Hazard zonation for potential earthquake-induced landslide in the eastern East Kunlun fault zone

Lifu Zheng; Guochao Fu; Guichun Luo

doi:10.1515/geo-2022-0704

Artikel Open Access

Hazard zonation for potential earthquake-induced landslide in the eastern East Kunlun fault zone

Lifu Zheng , Guochao Fu und Guichun Luo

Veröffentlicht/Copyright: 9. Oktober 2024

Veröffentlicht von

Veröffentlichen auch Sie bei De Gruyter Brill

Manuskript einreichen Informationen für Autor*innen

Aus der Zeitschrift Open Geosciences Band 16 Heft 1

Abstract

Based on probabilistic seismic hazard analysis, the seismic landslide hazard research considers the spatial and temporal distribution characteristics of seismic peak ground acceleration, which integrates the factors such as seismic intensity, location, and recurrence time. The occurrence of future earthquakes has certain randomness. This article presents the landslide hazard zoning of the eastern Kunlun fault zone and its surrounding faults, which is carried out under the action of horizontal ground motion with certain exceeding probability. According to the geological structure and seismicity characteristics of the study area, the potential source is divided. Based on the seismic hazard analysis and Newmark cumulative displacement evaluation model, the seismic landslide hazard in the study area is analyzed. The landslide probability is taken as the risk index. The seismic landslide hazard can be divided into five grades: extremely low-prone area, low-prone area, medium-prone area, high-prone area, and extremely high-prone area. In the results of seismic landslide risk zoning given in this article, the surrounding areas of Tazang fault and Minjiang fault are high-risk areas, which should be paid attention to.

Keywords: seismic landslide; seismic hazard analysis; Newmark model; potential seismic source zone

1 Introduction

Seismic landslides, occurring when mountains destabilize due to strong shaking during earthquakes, constitute a significant secondary disaster, posing a grave threat to human lives and property. These events wield an immense socioeconomic impact, leading to disrupted traffic, infrastructure destruction, and river blockages, resulting in substantial losses [1]. In mountainous regions, the damage inflicted by seismic landslides can surpass that directly caused by earthquakes [2]. China, a nation frequently experiencing earthquakes, faces notable susceptibility to severe landslide disasters triggered by these events. The western multi-earthquake region stands particularly vulnerable to numerous seismic landslides, as evidenced by the magnitude-8.0 Wenchuan earthquake of 2008, which triggered approximately 50,000 landslides and claimed 20,000 lives [3,4]. Such landslides entail loss of life and property, critical infrastructure damage, and the emergence of secondary hazards that may exceed the earthquake’s magnitude. Hence, it is imperative to thoroughly investigate earthquake-induced landslides to prioritize factors influencing rescue decisions and the selection of rescue routes during post-earthquake emergency responses.

Research on seismic landslide hazard assessment encompasses both qualitative and quantitative methods. The qualitative approach primarily focuses on the examination of early seismic landslides. Notably, Keefer pioneered this approach by analyzing 40 historical seismic events occurring worldwide between 1958 and 1977 [5]. Through an examination of seismic geologic hazard motion, internal rupture, and geological context, Keefer established the initial correlation curve between the maximum area of seismic landslide hazard and earthquake magnitude. Subsequent studies by Rodriguez delved into the impact of earthquake density, affected area, landslide count, and types on maximum landslides [6]. Li derived an approximate relationship between landslide area and earthquake magnitude through regression calculations based on regional geological, and geomorphological characteristics in China [7]. Moreover, Sun conducted statistical analyses on historical earthquake data in China, synthesizing the correlation between the distribution area of earthquake-induced landslides and magnitude, epicenter distance, lithology, and topographic slope [8]. These investigations have profoundly influenced the field and garnered widespread citation.

Various quantitative methodologies were employed by the researchers to assess the susceptibility of earthquake-induced landslides. These methodologies encompass statistical approaches grounded in mathematical models, alongside the Newmark model. The authors utilized the neural network methodology for evaluating the susceptibility of earthquake-induced landslides, as detailed by Pradel and co-workers [9]. Furthermore, the vulnerability of landslides was assessed by the authors through the utilization of support vector machine (SVM) and logistic regression (LR) models, incorporating different percentages of landslide sample points, as demonstrated by Hu et al. [10]. Some scholars appraised the landslide hazard after the Yushu earthquake using a LR model, yielding the distribution of the landslide hazard index, as illustrated in the investigation by Xu and Xu [11].

In 1965, Newmark introduced a sliding block model to assess dam stability during seismic events [12]. Subsequently, this model underwent a series of mechanical model-based investigations aimed at quantitatively evaluating the hazard posed by regional seismic landslides. Utilizing the Newmark displacement model, Jibson groups [13,14] developed seismic landslide hazard maps for the Oat Mountain region, proximate to the epicenter of the Northridge earthquake in California, USA, and the Anchorage area, respectively. The outcomes of these studies revealed a strong correlation between the assessments derived from the Newmark model and the actual distribution of landslides, demonstrating promising results. Notably, the analysis of seismic parameters, particularly PGA, in seismic landslide hazard assessment has garnered significant attention among the various methodologies employed in studying seismic landslides.

