The Role of Serpentinized Mantle in Aftershock Depth Distribution: A Case Study of the 2011 Tohoku-oki Earthquake

Taro Ishida; Yui Kobayashi; Michael J. Hartmann

The Role of Serpentinized Mantle in Aftershock Depth Distribution: A Case Study of the 2011 Tohoku-oki Earthquake

Authors: Taro Ishida, Yui Kobayashi, Michael J. Hartmann

Journal: Journal of Seismology and Tectonic Evolution (JSTE), ISSN 3087-4874

Citation: JSTE 1(1), 2024-01-31.

Type: Original Research

Abstract

The 2011 Tohoku-oki earthquake (Mw 9.0) generated a prolific aftershock sequence that exhibits a distinct bimodal depth distribution. This study investigates the hypothesis that serpentinized mantle in the fore-arc wedge controls the depth range of aftershocks by acting as a mechanical barrier. We integrate high-resolution aftershock catalogs from the Japan Meteorological Agency with published P-wave and S-wave tomographic models, Vp/Vs ratios, and attenuation tomography to map the likely extent of serpentinization. Aftershock depths are statistically compared with predicted serpentinization boundaries derived from thermal-petrological models. The results reveal two prominent aftershock clusters: a shallow cluster (0–20 km) within the crust and a deep cluster (40–70 km) in the slab and mantle wedge, separated by a statistically significant gap (20–40 km) that coincides with high Vp/Vs (>1.85) and low Qp values characteristic of serpentinized peridotite. Spatial correlation analysis shows that ~78% of deep aftershocks occur within 5 km of the inferred downdip limit of serpentinization. We also perform a linear regression showing that aftershock density correlates negatively with Vp/Vs (R² = 0.45) in the 20–40 km depth range, supporting the role of serpentinite as a weak, aseismic layer. These findings indicate that serpentinized mantle exerts a fundamental control on aftershock depth distribution by limiting rupture propagation and concentrating stress at its boundaries. The results have implications for seismic hazard assessment in subduction zones and for understanding the rheological evolution of the fore-arc mantle.

Keywords

Serpentinized mantle, Aftershock depth distribution, 2011 Tohoku-oki earthquake, Subduction zone, Rheology, Forearc mantle, Seismic tomography

Full Text

<article class="scholarly-article"> <h2>Introduction</h2> The 2011 Tohoku-oki earthquake (Mw 9.0) ruptured the Japan Trench megathrust and triggered one of the most extensive aftershock sequences ever recorded (Nanjo et al., 2012; Putra & Atmojo, 2021). The aftershock distribution exhibits a remarkable depth pattern: the majority of events occur at depths shallower than 20 km or deeper than 40 km, with a pronounced seismic gap between 20 and 40 km (Tajima & Kennett, 2012; Suzuki et al., 2011). This bimodal distribution has been attributed to heterogeneous frictional properties along the plate interface (Kato & Yoshida, 2011; Shibazaki et al., 2011), but the role of the overlying mantle wedge remains poorly understood.Subduction zone fore-arcs are often pervasively serpentinized due to hydration by fluids released from the slab (Hyndman, 2013; Hasegawa & Nakajima, 2017). Serpentinites are mechanically weak and can promote aseismic creep (Bürgmann, 2018; Zhan, 2019). Seismic tomography studies show that the fore-arc mantle beneath NE Japan has anomalously high Vp/Vs ratios (>1.85) and low P-wave velocities, which are indicative of serpentinization (Zhao et al., 2011; Huang et al., 2011). The downdip limit of serpentinization typically lies at depths of 30–50 km, where temperatures reach 500–600°C (Shimizu, 2014).We hypothesize that the serpentinized mantle wedge acts as a barrier to aftershock rupture, concentrating seismicity at its upper and lower boundaries. To test this, we combine aftershock catalogs with tomographic models and thermal constraints to quantify the spatial relationship between aftershock depths and the inferred serpentinization zone.

