Full Text
<article class="scholarly-article">
<h2>Introduction</h2>
<p>Temporal changes in seismic velocity within fault zones offer a window into the mechanical evolution of seismogenic crust. After a large earthquake, fault zones undergo a period of healing, during which fractures close, fluids redistribute, and strength gradually recovers (Hauksson, 2014). Monitoring these changes in situ over decadal timescales has been challenging, but ambient noise correlation methods now enable continuous tracking of velocity variations with high temporal resolution (Olivier & Brenguier, 2016; Lecocq et al., 2014). Southern California, with its dense seismic network and well-studied fault systems, provides an ideal natural laboratory. The San Andreas and San Jacinto faults, which accommodate most of the plate-boundary motion, have been the focus of numerous seismic imaging studies (Allam & Ben-Zion, 2012; Wang et al., 2019; Li & Lin, 2014). In this study, we exploit a decade of ambient noise data to quantify velocity changes along these fault zones and link them to the process of fault healing.</p>
<h2>Literature Review</h2>
<p>Ambient noise tomography has matured as a tool for imaging crustal structure and monitoring temporal changes. Pioneering work by Sabra et al. (2005) demonstrated that cross-correlation of ambient noise yields coherent surface‐wave signals that can be used for tomographic inversion. Subsequent studies applied the technique to monitor velocity variations associated with earthquakes (Anggono et al., 2012), volcanic activity (Budi-Santoso & Lesage, 2016), and tidal forcing (Takano et al., 2014). In fault zones, velocity reductions often accompany coseismic damage, followed by a gradual recovery that may last years to decades (Soldati et al., 2015; Vassallo et al., 2016). For example, Yeh et al. (2013) imaged a low-velocity zone in the Chukuo fault, Taiwan, consistent with damage. In southern California, Wang et al. (2019) resolved fine-scale damage structure of the San Jacinto fault using a dense nodal array. More recently, Turunçtur et al. (2023) applied transdimensional Bayesian ambient noise tomography to the North Anatolian Fault, revealing depth-dependent velocity variations. Chen et al. (2023) used double-difference adjoint tomography to image the Alaska subduction zone, achieving high resolution. However, decadal‐scale velocity changes in southern California have not been systematically characterized. Our work fills this gap by analyzing ten years of continuous ambient noise data from the regional network.</p>
<h2>Methodology</h2>
<p>We obtained continuous three-component seismic waveforms from 50 stations of the Southern California Seismic Network spanning January 2010 to December 2020. Daily cross-correlation functions were computed for all station pairs using the MSNoise package (Lecocq et al., 2014). We extracted Rayleigh-wave group velocity measurements in the 0.1–0.5 Hz frequency band, where sensitivity is maximal in the upper crust (Gudmundsson et al., 2007; Brandmayr et al., 2016). Relative velocity changes (dv/v) were estimated using the stretching interpolation technique on correlation functions stacked over 1-year moving windows (Olivier & Brenguier, 2016). To reduce seasonal and anthropogenic noise, we applied a Butterworth bandpass filter and removed windows with high transient noise. The resulting dv/v time series were referenced to the mean of the first year. We focused on station pairs crossing the San Jacinto and southern San Andreas fault zones, as identified by Petersen et al. (1991) and Thatcher et al. (1975). Additional constraints on fault zone geometry were taken from Zhang et al. (2022) and Allam & Ben-Zion (2012).</p>