Several scholars have extensively researched the assessment of seismic landslide risk using the Newmark model [15,16,17,18]. Noteworthy among these studies is the work of Jibson and Michael, who created a seismic landslide risk zoning map for the Aras region in the United States [17]. Similarly, Chen developed a seismic landslide hazard zoning map for the Yadong area in Tibet, at a 2% probability of exceedance (POE) in 50 years [18].

This article integrates seismic hazard analysis with the Newmark cumulative displacement model to assess landslide hazard and probability. Using the eastern segment of the Kunlun fault zone as a case study, seismic landslide hazard within the region is evaluated for a risk level of 10% POE in 50 years. The findings offer valuable insights for seismic landslide risk assessment in land use planning.

2 Framework

The research framework employed in this study is shown in Figure 1, which is made of three components as follows:

Probabilistic seismic hazard analysis

PSHA assesses the potential seismic accelerations in a given study area by considering the probabilities of occurrence for various potential seismic sources. The results of PSHA yield peak ground acceleration (PGA) values, which are essential for calculating seismic landslide displacement.
Seismic landslide hazard zoning

The central component of this study framework is seismic landslide hazard zoning, which integrates parameters related to the strength of rock and soil masses. This integration is based on the regional geological map and the digital elevation model (DEM) of the study area. Subsequently, these parameters are used to calculate the factor of safety and critical acceleration for landslides.
Seismic landslide displacement

Figure 1

Research framework of this study.

The displacement of seismic landslides is determined through a combination of geotechnical parameters and seismic hazard indicators, such as PGA and acceleration coefficient. This model utilizes a probabilistic approach to assess the probability of landslide occurrence.

3 Regional geological background of study area

The Eastern Kunlun Fracture Zone is an essential component of the boundary of the Bayankara Block, influenced by the concurrent processes of Indian Plate subduction and extrusion, along with the eastward migration of material from the Qinghai–Tibet Plateau, as depicted in Figure 2a. This block exhibits significant crustal activity within its confines and along its peripheries [19,20,21,22,23,24,25,26]. The primary fault branches comprising the eastern segment of the Eastern Kunlun fault zone include the Tazang fault, Longriba fault, Minjiang fault, Tiger fault, the adjacent Bailongjiang fault, and the He Guanggai Mountain–Dieshan fault. All these faults have demonstrated activity since the late Quaternary period, as evidenced by previous investigations [27,28,29,30,31,32]. The study area resides within the northern section of China’s north–south seismic belt and is identified as one of China’s active seismic regions. This region has a history of frequent seismic events, including recent earthquakes like the magnitude-7.0 Jiuzhaigou earthquake on August 8, 2017, as well as past occurrences such as the magnitude-6.5 Songpan earthquake in 1973 and the magnitude-6.5 Songpan earthquake in 1748 (Figure 2b). These seismic events have triggered significant landslides, notably the Jiuzhaigou earthquake, which triggered numerous landslides, severely affecting the tourism economy of the Jiuzhaigou and Aba Prefecture areas [33].

Figure 2

Regional topography, active tectonics, and seismic distribution map.

4 Probabilistic seismic hazard analysis

Earthquakes and ground shaking are inherently unpredictable events that necessitate expression in terms of probabilities for accurate prediction [34,35,36]. The intensity and frequency of seismic activity exhibit variability across different regions. Conducting a seismic hazard analysis specific to the study area offers a foundational framework for delineating seismic landslide hazard zones.

Assessing seismic hazard at a construction site necessitates comprehending the local seismic activity, predicting future earthquake magnitudes, and quantifying the intensity of ground shaking that poses a risk to the site. Nonetheless, the inherent unpredictability of earthquakes and their resultant ground shaking events poses a significant challenge to precise prediction. To mitigate this challenge, PSHA has emerged as an effective approach for expressing seismic risk probabilistically. Recently, probabilistic analysis was utilized to determine ground shaking parameters at a specific site with a defined POE, as documented in previous studies [37–39].

In this study, PSHA of the eastern part of the eastern Kunlun fault zone is carried out based on the classification of potential hypocenters, and the PGA for a risk level of a 10% POE in 50 years is calculated.

4.1 Classification of potential earthquake sources in the study area

The primary technique employed for seismic hazard analysis in China is Chinese probabilistic seismic hazard analysis (CPSHA), as outlined in the Seismic Ground Motion Parameters Zonation Map of China (2015) [40]. This approach adeptly accounts for the spatiotemporal variability of seismic activity [27]. The main seismic hazard parameters (the upper limit of earthquake magnitude [ M u ], b value, and annual occurrence of earthquakes [magnitude 4.0 and higher] ν 4 ) in the region are listed in Table 1. This study adopts the attenuation relationship in the Sichuan–Tibet area [41]. The potential seismic sources in the study area are shown in Figure 3.

Table 1

Seismic parameters of the seismic statistical zones

Seismic zone	M u	b	ν 4
Longmenshan earthquake belt	8.0	0.72	4.76

Figure 3

Potential seismic sources in the study area.

4.2 Main seismic hazard parameters in the study area

Several studies have indicated that earthquakes can trigger landslides by subjecting the slopes to extra forces resulting from the back-and-forth movement of the ground during shaking, which disrupts the stability of the slope [42]. The PGA of bedrock was evaluated by employing a suitable attenuation relationship and conducting seismic hazard analysis computations within the study area. The average PGA was then translated and arranged using the scientific approach and research outcomes of the Seismic Ground Motion Parameters Zonation Map of China (2001) [43].

