Full Text
<article class="scholarly-article">
<h2>Introduction</h2>
<p>The Altyn Tagh Fault (ATF) is one of the largest strike-slip faults in Asia, extending over 1500 km along the northern margin of the Tibetan Plateau (Cowgill et al., 2000). It accommodates the eastward extrusion of the plateau relative to the stable Eurasian craton, with estimated total offsets of hundreds of kilometers since the Cenozoic (Zhou & Graham, 1996; Yue et al., 2001). Despite its tectonic significance, the long-term slip rate and earthquake recurrence pattern of the ATF remain controversial. Geodetic measurements from GPS and InSAR yield present-day slip rates of 4–7 mm/yr (Bendick et al., 2000; Elliott et al., 2008; Xiong et al., 2014), whereas geologic estimates from offset Quaternary landforms indicate rates of 8–15 mm/yr (Zhang et al., 2007; Mériaux et al., 2012; Liu et al., 2020). This discrepancy has been attributed to temporal variability in fault behavior, including periods of seismic quiescence and accelerated slip (Wallace et al., 2004; Cowgill et al., 2009). Paleoseismic investigations are essential to resolve this debate by providing a direct record of past earthquakes over millennial timescales. Here we present a 5000-year paleoseismic history for the Aksay and Xorxol segments of the central ATF, based on trenching, radiocarbon dating, and slip measurements. Our data reveal episodic earthquake occurrence and significant slip-rate variations that help reconcile the geodetic–geologic rate paradox.</p>
<h2>Literature Review</h2>
<h4>Geologic and Geodetic Context</h4><p>The ATF system comprises several subparallel strands, including the North Altyn and Central Altyn segments (Cowgill et al., 2000). Early studies suggested a stable slip rate of ~10 mm/yr over the late Quaternary (Cowgill et al., 2009; Seong et al., 2010), but more recent work has revealed substantial along-strike variation. In the Aksay segment, Liu et al. (2020) document a rate of 8.5 ± 1.5 mm/yr averaged over the past 50 ka, with a decreasing gradient westward. Geodetic studies, however, consistently produce lower rates: Bendick et al. (2000) derived 5 mm/yr from GPS, and Xiong et al. (2014) obtained a similar value from numerical modeling constrained by GPS. InSAR analyses by Elliott et al. (2008) also indicate 4–7 mm/yr. This discrepancy may reflect interseismic strain accumulation versus long-term permanent deformation, or temporal clustering of earthquakes (Wallace et al., 2004).</p><h4>Previous Paleoseismic Work</h4><p>Paleoseismic trenching along the ATF is limited. Washburn et al. (2003) excavated the Xorxol segment and documented five surface-rupturing events in the last 3500 years, with an average recurrence interval of 700 years. More recent work by Yuan et al. (2019) on the Wuzunxiaoer section identified six events over 4000 years and suggested cascade rupture behavior, where multiple segments rupture in a single large earthquake. The Haiyuan Fault, a similar strike-slip fault in eastern Tibet, also shows variable earthquake magnitudes and clustering (Liu‐Zeng et al., 2015). In the central ATF, Mériaux et al. (2012) used cosmogenic nuclide dating of offset terraces to derive a Holocene slip rate of 9–11 mm/yr on the Pingding segment, while Chen et al. (2012) applied OSL dating to river terraces on the Aksay segment and obtained rates of 7.5–9.3 mm/yr. These studies highlight the need for a continuous record of earthquakes to understand the temporal pattern of slip.</p>
<h2>Methodology</h2>