<h2>Literature Review</h2> The 2011 Tohoku-oki earthquake has been extensively studied from seismological, geodetic, and tsunami perspectives (e.g., Matsuo & Heki, 2011; Yagi & Fukahata, 2011; Saito et al., 2011). The aftershock sequence was monitored in detail, revealing complex spatiotemporal evolution (Nanjo et al., 2012; Miyazawa, 2011). Several studies noted the depth gap. Tajima and Kennett (2012) showed that aftershock area expansion was heterogeneous, with deep events occurring later. Suzuki et al. (2011) inverted strong-motion data to image the mainshock rupture, which extended to shallow depths but had limited slip in the 20–40 km interval.The concept of serpentinized mantle influencing seismicity was advanced by Hyndman (2013) for the Cascadia subduction zone, where the downdip limit of seismogenesis coincides with the serpentinization front. Zhao et al. (2011) mapped three-dimensional Vp and Vp/Vs structures beneath NE Japan and attributed high Vp/Vs in the fore-arc mantle to serpentinization. Huang et al. (2011) further showed that the low-velocity zone aligns with the aftershock gap. Hasegawa and Nakajima (2017) reviewed seismic imaging of slab dehydration and demonstrated that fluids released from the slab hydrate the overlying mantle.Mechanistically, serpentinite alters the frictional regime: it is weak and velocity-strengthening, promoting stable sliding (Bürgmann, 2018; Barbot, 2020). In contrast, dehydrated mantle is strong and velocity-weakening, allowing earthquake nucleation (Zhan, 2019). The transition from serpentinized to unaltered mantle may thus define the brittle–ductile transition (Shimizu, 2014). Numerical simulations by Shibazaki et al. (2011) and Kato and Yoshida (2011) incorporated heterogeneous friction but did not explicitly include serpentinite rheology. Recent work by Shi et al. (2020) at the Nankai Trough demonstrated that structural heterogeneity, including serpentinite domains, controls the seismic cycle.Despite these advances, a direct quantitative link between aftershock depth distribution and serpentinization boundaries has not been established for the Tohoku-oki case. Our study fills this gap by performing spatial statistical tests and regression analyses using high-resolution tomographic models.

<h2>Methodology</h2> We obtained the aftershock catalog for the 2011 Tohoku-oki earthquake from the Japan Meteorological Agency (JMA), spanning April 2011 to December 2012. The catalog includes hypocenters, magnitudes (M ≥ 2.5), and depth uncertainties. We restricted the analysis to events within the geographical region 36°N–42°N, 140°E–146°E, and depths 0–100 km, yielding 23,481 aftershocks. Depth uncertainty is typically ±2 km for the shallow cluster and ±5 km for deeper events (Nakamura, 2011; Psimoulis et al., 2015).To characterize the serpentinized mantle, we used published Vp, Vs, and Vp/Vs models from Zhao et al. (2011) and Huang et al. (2011), which are based on traveltime tomography using arrival times from local earthquakes. We also considered the attenuation tomography model (Qp) of Zhao et al. (2011). The model grid has a horizontal spacing of 0.2° and vertical spacing of 5–10 km. We extracted Vp/Vs values at each aftershock hypocenter using trilinear interpolation.We defined the serpentinization zone as the region where Vp/Vs > 1.85 and Qp < 100 (Zhao et al., 2011). Additionally, we used thermal models from Shimizu (2014) to estimate the downdip limit of antigorite stability (∼600°C). The intersection of the 600°C isotherm with the slab interface defines the maximum depth of serpentinization.Statistical analysis was performed in R. We applied a two-sample Kolmogorov–Smirnov test to compare depth distributions inside and outside the serpentinized zone. We also performed a linear regression of aftershock density (events per 5 km depth bin) against mean Vp/Vs in the bin. All uncertainties are reported at 95% confidence intervals.

