is the onset of fault reactivation favorable to the present stress state thought to be minimal in our study area However, a previous study has existed since 1.5 Ma or earlier (Doke et al., 2012), which highlighted the importance of the effect of the 2011 Tohoku Earth- suggests that the crust has undergone long‐term deformation In quake on fault activity in the Iwaki area of Tohoku (Miyakawa & contrast, in the NKTZ where crustal deformation is clearly observed Otsubo, 2015b) Therefore, consideration of the influence of large in geodetic signals located in the Chubu–Kinki area, fault reactivation events is an important next step in such studies The seismological favorable to the present stress state has occurred since less than and structural evolution has previously been investigated on the scale 1.5 Ma (Doke et al., 2012) Therefore, crustal deformation under of an individual fault (e.g Wesnousky, 1988), although our study did the present stress state in the NKTZ possibly started relatively not consider individual faults Seismicity rates along major Quaternary recently (< 1.5 Ma) faults have also been investigated and correlated with fault develop- The evolution of crustal deformation from northeast to central ment and/or maturity in Japan (e.g Ishibe & Shimazaki, 2012) There- Japan, as delineated by geological and geodetic data and the onset of fore, further investigation of the correlation between seismicity recent fault activity, shows a relationship with observed spatial (seismicity rate) along these faults and slip tendency is an important variations in the evolution from non‐active to active faults identified subject for future work from slip tendency analysis In northeast Japan where the EMJS is located and recent fault activity began > 1.5 Ma (Doke et al., 2012), it ACKNOWLEDGMENTS is thought that recent deformation has taken place for a longer period This study was supported by MEXT KAKENHI (26109003) We are than in central Japan, where the NKTZ is located and recent fault activ- grateful to Prof Y Kanaori (Yamaguchi University), A Katsube, ity started in the last 1.0 my Consequently, the Tohoku area (i.e Shinjo K Imanishi, and Y Matsuzaki (AIST), and Y Yukutake and R Doke region) is in a later stage of evolution from non‐active to active fault (Hot Springs Research Institute of Kanagawa Prefecture) for helpful (Stage III), whereas the Kinki area (i.e Kyoto region) is in an earlier stage discussions We thank K Michibayashi (Associate Editor of Island Arc), (Stage II) The Chubu area (i.e Takayama region) is in Stage II–III H Tsutsumi (Kyoto University), and an anonymous reviewer who pro- and transitional between the Tohoku and Kinki areas Overall, the vided constructive reviews that greatly improved the manuscript spatial variation in the locations of potentially active faults in north- Some of the figures were drawn using the GMT software package east–central Japan, based on their slip tendency, reflects the temporal (Wessel and Smith, 1998) We also thank AIST for providing fault data evolution of crustal deformation related to the subduction tectonics from the Active Fault Database of Japan, and NIED for providing focal of the Japanese island arc system mechanism solutions from the F‐net Database | S U M M A R Y A N D F U T U R E WO R K We evaluate fault activity in northeast–central Japan based on a slip tendency analysis for both active and non‐active faults Slip tendency RE FE RE NC ES Argus, D F., & Gordon, R G (1990) Pacific–North American Plate motion from very long baseline interferometry compared with motion inferred from magnetic anomalies, transform faults, and earthquake slip vectors Journal of Geophysical Research Solid Earth, 95, 17315–17324 7 of MIYAKAWA AND OTSUBO Demets, C., Gordon, R G., & Arugus, D F (2010) Geologically current plate motions Geophysical Journal International, 181, 1–80 Doke, R., Tanikawa, S., Yasue, K., Nakayasu, A., Niizato, T., Umeda, K., & Tanaka, T (2012) Spatial patterns of initiation ages of active faulting in the Japanese Islands Active Fault Research 37, 1–15 (in Japanese with English abstract) Geological Survey of Japan (1989) Geological maps of Japan 1:200,000, Takayama, NJ‐53‐6 Tsukuba, Japan: National Institute of Advanced science and Technology Hardebeck, J L., & Michael, A J (2006) Damped regional‐scale stress inversions: Methodology and examples for southern California and the Coalinga aftershock sequence Journal of Geophysical Research Solid Earth, 111, B11310 doi:10.1029/2005JB004144 Huzita, K (1980) Role of the Median Tectonic Line in the Quaternary tectonics of the Japanese islands Memories of the Geological Society of Japan, 18, 129–153 (in Japanese with English abstract) Ishibe, T., & Shimazaki, K (2012) Characteristic earthquake model and seismicity around late Quaternary active faults in Japan Bulletin of the Seismological Society of America, 102, 1041–1058 Kamata, H., & Kodama, K (1999) Volcanic history and tectonics of the Southwest Japan Arc Island Arc, 8, 393–403 Kosaka, K., Kanaori, Y., Chigira, M., & Yoshida, S (Eds) (2011) Fault Map of Japan Tokyo, Japan: Baifukan (in Japanese) Lisle, R J., & Srivastava, D C (2004) Test of the frictional reactivation theory for faults and validity of fault‐slip analysis Geology, 32, 569–572 doi:10.1130/G20408.1 Mazzotti, S., Henry, P., & Le Pichon, X (2001) Transient and permanent deformation of central Japan estimated by GPS:2 Strain partitioning and arc–arc collision Earth and Planetary Science Letters, 184, 455–469 doi:10.1016/S0012-821X(00)00336-8 Miyakawa, A., & Otsubo, M (2015a) Applicability of slip tendency for understanding long‐term fault activity: a case study of active faults in northeastern Japan Journal of JSCE, 3, 105–114 doi:10.2208/ journalofjsce.3.1_105 Miyakawa, A., & Otsubo, M (2015b) Effect of a change in the state of stress on the likelihood of inland fault failure during the Mw 6.6 Iwaki earthquake resulting from the Mw 9.0 2011 Tohoku earthquake, Japan Tectonophysics, 661, 112–120 Morris, A., Ferrill, D A., & Henderson, D B (1996) Slip‐tendency analysis and fault reactivation Geology, 24, 275–278 doi:10.1130/00917613(1996)0242.3.CO;2 Ohta, Y., Ohzono, M., Miura, S., Iinuma, T., Tachibana, K., Takatsuka, K., … Umino, N (2008) Coseismic fault model of the 2008 Iwate–Miyagi Nairiku earthquake deduced by a dense GPS network Earth, Planets and Space, 60, 1197–1201 doi:10.1186/BF03352878 Okamura, Y (2002) Deformation zones since the Neogene along the eastern margin of the Japan Sea In M Otake, A Taira, & Y Ota (Eds.), Active Faults and Seismotectonics in the Eastern Margin of the Japan Sea (pp 111–121) Tokyo, Japan: University of Tokyo Press (In Japanese) Sagiya, T., Miyazaki, S., & Tada, T (2000) Continuous GPS array and present‐day crustal deformation of Japan Pure and Applied Geophysics, 157, 2303–2322 doi:10.1007/PL00022507 Sato, H (1994) The relationship between Late Cenozoic tectonic events and stress field and basin development in northeast Japan Journal of Geophysical Research, 99, 22261–22274 doi:10.1029/94JB00854 Seno, T (1999) Syntheses of the regional stress fields of the Japanese islands Island Arc, 8, 66–79 doi:10.1046/j.1440-1738.1999.00225.x Simpson, R W (1997) Quantifying Anderson’s fault types Journal of Geophysical Research Solid Earth, 102, 17909–17919 doi:10.1029/ 97JB01274 Taira, A (2001) Tectonic evolution of the Japanese island arc system Annual Review of Earth and Planetary Sciences, 29, 109–134 doi:10.1146/annurev.earth.29.1.109 The Research Group for Active Faults of Japan (1991) Active Faults in Japan, Sheet Maps and Inventories (Revised ed.) Tokyo, Japan: University of Tokyo Press (in Japanese with English abstract) Terakawa, T., & Matsu’ura, M (2010) The 3D tectonic stress fields in and around Japan inverted from centroid moment tensor data of seismic events Tectonics, 29, TC6008 doi:10.1029/2009TC002626 Townend, J., & Zoback, M D (2006) Stress, strain, and mountain building in central Japan Journal of Geophysical Research, 111, B03411 doi:10.1029/2005JB003759 Tsukahara, H., & Ikeda, R (1991) Crustal stress orientation pattern in the central part of Honshu, Japan—stress provinces and their origins Journal of the Geological Society of Japan, 97, 461–474 (in Japanese) Tsutsumi, H., Sato, K., & Yamaji, A (2012) Stability of the regional stress field in central Japan during the late Quaternary inferred from the stress inversion of the active fault data Geophysical Research Letters, 39, L23303 doi:10.1029/2012GL054094 Ukawa, M (1982) Lateral stretching of the Philippine Sea Plate subducting along the Nankai‐Suruga Trough Tectonics, 1, 543–571 Wesnousky, S G (1988) Seismological and structural evolution of strike‐slip faults Nature, 335, 340–343 Wessel, P., & Smith, W H (1998) New, improved version of Generic Mapping Tools released Eos, Transactions American Geophysical Union, 79, 579–579 Yukutake, Y., Takeda, T., & Yoshida, A (2015) The applicability of frictional reactivation theory to active faults in Japan based on slip tendency analysis Earth and Planetary Science Letters, 411, 188–198 doi:10.1016/j.epsl.2014.12.005 Zoback, M L (1992) First‐and second‐order patterns of stress in the lithosphere: the world stress map project Journal of Geophysical Research, 97, 11703–11728 How to cite this article: Miyakawa A, Otsubo M Evolution of crustal deformation in the northeast–central Japanese island arc: Insights from fault activity Island Arc 2017;e12179 https://doi.org/10.1111/iar.12179 ... LANG=en) The black rectangle outlines the study area for active faults in Japan and determined that the initiation of faulting spans the Pliocene and Pleistocene The range in the ages of fault initiation... deformation in northeast–central Japan to these stress states is Northeast–Southwest striking reverse faulting Our model of the evolution from non‐active to active faults indi- The timing of onset of. .. stress acting on the fault plane, σn is the normal stress We estimated the regional stress state in the study area The area acting on the fault plane, and μ is the friction coefficient of the fault