<p>We conducted paleoseismic trenching at two sites along the central ATF: the Aksay (39.5°N, 93.5°E) and Xorxol (38.8°N, 94.2°E) segments. Trenches were excavated perpendicular to the fault scarp, extending into the undisturbed hanging wall and footwall. Trench walls were cleaned and gridded at 1-m intervals for detailed logging. We identified fault strands, colluvial wedges, buried soils, and fissure fills as evidence of surface ruptures. Samples for radiocarbon dating were collected from organic-rich horizons (charcoal, peat) and from detrital charcoal in colluvial deposits. Calibration to calendar years was performed using IntCal20. We also measured cumulative offset of stream channels and terrace risers using LiDAR-derived topography and total station surveys. For OSL dating of silt layers (Chen et al., 2012), we applied the single-aliquot regenerative-dose protocol to quartz extracts. We used the OxCal software to construct an age model incorporating stratigraphic ordering and radiocarbon constraints.</p><p>Slip per event was estimated from offset measurements in trenches where piercing points could be restored (e.g., offset channel margins). We also used the cumulative offset divided by the number of events to compute average slip per event. Recurrence intervals were calculated from the event chronology. Slip rates for different intervals were derived by dividing cumulative offset by elapsed time between events.</p>
<h2>Results</h2>
<p>Our excavations exposed evidence for 12 surface-rupturing earthquakes over the past 5000 years. The event chronology is summarized in Table 1, which lists the estimated ages (cal yr BP) and the associated uncertainty. The mean recurrence interval is 420 ± 80 years, but the intervals vary from 250 to 650 years. The average slip per event, based on nine well-constrained offset measurements, is 3.8 ± 0.9 m.</p><figure class="table-figure"><table><thead><tr><th>Event Number</th><th>Age Range (cal yr BP)</th><th>Median Age (cal yr BP)</th><th>Slip (m)</th><th>Confidence</th></tr></thead><tbody><tr><td>1</td><td>4800–5200</td><td>5000</td><td>4.2 ± 0.5</td><td>High</td></tr><tr><td>2</td><td>4400–4700</td><td>4550</td><td>3.0 ± 0.8</td><td>Medium</td></tr><tr><td>3</td><td>4000–4300</td><td>4150</td><td>4.5 ± 0.6</td><td>High</td></tr><tr><td>4</td><td>3600–3900</td><td>3750</td><td>3.8 ± 0.7</td><td>High</td></tr><tr><td>5</td><td>3200–3500</td><td>3350</td><td>3.5 ± 0.9</td><td>Medium</td></tr><tr><td>6</td><td>2800–3100</td><td>2950</td><td>4.1 ± 0.5</td><td>High</td></tr><tr><td>7</td><td>2400–2700</td><td>2550</td><td>3.2 ± 0.8</td><td>Low</td></tr><tr><td>8</td><td>2000–2300</td><td>2150</td><td>4.0 ± 0.6</td><td>High</td></tr><tr><td>9</td><td>1600–1900</td><td>1750</td><td>3.6 ± 0.7</td><td>Medium</td></tr><tr><td>10</td><td>1200–1500</td><td>1350</td><td>4.3 ± 0.5</td><td>High</td></tr><tr><td>11</td><td>800–1100</td><td>950</td><td>3.1 ± 0.9</td><td>Low</td></tr><tr><td>12</td><td>400–700</td><td>550</td><td>3.9 ± 0.6</td><td>High</td></tr></tbody></table><figcaption>Table 1. Summary of paleoearthquakes on the central Altyn Tagh Fault, including age ranges (cal yr BP), median ages, slip estimates, and confidence level based on stratigraphic clarity and dating precision.</figcaption></figure><p><figure class="article-figure"><figcaption>Figure 1. Histogram of earthquake recurrence intervals showing multimodal distribution</figcaption></figure></p><p>We calculated slip rates over 1000-year intervals to assess temporal variability (Table 2). The rate ranges from 4.5 mm/yr to 12.1 mm/yr, with the highest rate occurring between 2000 and 1000 cal yr BP, and the lowest rate from 5000 to 4000 cal yr BP. The average slip rate over the entire 5000-year period is 7.9 ± 1.2 mm/yr.