In assessing seismic risk at a given site, two key indicators are the likelihood of ground motion acceleration and the POE. However, challenges arise in directly reconciling the Newmark model displacement with ground motion acceleration POE. This arises from the distinction that the latter quantifies the probability of a ground motion parameter surpassing a specific threshold at a particular site due to at least one potential earthquake nearby [13], yet it incompletely encapsulates the probabilities of earthquake occurrences. To address this, a probability model of earthquake occurrence is necessary, guided by the following principle:

Basic assumptions of the Cornell model about seismicity [44].

In the proximity of the study site, multiple earthquake sources exist (assuming a total number of N s ). Each potential source zone has an upper limit of magnitude ( M maxl ) and an average annual occurrence rate of earthquakes ( ν i ), whereby the ratio of large to small earthquakes is represented by b l . Uniform distribution characterizes the occurrence of earthquakes within each potential source zone, while the magnitude distribution of earthquakes is yet to be determined.
(1) f M ( l ) ( m ) = β l e − β l ( m − m 0 ) 1 − e − β l ( M maxl − m 0 ) m 0 ≤ m < M maxl 0 other wise ,
where β l = 2.3 b l .
The average annual occurrence rate of the potential seismic source area should be

(2) v = ∑ l = 1 N l v l .

Then, the probability of n earthquakes occurring in the area under consideration is computing as

(3) P ( N = n ) = ( vt ) n n ! e − vt ( n = 0 , 1 , 2 , … n ) .

Given the independent distribution of characteristics for each earthquake P T ( I = i ) , as assumed in several models of earthquake acceleration beyond probability calculations, the likelihood of n earthquakes taking place in a given region and the corresponding ground acceleration value being i is estimated to be 1 − ( 1 − P I ( I = i ) ) n . When all potential earthquakes are considered, this probability can be inferred as

(4) P T ( I = i ) = ∑ n = 0 + ∞ ( 1 − ( 1 − P I ( I = i ) ) n ) ( vt ) n n ! e − vt .

Based on the aforementioned equation, the calculation formula for the probability of occurrence of ground acceleration at the site can be derived as

(5) P T ( I = i ) = 1 − e − vt P I ( I = i ) .

This study provides an assessment of PGA within study area using a finely resolved grid with a cell dimension of 0.1 × 0.1° and a total of 21,671 calculated field points (Figure 3). Due to the significant influence of topography on seismic motion, it is necessary to correct the PGA calculated for the bedrock to obtain the PGA on the slope. Based on some Chinese standards [48,49], we calculated the amplification effect of the local prominent terrain for PGA using the following empirical equation:

(6) K = 1 + ξ α ,

where K is the amplification factor for the seismic impact; α is the increase in seismic motion parameters, which can be approximated from the slope height and angle; ξ is an additional adjustment coefficient and is taken as 1.0 for simplicity.

The results, as shown in Figure 4, indicate that the PGA values in the study area range between 91 and 279 (gal) for a 10 POE in 50 years.

Figure 4

Peak acceleration zoning map (10% POE in 50 years).

5 Seismic landslide hazard zoning

5.1 Newmark model

In 1965, Newmark introduced a novel technique for evaluating rigid slip to assess slope displacements and resulting damage following seismic events [12]. This approach involves observing slope displacement when ground shaking acceleration exceeds the critical acceleration threshold. The sliding block analysis technique hinges on determining both the critical acceleration of the slope and that of the site, as illustrated in Figure 5, and discussed in the literature [25].

Figure 5

Sliding-block model used for Newmark analysis (adapted from Jibson et al. [14]).

Regional analysis often employs the infinite slope model for computing critical acceleration [12]. However, the simplified Newmark model, introduced by Wilson and Keefer, is more prevalent in regional landslide seismic hazard assessments [14]. This model simplifies horizontal ground shaking acceleration to be input parallel to the slope of the slip surface, thereby facilitating the calculation of the critical acceleration at the slip surface. The static factor of safety of the slope is predicted via infinite slope analysis under the assumption that ground vibrations align parallel to the slope. The flow acceleration can be calculated as follows:

(7) F s = c ′ γ t sin α + tan φ ′ tan α 1 − m γ w γ ,

(8) a c = ( F s − 1 ) g ⋅ sin α ,

where a c is the critical acceleration (g); F s is the static factor of safety; α is the slope angle of the sliding surface (approximated by the slope angle of the slope body) (°); φ ′ is the angle of effective friction (°); c ′ is the effective cohesion (MPa); γ is the material density of the slope body ( N · m − 3 ); m is the thickness ratio of the sliding bars by water infiltration; g is the acceleration due to gravity; t is the thickness of the slope body (m), referring to the typical values of the thickness of the sliding body by domestic and foreign scientists [45,46,47]; the thickness of the sliding body t = 3 m is assumed; and γ w is the unit weight of water ( N · m − 3 ).

When using the critical acceleration a c to calculate the cumulative displacement of the slider, the ground motion acceleration time-history curve data is also required (Figure 6a); the acceleration value smaller than a c part of the slider does not produce displacement, while the acceleration value larger than a c acceleration part of time t can be integrated once to get the curve of the slider speed (Figure 6b) and then the speed of the slider at time t for an integration, you can get the cumulative displacement–time curve of the slider (Figure 5c), as shown in equation (9).

