Using fluvial terraces as distributed deformation offset markers: Implications for deformation kinematics of the North Qilian Shan Fault

Highlights

• Distributed deformation study across the North Qilian Shan Fault

• Dated strath river terraces used as markers to assess cumulative deformation patterns

• Trishear model used to explain fault geometries

• Fault splays are bending-moment faults linked to fault propagation folding.

• Vertical slip rates increase into the Holocene, suggesting large earthquake potential.

Abstract

Knowledge of the kinematics of thrusts is the key to understanding mountain building processes in compressive ranges; however, relatively little attention has been paid to distributed deformation (on-fault and off-fault such as rotation and warping) across a thrust fault. Distributed deformation is widespread along the northern front of the Qilian Shan, in the NE Tibetan Plateau. In the Daciyao River, of the central North Qilian Shan, the East Yumu Shan Fault (EYF) splays into four parallel faults at the piedmont. The Daciyao River flows across the deformation zone and its terrace landform record provides a valuable geomorphic marker for reconstructing the cumulative deformation. Terrace deformation across the fault splays was estimated with measured longitudinal profiles using differential GPS, and the fault slip was reconstructed by applying the trishear model, a widely applied model used to explain the geometry of basement-related structures. At least 10 strath terraces are formed with bedrock bases developed into Neogene red beds along the mountain front. The age of one of the terraces (T8) was constrained by Optically Stimulated Luminescence (OSL) dating of the overlying aeolian loess, which provide a minimum age (68 ka) for the age of terrace abandonment. The survey data and trishear model results suggest that the fault splays are bending-moment faults (BMF), originating from a fault propagation fold. Based on the fault and fold geometry across the EYF, the total vertical slip rate is estimated as 0.9–1.2 mm/yr. Slip rates on the EYF are consistent along-strike during the Late Pleistocene but they are twice as high during the Holocene, indicating a high potential for large earthquakes along this fault. This work also indicates that to calculate accurate total slip rate, terrace heights must be measured outside of the distributed broad zone.

Introduction

The estimation of reliable slip rates on active fault systems is central to determining seismic recurrence intervals (Budding et al., 1991; Chen et al., 2007; Lee et al., 2001; Wallace, 1970; Yeats, 1988), earthquake magnitude (Wells and Coppersmith, 1994), and long-term crust deformation patterns (Dolan et al., 2016; Friedrich et al., 2003; Gold et al., 2017; Hetzel et al., 2019; Mouslopoulou et al., 2009). However, faulting in continental crust can create complex surface rupture patterns, including combinations of discrete and distributed deformation (on- and off-fault deformation), which has been widely observed along fault zones (McCalpin, 2009; Yeats et al., 1997). Thus, in order to obtain comprehensive knowledge of the entire thrusting process and assess the amount of slip, it is necessary to conduct a survey of the folding and faulting across a wide fault zone, rather than focusing on a single fault scarp.

In recent decades, several studies have attempted to estimate the fault slip using geometries of folded fluvial surfaces combining with fault-related fold models (Amos et al., 2007; Haghipour et al., 2012; Hu et al., 2015, Hu et al., 2017; Lavé and Avouac, 2000; Scharer et al., 2006; Thompson et al., 2002; Wang et al., 2020; Wilson et al., 2009; Zhong et al., 2020). There are many advantages to using fluvial surfaces to constrain the kinematics of active structures (Lavé and Avouac, 2000; Pan et al., 2013; Wilson et al., 2009), e.g., (1) fluvial terraces are commonly preserved in uplifted regions and are accessible, (2) the initial terrace surface slope can be evaluated by comparison with the modern riverbed, and (3) terraces are convenient for dating. In order to capture the total slip, such studies commonly measure the deformation of fluvial terraces surfaces over several to a few tens of kilometers. Larger spatial surveys will capture more slip rather than trench- or scarp-based studies (Gold et al., 2006). These studies have shown that fluvial surfaces combined with a fault-related fold model are valuable and efficient means of providing a more comprehensive picture of deformation kinematics. In areas dominated by active folds, secondary faults (fold-accommodation faults, e.g., bending-moment faults and flexural-slip faults) commonly slice through the tilted surface and produce a series of subparallel geomorphic scarps (Li et al., 2015, Li et al., 2017, Li et al., 2018). Across a fault zone with several fault splays and folds, it remains a challenging task to estimate the thrusting kinematics and the complete deformation rate. On the front limbs of basement-related structures, when the fault tip line propagates from the basement into the weaker sedimentary units, the fault slip commonly dissipates within a triangular deformation zone within the sedimentary units (Erslev, 1991) (Fig. 1a). A widely applied model used to explain the geometry of the triangular deformation zone is Trishear (Fig. 1b), which is characterized by smooth fold profiles with footwall synclines and a cumulative wedge on the forelimb with progressive limb rotation (Erslev, 1991). Sandbox experiments predict that with progressive deformation, a series of secondary faults will form and breakthrough the surface within the trishear zone (Fig. 1a) (Mitra and Miller, 2013). Thus, according the trishear deformation pattern, the secondary faulting may be incorporated into the deformation kinematics.

The growth process of the Tibetan Plateau is key to understanding the formation of mountains and continental plateaus (Meyer et al., 1998; Tapponnier et al., 1990). Along the northeastern margin of the Plateau in northwestern China, basement-related structures associated with reverse faults are a common structural style in the foreland of the Qilian Shan mountain (Guo et al., 1993) (Fig. 2). Along the Qilian Shan mountain front (Fig. 2), the North Qilian Shan Fault (NQF) splays into several branches and cuts Late Quaternary fluvial and alluvial terraces (Cao et al., 2019; Liu et al., 2019b; Ren et al., 2019; Wang et al., 2020; Yang et al., 2018b). However, the kinematics of these secondary faults are poorly constrained, and in addition, the relationship between the secondary faults and the causal thrust is unknown.

Several studies presented Quaternary activity and paleo-earthquakes on the East Yumu Shan Fault (EYF) (Fig. 3). As shown in Fig. 3, the fault becomes a monocline at its southern tip (Cao et al., 2019) and splays into several sub-parallel branches at its northern tip (Ren et al., 2019), while the connection between these fault traces is unknown. Apart from the deformation pattern, the estimated slip rates vary from a minimum of ~0.3–0.4 mm/yr to a maximum of 1.7 mm/yr (Cao et al., 2019; Li et al., 1995; Ren et al., 2019), which is obviously questionable during the similar timescale. Here, for a better understanding of the kinematics of the complex thrusting zone with secondary fault scarps and folds, we estimated the deformation pattern across this fault zone using detailed topographic surveys and trishear modeling. Our study focused on the northern tip of the EYF, where the deformation of a series of discrete faults is extended over many kilometers and recorded by Late Pleistocene/Holocene fluvial terraces along the Daciyao River (Fig. 3).