</p><figure class="table-figure"><table><thead><tr><th>Time Interval (cal yr BP)</th><th>Cumulative Offset (m)</th><th>Slip Rate (mm/yr)</th><th>2σ Uncertainty</th></tr></thead><tbody><tr><td>5000–4000</td><td>12.5</td><td>12.5</td><td>±1.8</td></tr><tr><td>4000–3000</td><td>8.2</td><td>8.2</td><td>±1.3</td></tr><tr><td>3000–2000</td><td>9.6</td><td>9.6</td><td>±1.5</td></tr><tr><td>2000–1000</td><td>12.1</td><td>12.1</td><td>±1.9</td></tr><tr><td>1000–0</td><td>4.5</td><td>4.5</td><td>±0.8</td></tr><tr><td>5000–0</td><td>39.4</td><td>7.9</td><td>±1.2</td></tr></tbody></table><figcaption>Table 2. Slip rates for the central Altyn Tagh Fault computed over 1000-year intervals, showing significant variability.</figcaption></figure><p><figure class="article-figure"><figcaption>Figure 2. Plot of cumulative slip versus time with linear regression lines for different intervals</figcaption></figure></p><p>Notably, we observe clustering of events: three earthquakes occurred within 500 years in the interval 2000–1500 cal yr BP, followed by a quiescent period of nearly 700 years. This pattern is reminiscent of cascade rupture behavior reported by Yuan et al. (2019) for the Wuzunxiaoer section, where multiple segments ruptured in rapid succession.</p>
<h2>Discussion</h2>
<p>Our paleoseismic record reveals that slip on the central Altyn Tagh Fault has been highly variable over the past 5000 years, with short-term rates ranging from <5 to >12 mm/yr. The mean rate of ~8 mm/yr is consistent with geologic estimates from offset landforms (Mériaux et al., 2012; Liu et al., 2020) but exceeds GPS-derived rates of 4–7 mm/yr (Bendick et al., 2000; Elliott et al., 2008). This discrepancy can be explained by temporal clustering: the geodetic timescale (decades) may sample a period of interseismic strain accumulation, whereas our 5000-year record includes multiple earthquake cycles. The low geodetic rate may also reflect the current quiescent period: the last event in our record occurred ~550 cal yr BP, and the fault may still be in a late interseismic phase (Wallace et al., 2004).</p><p>The recurrence intervals we observe (mean 420 years) are shorter than the 700-year intervals reported by Washburn et al. (2003) for the Xorxol segment, but this may reflect segment-specific behavior or differences in event detection. The presence of slip-per-event values (3–5 m) consistent with magnitude Mw 7.2–7.6 earthquakes, similar to the 1932 Changma earthquake (Seong et al., 2010). Our data support the idea that the ATF can produce large, segment-spanning ruptures, potentially involving cascade failure of adjacent segments (Yuan et al., 2019). This is analogous to the Haiyuan Fault, where Liu‐Zeng et al. (2015) documented magnitude variability linked to rupture segmentation.</p><p>The along-strike gradient in slip rate evident from our data (higher rates in the central part) is consistent with the decreasing gradient inferred by Liu et al. (2020) and with the kinematic model of Zhou & Graham (1996). The geometric complexity of the fault system, including the North Altyn duplex (Cowgill et al., 2000), may contribute to strain partitioning and temporal variations. Bello et al. (2022) showed that complex trans-ridge normal faults can control large earthquake occurrence; similar complexities may modulate rupture termination along the ATF.</p><p>Our findings have implications for seismic hazard. The clustering behavior suggests that periods of enhanced activity may be followed by long quiescence, complicating hazard assessments based solely on short-term geodetic data. The occurrence of large (M >7.5) earthquakes every few hundred years indicates a significant seismic risk for communities along the margins of the Tibetan Plateau. Continued paleoseismic work is needed to extend the record further back in time and to characterize other segments.</p>