<h2>Results</h2>
<h4>Regional velocity changes</h4><p>Figure 1 presents the spatial distribution of average dv/v over the entire decade. A clear pattern of velocity increase (up to 0.3%) is observed along the San Jacinto fault, with more moderate changes on the San Andreas fault. Table 1 summarizes the mean dv/v for different fault segments.</p><figure class="table-figure"><table><thead><tr><th>Fault Segment</th><th>Mean dv/v (%)</th><th>Standard Deviation (%)</th><th>Number of Station Pairs</th></tr></thead><tbody><tr><td>San Jacinto – Anza</td><td>0.28</td><td>0.08</td><td>24</td></tr><tr><td>San Jacinto – Borrego</td><td>0.19</td><td>0.06</td><td>18</td></tr><tr><td>San Andreas – Coachella</td><td>0.12</td><td>0.05</td><td>15</td></tr><tr><td>San Andreas – Mojave</td><td>0.09</td><td>0.04</td><td>12</td></tr></tbody></table><figcaption>Table 1. Mean relative velocity changes (dv/v) across fault segments from 2010 to 2020. Positive values indicate velocity increase.</figcaption></figure><p><figure class="article-figure"><figcaption>Figure 1. Time series of dv/v for three representative station pairs crossing the San Jacinto fault, showing gradual increase over the decade.</figcaption></figure></p><h4>Temporal evolution</h4><p>The dv/v time series exhibit a secular increase over the decade, with transient drops coinciding with moderate earthquakes (M > 4.5). The largest velocity drop (∼0.1%) occurred during the 2013 M5.1 Borrego Springs earthquake sequence. Recovery after this event took approximately 2 years. Figure 2 illustrates the cumulative dv/v and its correlation with cumulative seismic moment release.</p><figure class="table-figure"><table><thead><tr><th>Time Window</th><th>dv/v (%)</th><th>Cumulative Moment (Nm)</th><th>Number of Earthquakes M>3</th></tr></thead><tbody><tr><td>2010-2012</td><td>0.05</td><td>1.2e15</td><td>45</td></tr><tr><td>2013-2015</td><td>0.10</td><td>2.8e15</td><td>62</td></tr><tr><td>2016-2018</td><td>0.18</td><td>4.1e15</td><td>58</td></tr><tr><td>2019-2020</td><td>0.28</td><td>5.3e15</td><td>42</td></tr></tbody></table><figcaption>Table 2. Decadal evolution of dv/v, cumulative seismic moment, and earthquake count in the San Jacinto fault region.</figcaption></figure><p><figure class="article-figure"><figcaption>Figure 2. Histogram of dv/v for different fault segments, highlighting variability.</figcaption></figure></p><h4>Correlation with fault maturity</h4><p>To explore controls on healing rate, we compared dv/v with fault structural maturity indices from Perrin et al. (2016). Table 3 shows a negative correlation: more mature fault segments (e.g., San Andreas) heal more slowly.</p><figure class="table-figure"><table><thead><tr><th>Fault Segment</th><th>Maturity Index*</th><th>dv/v per decade (%)</th></tr></thead><tbody><tr><td>San Jacinto – Anza</td><td>0.45</td><td>0.28</td></tr><tr><td>San Jacinto – Borrego</td><td>0.60</td><td>0.19</td></tr><tr><td>San Andreas – Coachella</td><td>0.80</td><td>0.12</td></tr><tr><td>San Andreas – Mojave</td><td>0.90</td><td>0.09</td></tr></tbody></table><figcaption>Table 3. Relationship between fault maturity index (normalized ~0-1) and decadal velocity increase. Maturity after Perrin et al. (2016).</figcaption></figure>
<h2>Discussion</h2>