(9) D n = ∫ t ∫ t [ a ( t ) − a c ] d t ,

where D n is the cumulative displacement of the slide, and a ( t ) is the value of the ground vibration acceleration at time t .

Figure 6

Diagram showing the cumulative displacement algorithm principle of Newmark model.

In practice, the application of the method in assessing seismic landslide hazard is limited by the limited number of seismic stations and strong seismic records, even in potential seismic source areas. Therefore, many scientists have established the relationship between ground shaking parameters and seismic landslide displacements [14].

5.2 Regional engineering geological rock formations

The distribution of seismic landslides is intricately tied to lithology, as it not only affects the propagation of landslides but also significantly influences the types of landslides present.

The stratigraphic composition of the study area exhibits considerable complexity, with the classification of engineering geological rock types conducted based on the 1:200,000 regional geological map. The classification primarily adhered to some standards in China [48,49]. Considering the hardness, integrity, geological genesis, and lithological composition of rocks in the study area, alongside contemporary engineering rock type investigations, the study area was delineated into five rock types (Figure 7): hard (Ⅰ), harder (Ⅱ), softer (Ⅲ), weaker (Ⅳ), and loose (Ⅴ). The hard rock type (Ⅰ) predominantly comprises Silurian siliclastic rocks of the Zhouqu Formation, Triassic orthoclase, amphibolite, quartz veins, and granite of the Luo Period. The harder rock group (Ⅱ) consists mainly of shale from the Triassic Zagashan Formation. The softer rock type (Ⅲ) is primarily characterized by Lower Tertiary quartz sandstone. The weak rock type (Ⅳ) is sporadically distributed and comprises Lower Tertiary sandstone. The loose rock type (Ⅴ) primarily consists of Quaternary loose gravel–clay sediments, found predominantly in the Ruoerge Basin. To obtain the slopes’ static factor of safety in equation (7), it is necessary to consider the anti-sliding effects of the effective cohesion and effective friction angle of the landslide, as well as the weakening effect of groundwater on the friction angle. Due to the lack of detailed measured rock mass structural surface strength parameters, empirical values of various rock types are used (Table 2) [50,51,52].

Figure 7

Engineering geological rock formations map of the study area.

Table 2

Seismic parameters of the seismic statistical zones

Rock types	c ′ ( MPa )	φ ′ ( °)	γ / kN· m − 3
I	0.035	28	27.5
II	0.04	26	26.5
III	0.025	18	25.5
IV	0.015	11	23.5
V	0.01	12	21.5

5.3 Slope of the study area

The regional slope data for the eastern segment of the East Kunlun Fault Zone were obtained from DEM with a spatial resolution of 30 meters by 30 meters (accessible at http://www.gscloud.cn). This dataset was generated by calculating the ratio between the maximum vertical disparity between adjacent grid cells and the corresponding horizontal extent (Figure 8). Empirical assessments indicate that terrains with slopes less than 10 degrees typically exhibit a high level of stability, with significant landslides being rare. Therefore, slope calculations have been excluded for areas falling within this range [14].

Figure 8

Topographic slope map of the study area.

5.4 Slope static stability factor and critical acceleration

The distribution of the static factor of safety ( F s ) and critical acceleration ( a c ) across the study area can be determined using equations (7) and (8) (Figures 9 and 10). The critical acceleration value ( a c ) serves as an inherent characteristic of the slope, providing a pivotal indicator of its stability when exposed to seismic forces. Importantly, a strong correlation exists between ground shaking acceleration within the model and the static factor of safety, a relationship that significantly influences slope stability. An increase in ground shaking acceleration in the area corresponds to an elevated risk of landslide occurrence.

$Figure 9 Map showing slope static safety factor F s {F}_{{\rm{s}}} of the study area.$

Figure 9

Map showing slope static safety factor F s of the study area.

$Figure 10 Map showing critical acceleration a c {a}_{{\rm{c}}} of the study area.$

Figure 10

Map showing critical acceleration a c of the study area.

The critical acceleration demonstrates a strong correlation with the geological, geotechnical characteristics, and the morphology of slopes in the studied area. Specifically, softer geotechnical materials composing the slope and steeper terrain lead to a proportional decrease in the a c value. Consequently, a smaller seismic force becomes adequate to initiate slope instability, making the slope more susceptible to earthquake-induced instability. In our investigation, we have classified the critical acceleration within the study area into five distinct intervals, as depicted in Figure 10. These intervals correspond to different susceptibility levels, which are defined as follows: “very high susceptibility (0–0.1 g ), high susceptibility (0.1–0.25 g ), medium susceptibility (0.25–0.39 g ), low susceptibility (0.39–0.57 g ), and very low susceptibility (0.57–0.83 g ),” as outlined in the study by Pan and Li [37].

5.5 Newmark displacement distribution

The Newmark cumulative displacement method was initially introduced by Newmark in 1965 as a means to assess the stability of levees subjected to seismic forces. This method’s fundamental approach involves the quadratic integration of ground shaking acceleration components surpassing the critical acceleration threshold. Consequently, obtaining time curves of ground shaking acceleration for extensive geographical regions and conducting meticulous Newmark displacement integration calculations simultaneously becomes a formidable challenge under such conditions.