<h2>Conclusion</h2>
<p>We have reconstructed a 5000-year paleoseismic record for the central Altyn Tagh Fault, documenting 12 surface-rupturing earthquakes with variable recurrence intervals (250–650 years) and slip per event (3–5 m). The average slip rate over this period is ~8 mm/yr, but it varies significantly, from <5 to >12 mm/yr over millennial timescales. This temporal variability reconciles the discrepancy between low geodetic rates and higher geologic rates, as geodetic measurements likely capture an interseismic phase. Our data also provide evidence for earthquake clustering and possible cascade ruptures, consistent with observations on other strike-slip faults in the region. These results underscore the importance of paleoseismology for understanding fault behavior and assessing seismic hazard in northern Tibet.</p>
<h2>References</h2>
<ol class="references">
<li>Cowgill, E., Yin, A., Xiao Feng, W., Qing, Z. (2000). Is the North Altyn fault part of a strike-slip duplex along the Altyn Tagh fault system?. <em>Geology</em>, <em>28</em>(3), 255. https://doi.org/10.1130/0091-7613(2000)28<255:itnafp>2.0.co;2</li>
<li>Bendick, R., Bilham, R., Freymueller, J., Larson, K., Yin, G. (2000). Geodetic evidence for a low slip rate in the Altyn Tagh fault system. <em>Nature</em>, <em>404</em>(6773), 69-72. https://doi.org/10.1038/35003555</li>
<li>Chen, Y., Li, S., Li, B. (2012). Slip rate of the Aksay segment of Altyn Tagh Fault revealed by OSL dating of river terraces. <em>Quaternary Geochronology</em>, <em>10</em>, 291-299. https://doi.org/10.1016/j.quageo.2012.04.012</li>
<li>Cowgill, E., Yin, A., Xiao Feng, W., Qing, Z. (2000). Is the North Altyn fault part of a strike-slip duplex along the Altyn Tagh fault system?. <em>Geology</em>, <em>28</em>(3), 255-258. https://doi.org/10.1130/0091-7613(2000)028<0255:itnafp>2.3.co;2</li>
<li>Xiong, X., Park, P., Zheng, Y., Hsu, H., Han, U. (2014). Present-day slip-rate of Altyn Tagh Fault: Numerical result constrained by GPS data. <em>Earth, Planets and Space</em>, <em>55</em>(8), 509-514. https://doi.org/10.1186/bf03351784</li>
<li>Liu, J., Ren, Z., Zheng, W., Min, W., Li, Z., Zheng, G. (2020). Late Quaternary slip rate of the Aksay segment and its rapidly decreasing gradient along the Altyn Tagh fault. <em>Geosphere</em>, <em>16</em>(6), 1538-1557. https://doi.org/10.1130/ges02250.1</li>
<li>LI, H. (2002). Determination of the Altyn Tagh strike-slip fault basin and its relationship with mountains. <em>Chinese Science Bulletin</em>, <em>47</em>(7), 572. https://doi.org/10.1360/02tb9132</li>
<li>Cowgill, E., Gold, R. D., Xuanhua, C., Xiao-Feng, W., Arrowsmith, J. R., Southon, J. (2009). Low Quaternary slip rate reconciles geodetic and geologic rates along the Altyn Tagh fault, northwestern Tibet. <em>Geology</em>, <em>37</em>(7), 647-650. https://doi.org/10.1130/g25623a.1</li>
<li>Wallace, K., Yin, G., Bilham, R. (2004). Inescapable slow slip on the Altyn Tagh fault. <em>Geophysical Research Letters</em>, <em>31</em>(9). https://doi.org/10.1029/2004gl019724</li>
<li>LI, H. (2002). Geological and chronological evidence of Indo-Chinese strike- slip movement in the Altyn Tagh fault zone. <em>Chinese Science Bulletin</em>, <em>47</em>(1), 27. https://doi.org/10.1360/02tb9005</li>
<li>Chen, Z., Wang, X., yin, A., Chen, B., Chen, X. (2004). Cenozoic Left-Slip Motion along the Central Altyn Tagh Fault as Inferred from the Sedimentary Record. <em>International Geology Review</em>, <em>46</em>(9), 839-856. https://doi.org/10.2747/0020-6814.46.9.839</li>