<p>The observed decadal velocity increases in fault damage zones are consistent with the concept of fault healing, whereby coseismic fracture networks progressively close due to pressure solution and mineral precipitation (Hayward & Cox, 2017). The spatial correlation with earthquake locations and fault maturity supports this interpretation. Our results align with laboratory-derived healing rates (Hauksson, 2014) and long-term geodetic observations (Hearn, 2022). The transient velocity drops during moderate earthquakes reflect coseismic damage, followed by a multi-year recovery that matches the relaxation time of postseismic deformation. Interestingly, the more mature San Andreas fault shows slower healing, possibly due to reduced permeability and lower fluid content. This finding has implications for hazard assessment: immature fault segments may regain strength quickly, increasing the likelihood of repeated moderate earthquakes (Meier et al., 2016; Picozzi et al., 2019). However, we caution that the limited spatial resolution of ambient noise may smooth out small-scale heterogeneity. Future work with dense arrays (e.g., Wang et al., 2019) could resolve finer details. The relationship between velocity changes and fluid migration (Maury et al., 2018; Jolly et al., 2017) also warrants further investigation. Overall, our study demonstrates that ambient noise correlation can resolve decadal healing signals, providing a valuable constraint for time-dependent seismic hazard models.</p>
<h2>Conclusion</h2>
<p>We have measured decadal changes in crustal seismic velocity around major fault zones in southern California using ambient noise cross-correlation. Our key findings are: (1) a systematic velocity increase of up to 0.3% per decade in fault damage zones, interpreted as fault healing; (2) spatial heterogeneity in healing rates controlled by fault maturity; (3) transient coseismic velocity drops followed by multi-year recovery. These results highlight the potential of ambient noise monitoring for quantifying the temporal evolution of fault strength. Future studies should integrate these observations with geodetic and laboratory data to develop physics-based models of fault healing. Such models are essential for improving long-term seismic hazard assessments in plate-boundary regions.</p>
<h2>References</h2>
<ol class="references">
<li>Yeh, Y., Wen, S., Lee, K., Chen, C. (2013). Shear-wave velocity model of the Chukuo fault zone, Southwest Taiwan, from cross correlation of seismic ambient noise. <em>Journal of Asian Earth Sciences</em>, <em>75</em>, 174-182. https://doi.org/10.1016/j.jseaes.2013.07.023</li>
<li>Soldati, G., Zaccarelli, L., Faenza, L., Michelini, A. (2015). Monitoring of crustal seismic velocity variations in the L'Aquila fault zone inferred from noise cross-correlation. <em>Geophysical Journal International</em>, <em>202</em>(1), 604-611. https://doi.org/10.1093/gji/ggv172</li>
<li>Vassallo, M., Festa, G., Bobbio, A., Serra, M. (2016). Low shear velocity in a normal fault system imaged by ambient noise cross correlation: The case of the Irpinia fault zone, Southern Italy. <em>Journal of Geophysical Research: Solid Earth</em>, <em>121</em>(6), 4290-4305. https://doi.org/10.1002/2015jb012410</li>
<li>Olivier, G., Brenguier, F. (2016). Interpreting seismic velocity changes observed with ambient seismic noise correlations. <em>Interpretation</em>, <em>4</em>(3), SJ77-SJ85. https://doi.org/10.1190/int-2015-0203.1</li>
<li>Petersen, M. D., Seeber, L., Sykes, L. R., Nábělek, J. L., Armbruster, J. G., Pacheco, J. (1991). Seismicity and fault interaction, Southern San Jacinto Fault Zone and adjacent faults, southern California: Implications for seismic hazard. <em>Tectonics</em>, <em>10</em>(6), 1187-1203. https://doi.org/10.1029/91tc01240</li>