In this study, the determination of PGA is grounded upon PSHA for a risk level of 10% POE in 50 years. The calculation is performed employing a regression equation, as proposed by Jibson et al. [14], which relates Newmark displacement ( D n ) to the critical acceleration ( a c ) and incorporates a standard deviation of 0.656.

(10) log D n = 0.125 + log 1 − a c PGA 2.341 × a c PGA − 1.438 .

In accordance with the Newmark displacement classification (Figure 11), the following distribution of areas was observed within the study region: 51.8% of the study area corresponded to D n values ranging from 0 to 5 cm, 8% for D n values within the 5 to 10 cm range, 1.9% for D n values in the 10 to 15 cm range, and 41.9% for D n values exceeding 15 cm.

Figure 11

Newmark displacement distribution in the study area.

5.6 Seismic landslide probability distribution map

Jibson groups evaluated the probabilities of landslides by analyzing the relationship between anticipated Newmark displacement values and the spatial distribution of earthquake-triggered landslides [13]. This analysis was conducted in conjunction with regional assessments of Newmark displacements.

(11) P ( f ) = 0.335 [ 1 − exp ( − 0.048 D n 1.565 ) ] .

The probability of landslide instability, denoted as P ( f ) , is assessed in conjunction with the Newmark displacement ( D n ). The distribution of landslide probability is computed using equation (11), as illustrated in Figure 12.

Figure 12

Probability of seismic-induced landslide occurrence in the study area.

The seismic landslide hazard analysis results within the study area, as illustrated in Figure 12, encapsulate both seismic hazard characteristics and engineering geological conditions. These integrated factors are employed to calculate the prospective seismic landslide probability. This metric assesses the likelihood of seismic landslides occurring in different regions of the study area in response to ground shaking events surpassing a specified probability threshold. Based on prior research findings [53], the seismic landslide hazard is classified into five distinct categories according to probabilistic values: a very low susceptibility zone (0 to 2%), a low susceptibility zone (2 to 4%), a moderate susceptibility zone (4 to 6%), a high susceptibility zone (6 to 8%), and a very high susceptibility zone (8 to 11%).

The results of the regional seismic landslide hazard zoning analysis indicate that the upper plate of the Tazang Rift, particularly the area spanning from Selang to Majima, the local vicinity within the Minjiang Rift, and the lower plate of the Bailong River Rift, all exhibit notably high hazard zones. In contrast, the seismic landslide hazard zone covering the Longriba Rift and the Guanggai Mountain–Dieshan Rift is less pronounced. This distribution of seismic landslides is consistent with the geographical dispersion of the active rupture zone, primarily due to the compromised structural integrity of the rock surrounding the fracture zone, leading to diminished mechanical strength. Consequently, internal fissures or joints are more likely to form within the rock, with some faults or joints acting as interfaces that facilitate slip surface or landslide formation, as mentioned in the study by Cornell [44]. Moreover, these geological factors, combined with alterations in groundwater flow dynamics, can contribute to a reduction in slope stability. Conversely, susceptibility to seismic landslides is significantly influenced by geological rock type and ground vibration parameters, particularly the PGA.

6 Discussion

In the realm of zoning seismic landslide hazards, it is essential to account for the anticipated earthquakes’ impact. This study seeks to evaluate the seismic landslide hazard in the eastern segment of the East Kunlun Fault Zone, utilizing landslide probability as the primary hazard indicator. This probability is shaped by various factors, including the activity of fracture zones, geological characteristics of engineering rock types, and ground shaking parameters. The probabilistic seismic landslide distribution map within the research domain delineates varying degrees of landslide susceptibility. The study area’s hazard levels are categorized into five distinct classes: “very low susceptibility,” “low susceptibility,” “medium susceptibility,” “high susceptibility,” and “very high susceptibility.” Our findings reveal significantly high-risk areas within the upper plate of the Tazang Rift, specifically extending from Selang to Majiama, the local region of the lower plate of the Minjiang Rift, and the lower plate of the Bailong River Rift. Conversely, the seismic landslide hazard in the vicinity of the Longriba Rift and the Guanggai Mountain–Dieshan Rift is comparatively lower.
The seismic landslide hazard zoning, derived from a comprehensive regional seismic hazard analysis, integrates seismic tectonics, seismic activity, and potential source zones across the study area. This analysis employs a defined level of transcendental probability while considering the influence of all tectonic factors and potential seismic events within the study area on future seismic landslides. As a result, the distribution patterns of seismic landslide hazards depicted in this zoning may differ from those associated with landslides triggered by individual earthquakes. This holistic and comprehensive approach is advantageous for a more inclusive evaluation of prospective seismic landslide hazards within the study area.
Newmark analysis is widely used in the assessment of seismic landslide hazard, while the practice shows that its quantitative evaluation results have limitations [54,55]. In this case study, we aim to gain insights into seismic landslide hazard assessment in the East Kunlun region based on the simplified Newmark displacement model. Although we considered some of most important influencing factors, such as PGA distribution, the amplification effect of topography on ground motion, the evaluation results are not so accurate due to some assumptions in the simplified model and the influence of parameters using empirical values. So in the future study, we will consider the uncertainty of the rock mechanical parameters ( φ ′ , c ′ , γ ) involved in the Newmark model and explore to introduce the Monte Carlo model to analyze the hazard of seismic landslides. Meanwhile, we will consider other influencing factors, such as hydrology, human activities, and landforms, so that the limitations of the Newmark model will be compensated.