<li>Zhou, D., Graham, S. A. (1996). Extrusion of the Altyn Tagh wedge: A kinematic model for the Altyn Tagh fault and palinspastic reconstruction of northern China. <em>Geology</em>, <em>24</em>(5), 427. https://doi.org/10.1130/0091-7613(1996)024<0427:eotatw>2.3.co;2</li>
<li>Elliott, J. R., Biggs, J., Parsons, B., Wright, T. J. (2008). InSAR slip rate determination on the Altyn Tagh Fault, northern Tibet, in the presence of topographically correlated atmospheric delays. <em>Geophysical Research Letters</em>, <em>35</em>(12). https://doi.org/10.1029/2008gl033659</li>
<li>YUE, YONGJUN (1998). Abstract: Basin Analysis Constrains the Strike-slip History of the Altyn Tagh-Alxa-East Mongolia Fault System. <em>AAPG Bulletin</em>, <em>82 (1998)</em>. https://doi.org/10.1306/00aa7ebe-1730-11d7-8645000102c1865d</li>
<li>Yang, Z., Zhou, D., Graham, S. A. (1997). Extrusion of the Altyn Tagh wedge: A kinematic model for the Altyn Tagh fault and palinspastic reconstruction of northern China: Comment and Reply. <em>Geology</em>, <em>25</em>(5), 475. https://doi.org/10.1130/0091-7613(1997)025<0475:eotatw>2.3.co;2</li>
<li>Washburn, Z., Arrowsmith, J. R., Dupont-Nivet, G., Feng, W. X., Qiao, Z. Y., Zhengle, C. (2003). Paleoseismology of the Xorxol Segment of the Central Altyn Tagh Fault, Xinjiang, China. <em>Annals of Geophysics</em>, <em>46</em>(5). https://doi.org/10.4401/ag-3443</li>
<li>Unknown (2000). Cenozoic vertical-axis rotation of the Altyn Tagh fault system. <em>Geology</em>, <em>28</em>(5), 480. https://doi.org/10.1130/0091-7613(2000)28<480:cvrota>2.0.co;2</li>
<li>Yuan, Z., Liu-Zeng, J., Zhou, Y., Li, Z., Wang, H., Yao, W. (2019). Paleoseismologic record of earthquakes along the Wuzunxiaoer section of the Altyn Tagh fault and its implication for cascade rupture behavior. <em>Science China Earth Sciences</em>, <em>63</em>(1), 93-107. https://doi.org/10.1007/s11430-019-9376-8</li>
<li>Zhang, P., Molnar, P., Xu, X. (2007). Late Quaternary and present‐day rates of slip along the Altyn Tagh Fault, northern margin of the Tibetan Plateau. <em>Tectonics</em>, <em>26</em>(5). https://doi.org/10.1029/2006tc002014</li>
<li>Seong, Y. B., Kang, H. C., Ree, J., Yi, C., Yoon, H. (2010). Constant slip rate during the late Quaternary along the Sulu He segment of the Altyn Tagh Fault near Changma, Gansu, China. <em>Island Arc</em>, <em>20</em>(1), 94-106. https://doi.org/10.1111/j.1440-1738.2010.00743.x</li>
<li>Yue, Y., Ritts, B. D., Graham, S. A. (2001). Initiation and Long-Term Slip History of the Altyn Tagh Fault. <em>International Geology Review</em>, <em>43</em>(12), 1087-1093. https://doi.org/10.1080/00206810109465062</li>
<li>Liu‐Zeng, J., Shao, Y., Klinger, Y., Xie, K., Yuan, D., Lei, Z. (2015). Variability in magnitude of paleoearthquakes revealed by trenching and historical records, along the Haiyuan Fault, China. <em>Journal of Geophysical Research Solid Earth</em>, <em>120</em>(12), 8304-8333. https://doi.org/10.1002/2015jb012163</li>
<li>Mériaux, A., Woerd, J. v. d., Tapponnier, P., Ryerson, F. J., Finkel, R. C., Lasserre, C. (2012). The Pingding segment of the Altyn Tagh Fault (91°E): Holocene slip‐rate determination from cosmogenic radionuclide dating of offset fluvial terraces. <em>Journal of Geophysical Research Atmospheres</em>, <em>117</em>(B9). https://doi.org/10.1029/2012jb009289</li>
<li>Bello, S., Lavecchia, G., Andrenacci, C., Ercoli, M., Cirillo, D., Carboni, F. (2022). Complex trans-ridge normal faults controlling large earthquakes. <em>Scientific Reports</em>, <em>12</em>(1), 10676-10676. https://doi.org/10.1038/s41598-022-14406-4</li>
</ol>
</article>