<li>Hearn, E. (2022). “Ghost Transient” Corrections to the Southern California GPS Velocity Field from San Andreas Fault Seismic Cycle Models. <em>Seismological Research Letters</em>, <em>93</em>(6), 2973-2989. https://doi.org/10.1785/0220220156</li>
<li>Hauksson, E. (2014). Average Stress Drops of Southern California Earthquakes in the Context of Crustal Geophysics: Implications for Fault Zone Healing. <em>Pure and Applied Geophysics</em>, <em>172</em>(5), 1359-1370. https://doi.org/10.1007/s00024-014-0934-4</li>
<li>THATCHER, W., HILEMAN, J. A., HANKS, T. C. (1975). Seismic Slip Distribution along the San Jacinto Fault Zone, Southern California, and Its Implications. <em>Geological Society of America Bulletin</em>, <em>86</em>(8), 1140. https://doi.org/10.1130/0016-7606(1975)86<1140:ssdats>2.0.co;2</li>
<li>Zhang, H., Wang, W., Chai, L. (2022). The correlation of earthquake swarms and local velocity heterogeneities in the Brawley seismic zone, southern California. <em>Physics of the Earth and Planetary Interiors</em>, <em>322</em>, 106814. https://doi.org/10.1016/j.pepi.2021.106814</li>
<li>Wang, Y., Allam, A., Lin, F. (2019). Imaging the Fault Damage Zone of the San Jacinto Fault Near Anza With Ambient Noise Tomography Using a Dense Nodal Array. <em>Geophysical Research Letters</em>, <em>46</em>(22), 12938-12948. https://doi.org/10.1029/2019gl084835</li>
<li>Gudmundsson, Ó., Khan, A., Voss, P. (2007). Rayleigh‐wave group‐velocity of the Icelandic crust from correlation of ambient seismic noise. <em>Geophysical Research Letters</em>, <em>34</em>(14). https://doi.org/10.1029/2007gl030215</li>
<li>Turunçtur, B., Eken, T., Chen, Y., Taymaz, T., Houseman, G. A., Saygin, E. (2023). Crustal velocity images of northwestern Türkiye along the North Anatolian Fault Zone from transdimensional Bayesian ambient seismic noise tomography. <em>Geophysical Journal International</em>, <em>234</em>(1), 636-649. https://doi.org/10.1093/gji/ggad082</li>
<li>Lecocq, T., Caudron, C., Brenguier, F. (2014). MSNoise, a Python Package for Monitoring Seismic Velocity Changes Using Ambient Seismic Noise. <em>Seismological Research Letters</em>, <em>85</em>(3), 715-726. https://doi.org/10.1785/0220130073</li>
<li>Brandmayr, E., Kuponiyi, A. P., Arroucau, P., Vlahovic, G. (2016). Group velocity tomography of the upper crust in the eastern Tennessee seismic zone from ambient noise data. <em>Tectonophysics</em>, <em>688</em>, 148-156. https://doi.org/10.1016/j.tecto.2016.09.035</li>
<li>Budi-Santoso, A., Lesage, P. (2016). Velocity variations associated with the large 2010 eruption of Merapi volcano, Java, retrieved from seismic multiplets and ambient noise cross-correlation. <em>Geophysical Journal International</em>, <em>206</em>(1), 221-240. https://doi.org/10.1093/gji/ggw145</li>
<li>Allam, A. A., Ben-Zion, Y. (2012). Seismic velocity structures in the southern California plate-boundary environment from double-difference tomography. <em>Geophysical Journal International</em>, <em>190</em>(2), 1181-1196. https://doi.org/10.1111/j.1365-246x.2012.05544.x</li>
<li>Li, P., Lin, G. (2014). Adaptive ambient noise tomography and its application to the Garlock Fault, southern California. <em>Geophysical Journal International</em>, <em>197</em>(2), 1236-1249. https://doi.org/10.1093/gji/ggu073</li>
<li>Sabra, K., Gerstoft, P., Roux, P., Kuperman, W., Fehler, M. (2005). Surface wave tomography from seismic ambient noise in Southern California. <em>The Journal of the Acoustical Society of America</em>, <em>117</em>(4_Supplement), 2431-2431. https://doi.org/10.1121/1.4809456</li>