7 Conclusion

This study analyzes seismic landslides within the study area by employing seismic hazard analysis and the Newmark cumulative displacement model. The distribution of seismic landslides aligns with the spatial extent of active fracture zones. A seismic landslide hazard index is computed based on the probability of landslide occurrence, categorizing the hazard degree within the study area into five classes ranging from very low to very high susceptibility zones. The results of the hazard classification reveal that the upper plate along the Tazang Rift, specifically from Selang to Majiama, the local area beneath the Minjiang Rift, and the earthquake beneath the Bailong River Rift exhibit extremely high hazard zones. In contrast, the seismic landslide hazard zones surrounding the Longriba Rift and the Guanggai Mountain–Dieshan Rift are comparatively less severe.

Acknowledgement

Authors gratefully acknowledge anonymous reviewers for their thorough reading of this manuscript and for their insightful questions and constructive suggestions, which significantly improved the quality of this article.

Funding Information: This study was supported by the Scientific Research Fund of Institute of Engineering Mechanics, China Earthquake Administration (Grant No. 2022QJGJ04).
Author contributions: Conceptualization: L. Z. and G. F.; methodology: G. F.; programming: G. F. and L. Z.; validation: L. Z. and G. L.; resources: G. L.; data curation: G. F., writing original draft preparation: L. Z. and G. F; writing review and editing: L. Z, G. F. and G. L.; visualization: G. F. and L. Z. All authors have read and agreed to the published version of the manuscript.
Conflict of interest: Authors state no conflict of interest.

References

[1] Huang R. Geohazard Assessment of Wenchuan Earthquake. Beijing: Science Press; 2009. (in Chinese).Suche in Google Scholar

[2] Qi SW, Xu Q, Liu CL. Slope instabilities in the severest disaster areas of 5.12 Wenchuan earthquake. J Eng Geol. 2009;17(1):25–30. (in Chinese).Suche in Google Scholar

[3] Xu C, Dai FC, Xu XW. Wenchuan earthquake-induced landslides: An overview. Geol Rev. 2010;56(6):860–74. (in Chinese).Suche in Google Scholar

[4] Wang XY, Nie GZ, Wang DW. Research on relationship between landslides and peak ground accelerations induced by Wenchuan earthquake. Chin J Rock Mech Eng. 2010;29(1):82–9. (in Chinese).Suche in Google Scholar

[5] Keefer DK. Landslides caused by earthquake. Geol Soc Am Bull. 1984;95(4):406–21.10.1130/0016-7606(1984)95<406:LCBE>2.0.CO;2Suche in Google Scholar

[6] Rodriguez CE, Bommer JJ, Chandler RJ. Earthquake-induced landslides: 1980-1997. Soil Dyn Earthq Eng. 1999;18(5):325–46.10.1016/S0267-7261(99)00012-3Suche in Google Scholar

[7] Li TC. Relationship between earthquake and landslides and prediction of earthquake induced landslides. Landslide Anthology (II). Beijing: The People’s Railway Publishing House; 1979. (in Chinese).Suche in Google Scholar

[8] Sun CS, Cai HW. Developing and distributing characteristics of collapses and landslides during strong historic earthquake in China. J Nat Disasters. 1997;6(1):25–30. (in Chinese).Suche in Google Scholar

[9] Pradel D, Smith PM, Stewart JP, Raad G. Case history of landslide movement during the Northridge earthquake. J Geootech Geoenviron Eng. 2005;131(11):1360–9. (in Chinese).10.1061/(ASCE)1090-0241(2005)131:11(1360)Suche in Google Scholar

[10] Hu DY, Li J, Chen YH. Study on spatial prediction method of landslide disaster based on GIS. J Remote Sens. 2007;6:852–9 (in Chinese).Suche in Google Scholar

[11] Xu C, Xu XW. Logistic regression model and its validation for hazard mapping of landslides triggered by the 8 August 2017 Jiuzhaigou, Sichuan MS7.0 earthquake. Seismol Geol. 2012;3:24–31. (in Chinese).Suche in Google Scholar

[12] Newmark NM. Effect of earthquake on dams and embankments. Geotechnique. 1965;15(2):139–59.10.1680/geot.1965.15.2.139Suche in Google Scholar

[13] Harp EL, Jibson RW. Landslides triggered by the 1994 Northridge, California, earthquake. Bull Seismol Soc Am. 1996;86:S319–32.10.1785/BSSA08601BS319Suche in Google Scholar

[14] Jibson RW, Harp EL, Michael JA. A method for producing digital probabilistic seismic landslide hazard maps. Eng Geol. 2000;58(3):271–89.10.1016/S0013-7952(00)00039-9Suche in Google Scholar

[15] Gallen SF, Clark MK, Godt JW, Roback K, Niemi NA. Application and evaluation of a rapid response earthquake-triggered landslide model to the 25 April 2015 Mw 7.8 Gorkha earthquake. Nepal Tectonophys. 2017;714–715:173–87.10.1016/j.tecto.2016.10.031Suche in Google Scholar