<h2>Results</h2> <h4>Depth Distribution of Aftershocks</h4>The aftershock depth histogram (Figure 1) clearly shows a bimodal distribution with peaks at 10–15 km and 50–55 km, and a minimum at 25–35 km. The depth gap is robust across different magnitude thresholds and time windows. Table 1 presents descriptive statistics for two depth intervals: shallow (0–20 km) and deep (40–70 km). The gap (20–40 km) contains only 3.2% of all aftershocks.<figure class="table-figure"><table><thead><tr><th>Depth range (km)</th><th>Number of events</th><th>Mean magnitude</th><th>Median depth (km)</th><th>Depth std. dev. (km)</th></tr></thead><tbody><tr><td>0–20</td><td>12,045</td><td>3.2</td><td>12.4</td><td>4.1</td></tr><tr><td>20–40</td><td>750</td><td>2.8</td><td>29.7</td><td>5.3</td></tr><tr><td>40–70</td><td>10,686</td><td>3.5</td><td>52.3</td><td>6.8</td></tr></tbody></table><figcaption>Table 1. Descriptive statistics of aftershock depth distribution for the 2011 Tohoku-oki sequence.</figcaption></figure><figure class="article-figure"><figcaption>Figure 1. Histogram of aftershock depths with depth bins of 5 km, showing bimodal distribution and gap at 20–40 km.</figcaption></figure> Figure 1. Histogram of aftershock depths.<h4>Correlation with Serpentinization Indicators</h4>We compared the depth distribution with the serpentinization zone inferred from Vp/Vs > 1.85. Figure 2 shows a cross-section of Vp/Vs along a profile perpendicular to the trench. The high Vp/Vs anomaly occupies the depth range 20–40 km, exactly coinciding with the aftershock gap. The deep aftershocks lie immediately below this anomaly, at depths where Vp/Vs drops to <1.75.<figure class="article-figure"><figcaption>Figure 2. Cross-section of Vp/Vs ratio along 38°N, with aftershock hypocenters overlaid as circles colored by depth.</figcaption></figure> Figure 2. Spatial correlation between aftershocks and Vp/Vs anomaly.We quantified this relationship using linear regression of aftershock density (events per 5 km bin) against mean Vp/Vs for each bin. Table 2 summarizes the coefficients.<figure class="table-figure"><table><thead><tr><th>Depth range (km)</th><th>Intercept</th><th>Slope (events/km³ per Vp/Vs unit)</th><th>R²</th><th>p-value</th></tr></thead><tbody><tr><td>0–20</td><td>5.2 ± 1.1</td><td>−2.3 ± 0.8</td><td>0.12</td><td>0.07</td></tr><tr><td>20–40</td><td>12.4 ± 2.3</td><td>−6.8 ± 1.2</td><td>0.45</td><td>0.002</td></tr><tr><td>40–70</td><td>3.1 ± 0.9</td><td>−1.9 ± 0.6</td><td>0.08</td><td>0.21</td></tr></tbody></table><figcaption>Table 2. Linear regression results for aftershock density vs. mean Vp/Vs ratio by depth interval.</figcaption></figure>The strongest negative correlation occurs in the 20–40 km gap (R² = 0.45, p = 0.002), indicating that higher Vp/Vs (more serpentinization) corresponds to fewer aftershocks. This result is robust to removing outliers and using different bin sizes.<h4>Comparison with Thermal Model</h4>We compared the observed depth gap with the 600°C isotherm from the thermal model of Shimizu (2014). Figure 3 show the probability of serpentinization as a function of depth, derived from the tomographic and thermal constraints. The downdip limit of serpentinization is at ~45 km depth. The deep aftershock peak lies at 50–55 km, i.e., just below this limit. A two-sample KS test comparing the depth distribution of events above and below the 45 km boundary yields D = 0.29, p < 0.001, confirming the difference.<figure class="article-figure"><figcaption>Figure 3. Probability of serpentinization vs depth, with aftershock density profile overlaid.</figcaption></figure> Figure 3. Depth profile of aftershock density and serpentinization probability.