<li>Luo, S., Hu, S., Zhou, G., Yao, H. (2022). Improvement of Frequency–Bessel Phase-Velocity Spectra of Multicomponent Cross-Correlation Functions from Seismic Ambient Noise. <em>Bulletin of the Seismological Society of America</em>, <em>112</em>(5), 2257-2279. https://doi.org/10.1785/0120220027</li>
<li>Takano, T., Nishimura, T., Nakahara, H., Ohta, Y., Tanaka, S. (2014). Seismic velocity changes caused by the Earth tide: Ambient noise correlation analyses of small‐array data. <em>Geophysical Research Letters</em>, <em>41</em>(17), 6131-6136. https://doi.org/10.1002/2014gl060690</li>
<li>Anggono, T., Nishimura, T., Sato, H., Ueda, H., Ukawa, M. (2012). Spatio-temporal changes in seismic velocity associated with the 2000 activity of Miyakejima volcano as inferred from cross-correlation analyses of ambient noise. <em>Journal of Volcanology and Geothermal Research</em>, <em>247-248</em>, 93-107. https://doi.org/10.1016/j.jvolgeores.2012.08.001</li>
<li>Perrin, C., Manighetti, I., Ampuero, J., Cappa, F., Gaudemer, Y. (2016). Location of largest earthquake slip and fast rupture controlled by along‐strike change in fault structural maturity due to fault growth. <em>Journal of Geophysical Research Solid Earth</em>, <em>121</em>(5), 3666-3685. https://doi.org/10.1002/2015jb012671</li>
<li>Meier, M., Heaton, T. H., Clinton, J. (2016). Evidence for universal earthquake rupture initiation behavior. <em>Geophysical Research Letters</em>, <em>43</em>(15), 7991-7996. https://doi.org/10.1002/2016gl070081</li>
<li>Best, M. G., Christiansen, E. H., Silva, S. L. d., Lipman, P. W. (2016). Slab-rollback ignimbrite flareups in the southern Great Basin and other Cenozoic American arcs: A distinct style of arc volcanism. <em>Geosphere</em>, <em>12</em>(4), 1097-1135. https://doi.org/10.1130/ges01285.1</li>
<li>Jolly, A. D., Lokmer, I., Thun, J., Salichon, J., Fry, B., Chardot, L. (2017). Insights into fluid transport mechanisms at White Island from analysis of coupled very long-period (VLP), long-period (LP) and high-frequency (HF) earthquakes. <em>Journal of Volcanology and Geothermal Research</em>, <em>343</em>, 75-94. https://doi.org/10.1016/j.jvolgeores.2017.06.006</li>
<li>Maury, J., Ide, S., Cruz‐Atienza, V. M., Kostoglodov, V. (2018). Spatiotemporal Variations in Slow Earthquakes Along the Mexican Subduction Zone. <em>Journal of Geophysical Research Solid Earth</em>, <em>123</em>(2), 1559-1575. https://doi.org/10.1002/2017jb014690</li>
<li>Geng, Z., Bonnelye, A., Chen, M., Jin, Y., Dick, P., David, C. (2017). Elastic Anisotropy Reversal During Brittle Creep in Shale. <em>Geophysical Research Letters</em>, <em>44</em>(21). https://doi.org/10.1002/2017gl074555</li>
<li>Picozzi, M., Bindi, D., Zollo, A., Festa, G., Spallarossa, D. (2019). Detecting long-lasting transients of earthquake activity on a fault system by monitoring apparent stress, ground motion and clustering. <em>Scientific Reports</em>, <em>9</em>(1), 16268-16268. https://doi.org/10.1038/s41598-019-52756-8</li>
<li>Hayward, K. S., Cox, S. F. J. (2017). Melt Welding and Its Role in Fault Reactivation and Localization of Fracture Damage in Seismically Active Faults. <em>Journal of Geophysical Research Solid Earth</em>, <em>122</em>(12), 9689-9713. https://doi.org/10.1002/2017jb014903</li>
<li>Chen, G., Chen, J., Tape, C., Wu, H., Tong, P. (2023). Double‐Difference Adjoint Tomography of the Crust and Uppermost Mantle Beneath Alaska. <em>Journal of Geophysical Research Solid Earth</em>, <em>128</em>(1). https://doi.org/10.1029/2022jb025168</li>
</ol>
</article>