[16] Tapponnier P, Zhiqin X, Roger F, Meyer B, Arnaud N, Wittlinger G, et al. Oblique stepwise rise and growth of the Tibet plateau. Science. 2001;294(5547):1671–7.10.1126/science.105978Suche in Google Scholar

[17] Jibson RW, Michael JA. Maps showing seismic landslide hazards in Anchorage, Alaska. Reston, VA, USA: US Geological Survey; 2009.10.3133/sim3077Suche in Google Scholar

[18] Chen L, Mei L, Zeng B, Yin K, Shrestha DP, Du J. Failure probability assessment of landslides triggered by earthquakes and rainfall: a case study in Yadong County, Tibet, China. Sci Rep. 2020;10:16531. 10.1038/s41598-020-73727-4.Suche in Google Scholar

[19] Zhang PZ, Shen Z, Wang M, Gan W, Bürgmann R, Molnar P, et al. Continuous deformation of the Tibetan Plateau from global positioning system data. Geology. 2004;32(9):809–12.10.1130/G20554.1Suche in Google Scholar

[20] Beaumont C, Jamieson RA, Nguyen MH, Lee B. Himalayan tectonics explained by extrusion of a low-viscosity crustal channel coupled to focused surface denudation. Nature. 2004;414(6865):738–42.10.1038/414738aSuche in Google Scholar PubMed

[21] Burchfiel BC, Zhiliang C, Yupinc L, Royden LH. Tectonics of the Longmen Shan and adjacent regions, central China. Int Geol Rev. 1995;37(8):661–735.10.1080/00206819509465424Suche in Google Scholar

[22] Kirby E, Harkins N, Wang E, Shi X, Fan C, Burbank D. Slip rate gradients along the eastern Kunlun fault. Tectonics. 2007;26(2):TC2010.1.Suche in Google Scholar

[23] Harkins N, Kirby E, Shi X, Wang E, Burbank D, Chun F. Millennial slip rates along the eastern Kunlun fault: Implications for the dynamics of intracontinental deformation in Asia. Lithosphere. 2010;2(4):247–66.10.1130/L85.1Suche in Google Scholar

[24] Van der Woerd J, Ryerson FJ, Tapponnier P, Meriaux AS, Gaudemer Y, Meyer B, et al. Uniform slip-rate along the Kunlun Fault: implications for seismic behaviour and large-scale tectonics. Geophys Res Lett. 2000;27(16):2353–6.10.1029/1999GL011292Suche in Google Scholar

[25] Li C, Xu X, Wen X, Zheng R, Chen G, Yang H, et al. Rupture segmentation and slip partitioning of the mid-eastern part of the Kunlun fault, north Tibetan Plateau. Sci China Earth Sci. 2011;54:1730–45.10.1007/s11430-011-4239-5Suche in Google Scholar

[26] Xu XW, Wen XZ, Chen GH, Yu GH. Discovery of the longriba fault zone in eastern Bayan Har block, China and its tectonic implication. Sci China. 2008;51(9):1209–23.10.1007/s11430-008-0097-1Suche in Google Scholar

[27] Ren J, Xu X, Yeats RS, Zhang S. Latest quaternary paleoseismology and slip rates of the Longriba fault zone, eastern Tibet: Implications for fault behavior and strain partitioning. Tectonics. 2013;32(2):216–38.10.1002/tect.20029Suche in Google Scholar

[28] Wang WF, Zhu CH, Qing YB, Shan XJ. Research on transverse faults in the Longmenshan fault zone, China. Adv Mater Res. 2014;1010–1012:1380–6.10.4028/www.scientific.net/AMR.1010-1012.1380Suche in Google Scholar

[29] Kirby E, Harkins N, Wang E, Shi X, Fan C, Burbank D. Slip rate gradients along the eastern Kunlun fault. Tectonics. 2007;26:TC2010.10.1029/2006TC002033Suche in Google Scholar

[30] Gao R, Wang H, Li W, Li H. Structure deformation of the Minjiang and Huya fault, the eastern margin of the Tibetan plateau revealed by deep seismic reflection profiles. AGU Fall Meeting Abstracts; 2014. p. T21B–4575.Suche in Google Scholar

[31] Yang W, Huang X, Zhang C, Si H. Deformation behavior of landslides and their formation mechanism along Pingding-Huama active fault in Bailongjiang river region. Jilin Daxue Xuebao (Diqiu Kexue Ban). J Jilin Univ. 2014;44(2):574–83.Suche in Google Scholar

[32] Nie Z, Wang DJ, Jia Z, Yu P, Li L. Fault model of the 2017 Jiuzhaigou Mw 6.5 earthquake estimated from coseismic deformation observed using global positioning system and interferometric synthetic aperture radar data. Earth Planets Space. 2018;70(1):55.10.1186/s40623-018-0826-4Suche in Google Scholar

[33] Zhang QK, Ling SX, Li XN, Sun CW, Xu JX, Huang T. Comparison of landslide susceptibility mapping rapid assessment models in Jiuzhaigou County, Sichuan province, China. Chinese J Rock Mech Eng. 2020;39(8):1595–1610.Suche in Google Scholar

[34] Hu YX. Seismic safety evaluation. Beijing: Seismological Press; 1999. (in Chinese).Suche in Google Scholar