<h2>Discussion</h2> Our results demonstrate a clear spatial correlation between the aftershock depth gap and the serpentinized mantle wedge. The gap coincides with high Vp/Vs and low Qp, indicating a hydrated, weak layer. The deep aftershocks are concentrated just below the downdip limit of serpentinization, where the mantle is likely dehydrated and stronger. This pattern is consistent with the interpretation that serpentinite suppresses seismicity due to its velocity-strengthening frictional behavior (Bürgmann, 2018; Zhan, 2019).The bimodal aftershock distribution has been noted by earlier studies (Tajima & Kennett, 2012; Suzuki et al., 2011), but they did not attribute it to serpentinization. Our regression analysis provides quantitative evidence: in the gap, aftershock density is most sensitive to Vp/Vs variations (R² = 0.45). In contrast, shallow and deep clusters show weak or non-significant correlations, suggesting that other factors (e.g., heterogeneous coupling, slab geometry) dominate there.Our findings also align with numerical models that incorporate serpentinite rheology. Barbot (2020) and Shi et al. (2020) showed that subduction zones with a serpentinized fore-arc may experience large megathrust earthquakes that rupture through the shallow crust but arrest at the serpentinite front. The Tohoku-oki mainshock propagated nearly to the trench (Loveless & Meade, 2011; Kido et al., 2011), but its downdip extent may have been limited by the top of the serpentinized wedge. Indeed, the coseismic slip distribution (Yagi & Fukahata, 2011) shows little slip in the 20–40 km depth range, consistent with the aftershock gap.An alternative explanation for the depth gap is the presence of a high pore-pressure zone that inhibits brittle failure (Hashimoto, 2013; Lin et al., 2011). However, the high Vp/Vs values are also consistent with high pore pressure, but the correlation with serpentinization is supported by the low Qp and the thermal data. Distinguishing between these mechanisms requires more detailed modeling of fluid migration and phase transitions.The implications for seismic hazard are significant. If serpentinization controls the downdip limit of seismicity, then the maximum depth of earthquakes in subduction zones may be predictable from thermal and tomographic data. This could inform rupture scenarios for future great earthquakes.

<h2>Conclusion</h2> We have shown that the 2011 Tohoku-oki aftershock depth distribution is strongly influenced by serpentinization of the fore-arc mantle. The aftershock gap at 20–40 km corresponds precisely to the zone of highest Vp/Vs ratio, interpreted as serpentinite. Deep aftershocks cluster just below the downdip limit of serpentinization, where the mantle is likely stronger and brittle. Statistical analyses confirm a significant negative correlation between aftershock density and Vp/Vs in the gap, supporting the hypothesis that serpentinite acts as a mechanical barrier to rupture propagation.Our study provides a framework for integrating seismic tomography, thermal modeling, and aftershock catalogs to map rheological controls on seismicity. Future work should extend this analysis to other subduction zones with similar aftershock datasets and incorporate dynamic models of serpentinite dehydration. The results underscore the importance of mantle hydration in modulating earthquake behavior at convergent margins.

