3D geomechanical modeling of the Xianshuihe fault zone, SE Tibetan Plateau: Implications for seismic hazard assessment

Highlights

• A refined 3D geomechanical model of the Xianshuihe fault zone is established.

• The model provides high quality spatially continuous crustal kinematics and stress.

• The stress state on the Xianshuihe fault is inhomogeneous.

• The Songlinkou-Selaha segment and the Moxi fault are of high seismic hazard.

Abstract

The Xianshuihe fault on the southeastern margin of the Tibetan Plateau is one of the most active intracontinental faults worldwide. At least 16 strong earthquakes (M > 6.5) have occurred on this fault since 1327 CE. To reasonably assess the future seismic hazard of this fault region, it is essential to gain a comprehensive knowledge of the fault kinematics and stress state. In this study, we developed a fine 3D geomechanical model of the Xianshuihe fault and its adjacent area, and obtained a spatially continuous contemporary kinematics and crustal background stress field of this region. Our modeled results show that the slip rate of the Xianshuihe fault is as high as 11 mm/a on the northwestern segment and generally decreases to ~9 mm/a on the southeastern segment (i.e. Moxi fault). The Yulongxi fault, a branch oblique to the Moxi fault, has different slip senses in different segments with a very low horizontal slip rate (<0.6 mm/a). In terms of stress, the study area is dominated by a transtensional faulting stress regime with E-W-trending maximum horizontal stress S_H, indicating the study area is mainly subjected to an E-W compression and N-S extension. Localized normal and thrust faulting regimes appear alternately along the Xianshuihe fault, showing an inhomogeneous stress state. The high normal stresses on the Songlinkou-Selaha segment and the Moxi fault combined with their relatively high slip rates, make us speculate that these two segments have the potential to produce strong earthquakes in the future. A relatively high seismic hazard is also ascribed to the Yulongxi fault based on the analysis of its geological and geodynamic environment.

Introduction

The Xianshuihe fault on the southeastern margin of the Tibetan Plateau is one of the most active intracontinental faults in the world (Fig. 1) (Allen et al., 1991). Geological and geodetic investigations show that this fault is a left-lateral strike-slip fault moving at a high rate of 10–20 mm/a (Allen et al., 1991; Bai et al., 2018; Bai et al., 2021; Gan et al., 2007; Shen et al., 2005), leading to frequent large earthquakes in this area. Since the year 1327, over 16 strong earthquakes of M > 6.5 have been recorded on different segments of the Xianshuihe fault (see Fig. 1), and 8 of these earthquakes reach a magnitude larger than 7.0 (Wen et al., 2008). The latest moderately strong earthquakes are the Kangding M_S 6.3 and M_S 5.8 earthquakes that occurred in 2014, and caused 5 deaths and a direct economic loss of approximately 700 million US dollars (Fang et al., 2015). However, some studies suggested that these two earthquakes did not fully release the energy accumulated in the focal area (Bai et al., 2018; Jiang et al., 2015), and a strong earthquake with a magnitude of M_W 7.0 is still highly possible to strike the Kangding area (Bai et al., 2021).

In recent years, the great area of the Xianshuihe fault has witnessed an increasing number of key infrastructures being constructed, e.g., hydropower stations and the Sichuan-Tibet railway (Fig. 1), accompanied by a growing population. The increased human activities and properties greatly enhanced the demand for seismic hazard assessment for the Xianshuihe fault and its neighboring area. Seismic hazard of a region is often assessed based on knowledge of kinematics and stress state of dominating faults. Fault slip rates are commonly used to investigate the seismic activities of faults and to calculate the recurrence time of potentially damaging earthquakes; the distribution characteristics of stress along a fault plane can be used to directly outline high-stress zones, and provide useful information on the potential hypocenter area of future large earthquakes. Past studies have mainly focused on using fault slip rate (Bai et al., 2018; Bai et al., 2021; Chen et al., 2016; Chevalier et al., 2016) or stress state (Papadimitriou et al., 2004; Parsons et al., 2008; Shan et al., 2013; Toda et al., 2008) to assess the seismic hazard. For example, by using offset geomorphic features near Kangding with ¹⁰Be age date, Bai et al. (2021) obtained a Quaternary horizontal slip rate of 8–13.4 mm/a on the southeastern segment of the Xianshuihe fault and suggested that a high seismic hazard with an M_W 7.0 earthquake exists on that segment, especially near the Kangding city. Papadimitriou et al. (2004) and Shan et al. (2013) calculated the changes of Coulomb Failure Stress (∆CFS) on the Xianshuihe fault caused by historical earthquakes and suggested that in general, the southeastern segment of the Xianshuihe fault has a high seismic hazard due to the positive ∆CFS.

However, using only kinematic or ∆CFS analysis may fail to give a reliable assessment of a fault's seismic hazard. For example, before the 2008 Wenchuan M_S 8.0 earthquake, the Longmen Shan fault zone was assigned to have a moderate-to-low seismic hazard due to its low slip rate of 1–2 mm/a (Zhang et al., 1999). ∆CFS can only identify faults moving toward or away from failure, and using it solely to evaluate the potential of future earthquakes is known to have great uncertainty as the absolute level of crustal stress is generally unavailable (Toda et al., 2008). An important lesson learned from the Wenchuan M_S 8.0 earthquake is that enhancing seismic hazard evaluation requires multi-source information about a fault from comprehensive observations (Zhang, 2013). However, so far the seismic hazard assessment for the Xianshuihe area has been conducted mostly based on a single information source, such as fault slip rate (Bai et al., 2018; Bai et al., 2021), ∆CFS (e.g., Shan et al., 2013), or historical seismic activity (Wen et al., 2008). Little attention has been paid to combining these information sources to give a comprehensive seismic hazard assessment. In addition, the sparse pointwise information on the fault slip rates (Bai et al., 2018; Bai et al., 2021; Chen et al., 2016) and stress measurements (Hu et al., 2017) distributed on the Xianshuihe fault also limits the reasonable seismic hazard assessment.

To improve the seismic hazard assessment, spatially continuous kinematics and stress information of the Xianshuihe fault is required. Fortunately, numerical modeling is such a powerful tool that can obtain the kinematics and stress of large regions simultaneously. In our previous study (Li et al., 2021, Li et al., 2022), a large-scale 3D geomechanical model that covers the eastern Tibetan Plateau was established. Based on the first-order features of these simulations, we have identified that the southeastern segment of the Xianshuihe fault has a stress environment that can generate large earthquakes. However, the geometry of the Xianshuihe fault implemented in the model was simplified as one almost vertical surface, and other secondary faults in/near the fault zone were ignored due to its low resolution. To test the inference of the seismic hazard on the southeastern segment of the Xianshuihe fault, a refined model that specifically focuses on the Xianshuihe fault and its adjacent area is needed. The detailed fault data of the SE Tibetan Plateau recently publicized by the China Seismic Experimental Site (CSES) (Lu, 2019) provide a solid foundation for such refined modeling.

In this paper, we construct a more detailed three-dimensional (3D) geomechanical model including the latest geometric data of the fault system, inhomogeneous rock properties, tectonic forces, and gravity for the Xianshuihe fault and its adjacent area. A higher resolution spatially continuous kinematics and stress state of the study area is obtained and calibrated by comparison with model-independent data. Finally, based on the calibrated model results, we synthetically analyze the seismic hazard in the study area. The geological background and model setup of the study area is presented in Section 2. The modeled results and discussion of the seismic hazard potential in the study area are given in 3 Results, 4 Discussion, respectively. The conclusions are drawn in Section 5.