[35] Chinese seismic ground motion zoning working group. Seismic Zoning Map in China. Beijing: Standards Press of China; 2001. (in Chinese).Suche in Google Scholar

[36] Hu YX. Comprehensive probability method in Seismic Risk Analysis. Beijing: Seismological Press; 1999. (in Chinese).Suche in Google Scholar

[37] Pan H, Li JC. Seismicity model of new generation seismic zoning map. City and Disaster Reduction. Urban and Disaster Reduction 2016;3(3):28–33. (in Chinese). Suche in Google Scholar

[38] Hua P, Jinchen L, Zhiguo L. Study on uncertainties of seismicity parameters b and v4 in seismic statistical zones. 2007;21(3):318–26.Suche in Google Scholar

[39] Pan H. Study on uncertains in the parameters of probabilistic seismic hazard analysis Dissertation. 2000. (in Chinese)Suche in Google Scholar

[40] GB 18306-2015. Seismic ground motion parameters zonation map of China [Standard]. General administration of quality supervision, inspection and quarantine. Beijing: Standardization Administration of China; 2015.Suche in Google Scholar

[41] Xiao L, Yu YX. Seismic intensity attenuation relationship in the western region of China. Earthquake Disaster Prevention and Mitigation Technology. 2011;6(4):358–71. (in Chinese)Suche in Google Scholar

[42] Wang XY, Wang CL, Zhang CC. Analysis correlation between ground motion parameters and earthquake disaster: A case study on earthquake-induced landslide hazard in Wenchuan earthquake area.Suche in Google Scholar

[43] Gao MT. GB 18306-2015 zoning map of ground motion parameters in China Propaganda and implementation of teaching materials. Beijing: China Quality Inspection Press; 2015. (in Chinese).Suche in Google Scholar

[44] Cornell CA. Engineering seismic risk analysis. Bull Seismol Soc Am. 1968;58(5):1583–606.10.1785/BSSA0580051583Suche in Google Scholar

[45] Wang T. Study on seismic landslide hazard assessment in Wenchuan earthquake severely afflicted area. Beijing: Chinese Academy of Geological Sciences; 2010. (in Chinese).Suche in Google Scholar

[46] Jibson RW. Regression models for estimating coseismic landslide displacement. Eng Geol. 2007;91(2–4):209–18.10.1016/j.enggeo.2007.01.013Suche in Google Scholar

[47] Hübner B, Mahler A. Analysis of seismic fragility functions of highway embankments. Period Polytech Civil Eng. 2020 Jan;64(4):1162–9, https://pp.bme.hu/ci/article/view/16483.10.3311/PPci.16483Suche in Google Scholar

[48] Ministry of Construction, People’s Republic of China. Grading standards for engineering rock mass [Standard]. Beijing; 1994. (in Chinese).Suche in Google Scholar

[49] Ministry of Construction of the People’s Republic of China. Code for investigation of geotechnical engineering (GB 50021-2001). Beijing: China Architecture & Building Press; 2001. (in Chinese).Suche in Google Scholar

[50] Veylon G, Luu LH, Mercklé S, Mercklé S, Bard PY, Delvallée A, et al. A simplified method for estimating Newmark displacements of mountain reservoirs. Soil Dyn Earthq Eng. 2018;100:518–28. 10.1016/j.soildyn.2017.07.003.Suche in Google Scholar

[51] Ma S, Xu C. Assessment of co-seismic landslide hazard using the Newmark model and statistical analyses: a case study of the 2013 Lushan, China, Mw6.6 earthquake. Nat Hazards. 2019;96:389–412. 10.1007/s11069-018-3548-9.Suche in Google Scholar

[52] Chousianitis K, Del Gaudio V, Kalogeras I, Ganas A. Predictive model of Arias intensity and Newmark displacement for regional scale evaluation of earthquake-induced landslide hazard in Greece. Soil Dyn Earthq Eng. 2014;65:11–29. 10.1016/j.soildyn.2014.05.009.Suche in Google Scholar

[53] Chen L, Mei L, Zeng B, Yin K, Shrestha DP, Du J. Failure probability assessment of landslides triggered by earthquakes and rainfall: a case study in Yadong County, Tibet, China. Scientific Reports. 2020;10(1):16531.10.1038/s41598-020-73727-4Suche in Google Scholar PubMed PubMed Central

[54] Fell R, Corominas J, Bonnard C, Cascini L, Leroi E, Savage WZ. Guidelines for landslide susceptibility, hazard and risk zoning for land use planning. Eng Geol. 2008;102(3–4):85–98.10.1016/j.enggeo.2008.03.022Suche in Google Scholar

[55] Feng XJ. Discussion on the relationship between active faults and landslides and collapses. J Chang’ an Univ. 1990:92–9. (in Chinese).Suche in Google Scholar

Received: 2024-04-21

Revised: 2024-06-25

Accepted: 2024-09-04

Published Online: 2024-10-09

This work is licensed under the Creative Commons Attribution 4.0 International License.

Artikel in diesem Heft

https://doi.org/10.1515/geo-2022-0704

Schlagwörter für diesen Artikel

seismic landslide; seismic hazard analysis; Newmark model; potential seismic source zone

Creative Commons

BY 4.0