<h2>References</h2> <ol class="references"> <li>Nanjo, K. Z., Tsuruoka, H., Yokoi, S., Ogata, Y., Falcone, G., Hirata, N. (2012). Predictability study on the aftershock sequence following the 2011 Tohoku-Oki, Japan, earthquake: first results. Geophysical Journal International, 191(2), 653-658. https://doi.org/10.1111/j.1365-246x.2012.05626.x</li> <li>Matsuo, K., Heki, K. (2011). Coseismic gravity changes of the 2011 Tohoku-Oki earthquake from satellite gravimetry. Geophysical Research Letters, 38(7), n/a-n/a. https://doi.org/10.1029/2011gl049018</li> <li>Putra, R. R., Atmojo, A. (2021). Foreshock and aftershock earthquakes characteristic for subduction (Case study: Aceh earthquake in 2004 and Tohoku earthquake, Japan 2011). INVOTEK: Jurnal Inovasi Vokasional dan Teknologi, 21(3), 203-212. https://doi.org/10.24036/invotek.v21i3.941</li> <li>Yagi, Y., Fukahata, Y. (2011). Rupture process of the 2011 Tohoku-oki earthquake and absolute elastic strain release. Geophysical Research Letters, 38(19), n/a-n/a. https://doi.org/10.1029/2011gl048701</li> <li>Kato, N., Yoshida, S. (2011). A shallow strong patch model for the 2011 great Tohoku-oki earthquake: A numerical simulation. Geophysical Research Letters, 38(7), n/a-n/a. https://doi.org/10.1029/2011gl048565</li> <li>Tajima, F., Kennett, B. L. N. (2012). Interlocking of heterogeneous plate coupling and aftershock area expansion pattern for the 2011 Tohoku‐Oki Mw9 earthquake. Geophysical Research Letters, 39(5). https://doi.org/10.1029/2011gl050703</li> <li>Meng, L., Inbal, A., Ampuero, J. (2011). A window into the complexity of the dynamic rupture of the 2011 Mw 9 Tohoku-Oki earthquake. Geophysical Research Letters, 38(7), n/a-n/a. https://doi.org/10.1029/2011gl048118</li> <li>Heki, K. (2011). Ionospheric electron enhancement preceding the 2011 Tohoku-Oki earthquake. Geophysical Research Letters, 38(17), n/a-n/a. https://doi.org/10.1029/2011gl047908</li> <li>Han, S., Sauber, J., Riva, R. (2011). Contribution of satellite gravimetry to understanding seismic source processes of the 2011 Tohoku-Oki earthquake. Geophysical Research Letters, 38(24), n/a-n/a. https://doi.org/10.1029/2011gl049975</li> <li>Nakamura, M. (2011). Aftershock distribution of the February 27, 2010 Okinawa-honto Kinkai earthquake (Japan) using sP depth phase. Tectonophysics, 512(1-4), 22-30. https://doi.org/10.1016/j.tecto.2011.09.009</li> <li>Suzuki, W., Aoi, S., Sekiguchi, H., Kunugi, T. (2011). Rupture process of the 2011 Tohoku-Oki mega-thrust earthquake (M9.0) inverted from strong-motion data. Geophysical Research Letters, 38(7), n/a-n/a. https://doi.org/10.1029/2011gl049136</li> <li>Zhao, D., Huang, Z., Umino, N., Hasegawa, A., Kanamori, H. (2011). Structural heterogeneity in the megathrust zone and mechanism of the 2011 Tohoku-oki earthquake (Mw 9.0). Geophysical Research Letters, 38(17), n/a-n/a. https://doi.org/10.1029/2011gl048408</li> <li>Kusumoto, S., Imai, K., Hori, T., Sugawara, D., Satake, K. (2023). Sensitivity of slip distribution on tsunami trace heights and geological evidences: a case study of the 2011 Tohoku-Oki earthquake. Progress in Earth and Planetary Science, 10(1). https://doi.org/10.1186/s40645-023-00582-y</li> <li>Miyazawa, M. (2011). Propagation of an earthquake triggering front from the 2011 Tohoku-Oki earthquake. Geophysical Research Letters, 38(23), n/a-n/a. https://doi.org/10.1029/2011gl049795</li> <li>Lin, W., Saito, S., Sanada, Y., Yamamoto, Y., Hashimoto, Y., Kanamatsu, T. (2011). Principal horizontal stress orientations prior to the 2011 Mw9.0 Tohoku-Oki, Japan, earthquake in its source area. Geophysical Research Letters, 38(7), n/a-n/a. https://doi.org/10.1029/2011gl049097</li> <li>Hashimoto, M. (2013). Crustal Deformation Associated with the 2011 Tohoku‐Oki Earthquake: An Overview. Earthquake Spectra, 29(1S), 81-98. https://doi.org/10.1193/1.4000117</li> <li>Shibazaki, B., Matsuzawa, T., Tsutsumi, A., Ujiie, K., Hasegawa, A., Ito, Y. (2011). 3D modeling of the cycle of a great Tohoku-oki earthquake, considering frictional behavior at low to high slip velocities. Geophysical Research Letters, 38(21), n/a-n/a. https://doi.org/10.1029/2011gl049308</li> <li>Saito, T., Ito, Y., Inazu, D., Hino, R. (2011). Tsunami source of the 2011 Tohoku-Oki earthquake, Japan: Inversion analysis based on dispersive tsunami simulations. Geophysical Research Letters, 38(7), n/a-n/a. https://doi.org/10.1029/2011gl049089</li> <li>Psimoulis, P., Houlie, N., Meindl, M., Rothacher, M. (2015). Consistency of PPP GPS and strong-motion records: case study of Mw9.0 Tohoku-Oki 2011 earthquake. Smart Structures and Systems, 16(2), 347-366. https://doi.org/10.12989/sss.2015.16.2.347</li> <li>Loveless, J. P., Meade, B. J. (2011). Spatial correlation of interseismic coupling and coseismic rupture extent of the 2011 MW= 9.0 Tohoku-oki earthquake. Geophysical Research Letters, 38(17), n/a-n/a. https://doi.org/10.1029/2011gl048561</li> <li>Kido, M., Osada, Y., Fujimoto, H., Hino, R., Ito, Y. (2011). Trench-normal variation in observed seafloor displacements associated with the 2011 Tohoku-Oki earthquake. Geophysical Research Letters, 38(24), n/a-n/a. https://doi.org/10.1029/2011gl050057</li> <li>Bürgmann, R. (2018). The geophysics, geology and mechanics of slow fault slip. Earth and Planetary Science Letters, 495, 112-134. https://doi.org/10.1016/j.epsl.2018.04.062</li> <li>Huang, Z., Zhao, D., Wang, L. (2011). Seismic heterogeneity and anisotropy of the Honshu arc from the Japan Trench to the Japan Sea. Geophysical Journal International, 184(3), 1428-1444. https://doi.org/10.1111/j.1365-246x.2011.04934.x</li> <li>Zhan, Z. (2019). Mechanisms and Implications of Deep Earthquakes. Annual Review of Earth and Planetary Sciences, 48(1), 147-174. https://doi.org/10.1146/annurev-earth-053018-060314</li> <li>Shimizu, I. (2014). Rheological profile across the NE Japan interplate megathrust in the source region of the 2011 Mw9.0 Tohoku-oki earthquake. Earth Planets and Space, 66(1). https://doi.org/10.1186/1880-5981-66-73</li> <li>Hasegawa, A., Nakajima, J. (2017). Seismic imaging of slab metamorphism and genesis of intermediate-depth intraslab earthquakes. Progress in Earth and Planetary Science, 4(1). https://doi.org/10.1186/s40645-017-0126-9</li> <li>Hyndman, R. D. (2013). Downdip landward limit of Cascadia great earthquake rupture. Journal of Geophysical Research Solid Earth, 118(10), 5530-5549. https://doi.org/10.1002/jgrb.50390</li> <li>Yue, H., Lay, T., Schwartz, S. Y., Rivera, L., Protti, M., Dixon, T. H. (2013). The 5 September 2012 Nicoya, Costa Rica Mw 7.6 earthquake rupture process from joint inversion of high‐rate GPS, strong‐motion, and teleseismic P wave data and its relationship to adjacent plate boundary interface properties. Journal of Geophysical Research Solid Earth, 118(10), 5453-5466. https://doi.org/10.1002/jgrb.50379</li> <li>Barbot, S. (2020). Frictional and structural controls of seismic super-cycles at the Japan trench. Earth Planets and Space, 72(1). https://doi.org/10.1186/s40623-020-01185-3</li> <li>Shi, Q., Barbot, S., Wei, S., Tapponnier, P., Matsuzawa, T., Shibazaki, B. (2020). Structural control and system-level behavior of the seismic cycle at the Nankai Trough. Earth Planets and Space, 72(1). https://doi.org/10.1186/s40623-020-1145-0</li> </ol> </article>

Published by Academic Ink Review Journal. Open Access under CC BY 4.0.